Local Area Unemployment Statistics Program

Local Area Unemployment Statistics
Program Manual

..and the model should fit CPS more closely.
Rate
10.0
CPS

MODEL

9.0
8.0

7.0
6.0
5.0
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

U.S. Department of Labor
Bureau of Labor Statistics
August 2010

Table of Contents
1. Local Area Unemployment Statistics Program: Introduction & Overview
History 1-3
Data Sources 1-10
Summary of Estimation Methods 1-12
Publication and Administrative Uses of LAUS Estimates 1-15
2. Inputs to LAUS Estimation: The Current Population Survey
Introduction to the CPS 2-1
CPS Labor Force Concepts and Definitions 2-4
Reliability of CPS Estimates 2-13
Sample Design 2-17
Data Collection 2-23
Estimation Procedures 2-25
3. Inputs to LAUS Estimation: The Unemployment Insurance System
Fe deral Role 3-2
UI Programs 3-15
UI Process 3-17
UI Claims Data for LAUS Estimation 3-23
Differences: UI Data versus the CPS 3-25
BLS Standards for UI Data 3-26
Residency Adjustment System 3-32
Appendix 3-1, LADT Transfer Record Layout 3-35
Appendix 3-2, LADT Agent Summary Report 3-45
Appendix 3-3, LADT Commuter Claims Report 3-46
Appendix 3-4, Residency Adjustment System User’s Guide 3-47
4. Inputs to LAUS Estimation: Establishment Data Sources
The Current Employment Statistics Program 4-1
The Quarterly Census of Employment and Wages Program 4-6
Differences: Establishment Data Sources versus the CPS 4-10
Uses of CES Data in the LAUS Program 4-11
5. Inputs to LAUS Estimation: Census Data
The Decennial Census: Enumerated and Sample -Based Data 5-2
Differences: Census versus CPS/LAUS Estimates 5-5
Uses of Decennial Census Data in LAUS 5-7

6. Development of Statewide Estimates
Background 6-1
Model Structure 6-5
Description of the Employment Model 6-18
Description of the Unemployment Model 6-20
Detailed Description of the Estimation Process 6-22
Official Estimates 6-25
Smoothed Seasonal Adjustment 6-29
Error Measures 6-37
7. LAUS Estimation: Labor Market Area Estimates
Introduction 7-1
Labor Market Area Employment 7-4
Labor Market Area Unemployment 7-20
Monthly Step 3 Ratios 7-31
CPS Agricultural Employment Estimation Factors 7-33
Quarterly Exhaustee Survival Rates 7-41
8. LAUS Estimation: Geography
Introduction 8-1
Standards for Defining Labor Market Areas and Combined Statistical Areas 8-5
Standards for Defining Small Labor Market Areas 8-8
9. LAUS Estimation: Additivity
Introduction 9-1
Adjustment to Independent Statewide Estimates-The Handbook Share Method 9-2
Interstate Areas 9-3
10. LAUS Estimation: Disaggregation
Introduction 10-1
The Population-Claims Method of Disaggregation 10-3
Claims-Based Unemployment Disaggregation 10-7
11. Annual Processing
Introduction 11-1
Annual Model Review 11-1
Population Controls 11-2
Annual Re-Estimation 11-6
Annual Sub-State Area Estimates 10-11
Appendix , Guidelines for Reviewing Benchmark Compare Edit Results 11-11
12. STARS Macro Output
Introduction 12-1
STARS Review Estimates 12-2
STARS Table s 1 through 12 12-4
Figures 1 through 6 12-13
Appendix, Notes for STARS Tables 12-21
Glossary G-1

1

Local Area Unemployment
Statistics Program:
Introduction and Overview

Introduction
he Local Area Unemployment Statistics (LAUS) program is a
Federal/State cooperative program which produces monthly employment
and unemployment estimates for approximately 7,300 geographic areas.
The areas include all States, the District of Columbia, labor market areas (LMAs),
counties, cities with a population of 25,000 or more, and all cities and towns in
New England, regardless of population. These estimates are key indicators of
local economic conditions. They are used by State and local governments for
planning and budgetary purposes and as determinants of the need for local
employment and training services and programs. LAUS estimates are also used
to determine the eligibility of an area for preferential treatment or benefits under
various Federal assistance programs.

T

The underlying concepts and definitions of all labor force data developed by the
LAUS program are consistent with those of the Current Population Survey (CPS).
Each month, a tiered approach to estimation is used. Model-based estimates are
produced for the nine Census divisions that geographically exhaust the nation.
(These models use inputs only from the CPS.) The division estimates are
benchmarked to the national levels of employment and unemployment each
month. The benchmarked division model estimate is then used as the benchmark
for the States within the division. Monthly estimates for all States, the District of
Columbia, New York City, Los Angeles-Long Beach, and the respective balances
of New York and California, are produced us ing estimating equations based on
time series and regression techniques. These “signal plus noise” models combine
current and historical data from the CPS, the Current Employment Statistics
(CES) program, and State unemployment insurance (UI) systems. Models are
August 2010

LAUS Program Manual 1-1

also utilized for five additional substate areas and their respective State balances.
These models are like the division models in that they only utilize data from the
CPS. The areas are: the Chicago-Naperville-Joliet, IL metropolitan division; the
Cleveland-Elyria-Mentor, OH metropolitan area; the Detroit-Warren-Livonia, MI
metropolitan area; the Miami- Miami Beach-Kendall, FL metropolitan division;
and the Seattle-Bellevue- Everett, WA metropolitan division. Area and balance of
State models are controlled directly to their respective State totals.
Monthly estimates for Puerto Rico are produced from a survey modeled on the
CPS survey. This survey is conducted by Puerto Rico and the resultant estimates
are provided to BLS by the Puerto Rican Bureau of Employment Security.
Estimates for substate labor market areas (other than the substate areas noted
above) are produced using a standard methodology called the “Handbook”
method. This method also uses data from several sources, including the CPS,
CES, State UI systems, and the decennial census, to create estimates which are
then adjusted to the model-based measures of employment and unemployment for
the State or balance-of-State area, as appropriate. Below the labor market area
level, estimates are prepared for all counties and cities of 25,000 population, using
disaggregation techniques based on decennial and annual population estimates
and current UI statistics.

August 2010

LAUS Program Manual 1-2

History
Since the late 1940’s, sub-national estimates of employment and unemployment
have been developed by States under the guidance of the Federal government.
These estimates were initially developed in response to a need to quantify labor
dislocations during World War II. To meet this need, the War Manpower
Commission developed an estimation program to supply figures on local area
labor and material shortages. With the end of the war, the Labor Department’s
Bureau of Employment Security (BES) (now the Employment and Training
Administration) took over responsibility for manpower programs.
As the need for more detailed statistics increased, there was also a need for more
conformity in estimation in the individual States. In 1950, BES introduced
guidelines on estimation, entitled Techniques for Estimating Employment, and
distributed them to the States. A decade later, revised and updated techniques
were republished in the Handbook on Estimating Unemployment. This was a 70step method of estimating procedures for producing unemployment data for the
State and for labor market areas. The Handbook method used a series of
“building blocks”, including establishment employment and unemployment
insurance data, to produce unemployment rates equivalent to the Current
Population Survey (CPS) but without the high cost of a household survey. As
early as 1961, the local area unemployment statistics were used to distribute
federal funds to local areas under such programs as the Area Redevelopment Act.
In 1962, the President’s Committee to Appraise Employment and Unemployment
Statistics (the Gordon Committee) criticized the validity of the Handbook method.
This was followed by a series of independent studies comparing the Handbook
estimates to those from the Census or from the CPS. They reported the existence
of biases and inaccuracies in the Handbook procedures. In 1971, the General
Accounting Office, after a year-long audit of two States’ unemployment
estimating programs, came to the same conclusions, and also found that States
were independently introducing their own changes into the Handbook Method.
The GAO recommended that the States’ procedures be reviewed and monitored in
order to reestablish methodological conformity, in that any State change which
improved the accuracy and comparability of the statistics be integrated into the
methodology, and that “high priority” be given to a general improvement in the
estimating methods.
In the early 1970’s, BLS was publishing CPS-based labor force statistics for
selected States and large areas while BES was publishing Handbook-based
statistics for all States and areas. Shortly after the GAO report was issued, OMB,
as part of its review of statistical programs in the Department of Labor,
determined that general purpose statistics should be the responsibility of BLS. In
November 1972, the responsibility for local area unemployment statistics was
transferred to BLS. Therefore, beginning in 1973, BLS (with the cooperation of
all States) published monthly labor force data for all States and labor market
areas, based on the Handbook procedures. One year later, BLS introduced the
August 2010

LAUS Program Manual 1-3

first major revisions to the program. The revisions had a two- fold purpose: to
introduce more conformity between LAUS and CPS data, and to achieve a greater
level of consistency of procedures among the States.
The most important of the methodological changes introduced by BLS in
November 1973 was the direct use of CPS data. At that time, the CPS was a
nationally-based sample. In order to identify usable State CPS data, a reliability
criterion was established which required that State samples be sufficiently large to
estimate the unemployment level with a coefficient of variation (CV) of no more
than 10 percent at one standard error when the unemployment rate is 6 percent.
Applying this standard resulted in the identification of 19 States and 30
metropolitan areas for which CPS data could be used directly as the annual
average benchmark for 1970-73. During 1974, the Census Bureau revised the
procedure used to weight up State sample data to reflect the universe, which
resulted in a lower estimated variance. Thus, 8 more States were able to be
benchmarked to the CPS. In 1975, BLS contracted with the Census Bureau to
expand the sample by 9,000 households in the 23 remaining States and the
District of Columbia, so that all States were able to be benchmarked to annual
average CPS estimates in 1976.
In 1978, BLS broadened the applicability of the reliability criterion for use of CPS
data by also considering monthly data, within the context of a budget proposal to
expand the CPS to yield monthly employment and unemployment data for all
States by June 1981. Under the expanded criterion, which specified a 10-percent
CV on monthly data, monthly CPS levels were used directly for 10 States, 2
areas, and the respective balance-of-State areas. The use of annual average CPS
data for the other 28 metropolitan areas was discontinued at that time, so that all
substate areas not meeting the monthly reliability criterion would be treated the
same. Ultimately, the budget proposal which initiated the direct use of monthly
State CPS data was rejected as too costly.
In addition to the 1975-76 increase to the CPS to obtain reliable annual average
data for all States, in 1980, 9,000 households were added to improve the
reliability in the 40 nondirect-use States. A final sample increase of 6,000 was
implemented in 1981 to improve the reliability of data in 30 specific metropolitan
areas, 10 of their central cities, and the respective balance-of-State areas. In 1982,
however, because of the Federal budget cut, the 1981 supplement and one- half of
the 1980 supplement were eliminated.
Another part of the improvement commitment supported by the budget
supplements was a $2.5 million effort to standardize and improve the
unemployment insurance data which provide the only current unemployment
measure for all substate areas. Funding for this initiative was provided to
BLS in 1975-76, and used through contracts with States to correct and
augment unemployment insurance statistics to make them more appropriate
for use in LAUS estimation. Inconsistencies within and among States were
eliminated, quality control measures were instituted, and manual tallies were
August 2010

LAUS Program Manual 1-4

replaced with computer-generated tabulations. Through such improvements
as the use of place-of-residence of the claimant, the CPS reference week, and
the elimination of claimants who had earnings due to employment, closer
adherence to CPS concepts was achieved. The resultant improved
unemployment insurance data were implemented in LAUS estimation in
1978.
In July 1985, the CPS redesign based on the incorporation of 1980 Census data
was fully implemented. A key part of the redesign involved a change in the
sample structure of the CPS from a national-based one to a State-based stratified
sample. Based on the redesign and sample restructuring, the reliability of the CPS
data at the State level was improved such that the monthly and annual CVs for
direct use of CPS data were reduced to 8 percent.
In 1986, updated inputs to the Handbook based on 1980 Census data and a
number of important methodological improvements in component groups of the
Handbook employed and unemployed not covered by unemployment insurance
were implemented.
Also in 1986, efforts to utilize econometric techniques to estimate monthly State
employment and unemployment were strengthened. The earliest BLS attempts to
explore regression methods go back to the late 1970’s. In addition to internal
work, BLS contracted with Mathematica Policy Research, Inc., which conducted
extensive State and area research using time series and cross-sectional models.
Their final report was delivered in 1981.
Internal model research continued through the 1980’s, resulting in the
identification of variable coefficient models as a possible substitute for the
nondirect- use State method used at that time—Handbook estimation adjusted to a
six-month moving average CPS. In 1986, the State Research Group was
established with participation of selected State research directors to facilitate the
evaluation of model-based estimates and to ensure adequate communication of
State needs. In 1987, a subsequent group of State research directors was
established—the Regression Implementation Committee—to further evaluate the
model approach during the one-year period of dual estimation in 1988. In
addition to internal and State review of the model-based estimating method,
Professor Art Dempster of Harvard University participated in the evaluation
effort. The result of these efforts was the implementation of variable coefficient
models in the nondirect-use States in January 1989.
Following the incorporation of the first generation of State econometric models,
model research continued. Since seasonally-adjusted estimates were available for
the direct-use States, efforts were focused on the seasonal adjustment of the
model-based State estimates. A BLS workgroup was established in 1989 to
evaluate the appropriateness of seasonally adjusting the model estimates using the
X-11 ARIMA software used for the CPS. The workgroup’s positive report in the
fall of 1991 led to the introduction of monthly seasonally-adjusted nondirect-use
August 2010

LAUS Program Manual 1-5

State estimates in 1992.
Also in the early 1990’s, a major effort was undertaken to improve the model
specifications. Research was conducted to explicitly account for important
characteristics of the CPS sample design. This led to better control of the effects
of sampling error on the model estimates. In addition, a more flexible modeling
of State-specific seasonal and trend effects was identified. The resultant secondgeneration models were referred to as “signal-plus-noise” models. These
modeling results were provided to States for comment in early 1993, and were
implemented in 1994.
In January 1996, the Bureau reduced the number of households in the Current
Population Survey, to accommodate lower funding levels for the labor force
program. One result was that the sample was no longer sufficient to provide
monthly data directly for the 11 large States, New Yo rk City, and the Los Angeles
Metropolitan Area. In response, monthly estimation for these States and areas
was replaced by the time series modeling methodology used for the other 39
States and the District of Columbia. Also in January 1996, the LAUS substate
estimation process was streamlined and input options were eliminated to
accommodate the reduction of resources for the LAUS program.
In January 2005, a major program redesign was implemented. Work on the
Redesign began in Fiscal Year 2001, with a budget initiative to enhance the
quality and quantity of LAUS program statistics. Major LAUS Redesign
components included improvements to the method of State and large area
estimation, including real-time benchmarking, extending the model-based
estimation methodology to additional substate areas, improving the methods used
in all other areas through better techniques and input data, and updating the
geography with 2000 Census-based areas.
The 2005 LAUS Redesign introduced a new generation of LAUS models. The
objectives of the new generation models were to implement direct model-based
seasonal adjustment with reliability measures and to improve the benchmarking
procedure by incorporating real-time monthly benchmarking. At the same time, 6
area models were introduced along with corresponding Balance of State models.
Real-time benchmarking addressed a number of concerns with the prior
generation of LAUS models. It reduced annual revisions by incorporating the
CPS benchmark on a current basis. It eliminated prior model biases and
benchmarking issues. It ensured that national events and shocks to the economy
were reflected in State estimates as they occurred. It also eliminated the
discrepancy between the sum-of-States estimates and the national not-seasonallyadjusted totals. Measures of error on the seasonally- adjusted and not-seasonallyadjusted estimates and the over-the- month change were introduced for Division,
State, and area model estimates.

August 2010

LAUS Program Manual 1-6

The LAUS Redesign also included projects to improve methodology, update
geography and decennial census inputs, and improve the quality of inputs to the
estimates. The Redesign resulted in significant improvements to the accuracy of
the LAUS labor force estimates and has enhanced the ability to analyze labor
market behavior. Methodological changes included improvements to substate
unemployment estimation that addressed long-standing inadequacies with the
previous method and an innovative approach to adjusting place-of-work
employment to place-of-residence that more accurately reflects complexities of
commuting. In addition, all LAUS areas were revised for the latest OMB and
BLS geographic definitions.
In January 2010, BLS introduced additional improvements to the LAUS models
with the implementation of smoothed-seasonally- adjusted (SSA) estimates. The
SSA estimates incorporate a long-run trend smoothing procedure, resulting in
estimates that are less volatile than those previo usly produced by the LAUS
estimation methodology. The use of the SSA methodology is effective in
reducing the number of spurious turning points in current estimates. More
importantly, SSA estimation can reduce revisions in historical estimates and
remove the potential disconnection between historically benchmarked and current
estimates.

August 2010

LAUS Program Manual 1-7

LAUS Time Line
Year

Historical Developments Related to LAUS

1933

Wagner-Peyser Act created Employment Service for
registering the unemployed

1935

Social Security Act created Unemployment Insurance
System

1937

Works Projects Administration began collecting
household-based labor force data

19391945

War Manpower Commission developed program on
local area labor and material shortages

1943

Responsibility for conducting household survey
transferred to Bureau of the Census

1948

Monthly Report on the Labor Force renamed Current
Population Survey (CPS)

1950

BES (now ETA) published the manual “Techniques for
Estimating Unemployment”

1959

Responsibility for analyzing and publishing CPS data
given to BLS; Census continues to conduct survey

1960

Manual for estimating area unemployment revised by
BES, title changed to “Handbook on Estimating
Unemployment” (70-Step Method)

1961

Area Redevelopment Act passed

1962

President’s Committee to Appraise Employment &
Unemployment Statistics (Gordon Committee) issued
final report

1965

Public Works and Economic Development Assistance
Act (PWEDA) passed

1972

Responsibility for LAUS program transferred to BLS

1973

Comprehensive Employment and Training Act (CETA)
passed

1975

CPS sample expansion; CPS benchmarking extended to
27 States

1975

First round of UI Database Survey conducted by BLS

1976

CPS benchmarking extended to all States.

1976

Public Works Employment Act (PWEA) passed

August 2010

LAUS Program Manual 1-8

LAUS Time Line (Continued)
Year

Historical Developments Related to LAUS

1976

National Commission on Employment and
Unemployment Statistics (Levitan) and National
Commission on Unemployment Compensation established

1978

Direct use of monthly CPS estimates for limited
number of States and areas introduced

1978

First UI database improvements incorporated into the
Handbook estimates

1979

Levitan Commission issued recommendations

1982

Job Training Partnership Act (JTPA) replaced CETA

1983

Second round of UI Database Survey conducted by
BLS

1985

Updated State-based CPS sample based on 1980
Census introduced

1986

Major revisions to Handbook methodology
incorporated

1989

Variable coefficient model estimates incorporated for
nondirect- use States

1992
1994
1996

Seasonal adjustment of model-based estimates
introduced.
Second generation of LAUS models introduced; 1990
Census data incorporated into LAUS; new CPS
questionnaire and data collection method implemented
Direct-use States adopt model based estimation
method; Handbook method streamlined to 13 steps

2005

Third generation of LAUS models introduced, bringing
real-time benchmarking and model-based seasonal
adjustment to the methodology; improvements to
Handbook methodology include dynamic ratio
adjustment for place-of-work employment and updated
new and reentrant unemployed estimation; 2000
Census data incorporated into LAUS; and redesign of
the State estimating system.

2010

Smoothed-seasonally-adjusted methodology
implemented to develop official seasonally-adjusted
estimates.

August 2010

LAUS Program Manual 1-9

Data Sources
LAUS estimates are designed to reflect the labor force concepts embodied in the
Current Population Survey (CPS) through the direct use of CPS data in the
estimation and the incorporation of official labor force specifications in other
program inputs. This allows LAUS estimates for a State to be conceptually
comparable to national labor force measures and to LAUS estimates in other
States.
LAUS estimates are based on data from a number of different sources. Primary
source data for the creation of employment and unemp loyment estimates include
the CPS; the State Unemployment Insurance (UI) systems; the Current
Employment Statistics (CES) program; the Quarterly Census of Employment and
Wages (QCEW); and the Decennial Census. Each of these inputs to LAUS
estimation is described in detail in the following four chapters. A brief summary
of each data source is provided below.

The Current Population Survey
The CPS is a monthly sample survey of households, conducted by the Bureau of
the Census under contract to the Bureau of Labor Statistics. It provides statistics
on the labor force status of the civilian noninstitutional population 16 years of age
and over. CPS data are collected each month from a probability sample of
approximately 60,000 occupied households and yield estimates of demographic,
social, and economic characteristics of the population.
The Bureau of Labor Statistics has responsibility for analyzing and publishing
monthly employment and unemployment estimates for the Nation. CPS data are
valuable inputs into LAUS monthly estimation due to their regular availability,
comparability across States, and measurable statistical error.

Unemployment Insurance Systems
Under the Unemployment Insurance system, an employer must pay a tax for each
employee covered by the State law. Coverage includes the State UI program and
the Federal Civilian Employment program. This tax is, in effect, an insurance
premium paid to provide for possible unemployment benefits. When any
employee in a covered job becomes unemployed, he/she may file an Initial Claim
for unemployment insurance benefits. For the first initial claim filed in a year, a
monetary determination as to whether the individual is covered by the State
unemployment insurance system and, if so, how much in benefits is due, will then
be made by the State. A nonmonetary determination follows, looking into the
nature of the individual’s job loss. A qualifying claimant will receive weekly
compensation until the maximum benefit amount is exhausted or until the person
returns to work, whichever is earlier.
August 2010

LAUS Program Manual 1-10

The UI administrative statistics created in this process are useful for LAUS
estimation because they are current and area-specific, allowing for their use in
estimation for a great many geographic areas. (See Chapter 3 for more details
on the UI system.)

Current Employment Statistics and QCEW
Both the CES and QCEW programs are Federal/State cooperative programs
which obtain employment data from employers.
The CES is a voluntary sample survey of establishments covered by State and
Federal UI laws. It is designed to produce monthly estimates of employment,
hours, and earnings for the Nation, all States, and most major metropolitan areas.
The QCEW data series is a universe of monthly employment and quarterly wage
information by industries covered for unemployment insurance, State, and county.
Completion of a quarterly contribution report, which is basis for the QCEW, is
mandatory in industries covered by Federal and State UI laws.
Data obtained through these two programs are used in LAUS employment
estimation. (See Chapter 4 for a detailed discussion of the CES and QCEW
programs.)

Decennial Census
The Decennial Census is a universe count of the national population conducted
each decade by the Bureau of the Census. The primary purpose of the Decennial
Census is to apportion seats to the U.S. House of Representatives and for
determining legislative district boundaries. Through the 2000 Census, the census
also was a source of socioeconomic and demographic data in great geographic
detail.
The LAUS program methodology uses decennial census data, in part, for
adjusting establishment-based employment estimates to residency-based
employment estimates, for estimating certain employment and unemployment
components in the Handbook methodology, and disaggregating or apportioning
labor market area estimates to smaller areas. (See Chapter 5 for additional details
on the decennial census.)

August 2010

LAUS Program Manual 1-11

Summary of Estimation Methods
Monthly estimates of employment and unemployment are prepared for
approximately 7, 300 geographic areas, which include all States, labor market
areas, counties, cities with a population of 25,000 or more, and all cities and
towns in New England, regardless of population. At each level of geographic
detail, the estimation method used depends on the most current data sources
available.
Statewide Estimates
A tiered approach to estimation is used for statewide estimates. First, modelbased estimates are developed for the nine Census Divisions that geographically
exhaust the nation using univariate signal-plus- noise models. The Division
models are similar to the State models, but do not use unemployment insurance
claims or payroll employment as input variables. The Division estimates are
benchmarked to the national levels of employment and unemployment on a
monthly basis. The benchmarked Division model estimate is then used as the
benchmark for the States within the Division.
Monthly labor force estimates for all States, the District of Columbia, the Los
Angeles- Long Beach metropolitan area, New York City, and the respective
balances of California and New York are based on dynamic time series regression
models that utilize data from the CPS, UI systems, and the CES survey. Both
smoothed-seasonally-adjusted and not-seasonally-adjusted estimates are produced
each month.
The model methodology is also utilized for five additional substate areas and their
respective balances of States. These models are univariate, like the Division
models, in that they do not use UI or CES inputs. The areas are: the ChicagoNaperville-Joliet, IL metropolitan division; the Cleveland-Elyria-Mentor, OH
metropolitan area; the Detroit-Warren-Livonia, MI metropolitan area; the MiamiMiami Beach-Kendall, FL metropolitan division; and the Seattle-BellevueEverett, WA metropolitan division. The substate area and the balance-of-State
estimates are benchmarked to the statewide control totals of not-seasonallyadjusted employment and unemployment estimates. (See Chapter 6.)
Labor Market Estimates
States are divided into Labor Market Areas (LMAs) which exhaust the geographic
area of the State. LMAs are economically integrated geographic areas within
which individuals can reside and find employment within a reasonable distance or
can readily change employment without changing their place of residence. Other
than the areas noted above for which model-based estimation is used, independent
estimates are produced for all LMAs using a standard procedure known as the
“Handbook” method. The Handbook method yields employment and
unemployment estimates for an area comparable to what would be produced by a
August 2010

LAUS Program Manual 1-12

random sample of households in the area, but without the expense of a CPS-like
labor force survey. Handbook estimates are adjusted to add to the LAUS
Statewide or balance-of-State employment and unemployme nt estimates to create
the official LMA LAUS estimates. LAUS estimates for sub-LMA areas, such as
individual counties within multi-county LMAs and cities with populations over
25,000, are derived by a disaggregation technique using population estimates and
UI statistics, or data from the decennial census. (See Chapters 7, 8, and 9 for
further details.) At the end of the year, State and substate areas are revised and
benchmarked to reflect updated, revised input data and model estimation. (See
Chapter 10.)

August 2010

LAUS Program Manual 1-13

LAUS Estimation Techniques

Area

Estimation Method

Nine Census Divisions

Signal-plus-noise univariate
regression model

50 States

Signal-plus-noise bivariate
regression model

District of Columbia

Signal-plus-noise bivariate
regression model

New York City, Balance of NY
State

Signal-plus-noise bivariate
regression model

Los Angeles, Balance of
California

Signal-plus-noise bivariate
regression model

Chicago, Cleveland, Detroit,
Miami, Seattle and balances of
Illinois, Ohio, Michigan,
Florida, and Washington

Signal-plus-noise univariate
regression model

Remaining Labor Market
Areas (LMAs)

Handbook, Additivity

Sub-LMA Areas

Disaggregation

August 2010

LAUS Program Manual 1-14

Publication and Administrative Uses of LAUS Estimates
The Bureau of Labor Statistics was given responsibility to develop and publish
the most current national, State, and local labor force and unemployment data by
the Office of Management and Budget in Statistical Policy Directive No. 11,
“Standard Data Source for Statistical Estimates of Labor Force and
Unemployment.” This directive also requires the use of BLS-developed LAUS
estimates by federal executive departments, agencies, and establishments in
allocations of federal resources and eligibility determinations. The complete text
of this Directive is provided at the end of this Chapter.

Publication of LAUS Estimates
Each month, labor force, employment, unemployment and
unemployment rate estimates for all 7,300+ LAUS areas are
published by BLS. Data from the LAUS program are made
available to users in a variety of ways.
•

The monthly “Regional and State Employment and
Unemployment” news release is issued approximately two weeks after the
national release of labor force data. It presents data for the Census regions
and Divisions and all modeled LAUS estimates. The Bureau’s public
database, LABSTAT (http://www.bls.gov/lau/), is also updated with these
estimates at that time.

•

The monthly “Metropolitan Area Employment and Unemployment” news
release is issued about 12 days after the Region and State release and
contains labor force and unemployment estimates for all metropolitan
areas in the nation. These estimates are also issued in LABSTAT at that
time.

•

Estimates of labor force, employment and unemployment for micropolitan
areas, small labor market areas, counties, cities with a population of
25,000 or more, and all cities and towns in New England are issued in
LABSTAT at the same time the Metropolitan area data are released.

•

Annual average emp loyment status data are provided each year in a press
release entitled “State and Regional Unemployment, Annual Averages”,
which is typically issued at the end of February. It presents data on the
population, civilian labor force, employed, unemployed, and
unemployment rate for regions, Divisions, and States.

•

The annual publication, Geographic Profile of Employment and
Unemployment, provides annual average CPS data for census regions and
Divisions, the 50 States and the District of Columbia, 50 large
metropolitan areas, and 17 central cities. Data are provided on the
employed and unemployed by selected demographic and economic
characteristics.

August 2010

LAUS Program Manual 1-15

Legislative Uses of LAUS Estimates
Each year, LAUS estimates are used to distribute federal funds to States and areas
or make eligibility determinations by a number of federal programs. The
following table, “Administrative Uses of Local Area Unemployment Statistics”,
presents information on the federal programs that utilize LAUS data in allocating
funds to States. Allocation formulas, reference periods, and geographic coverage
information are presented. Total funding for these programs amounted to
$60,849.3 million in Fiscal Year 2009. These programs are described in greater
detail in the section below. The American Recovery and Reinvestment Act of
2009 used LAUS data to allocate an additional $144,350.0 million to States.
These programs are also described in greater detail below and on the Bureau’s
website at http://www.bls.gov/lau/lauadminuses.pdf.

August 2010

LAUS Program Manual 1-16

ADMINISTRATIVE USES OF LOCAL AREA UNEMPLOYMENT STATISTICS
2009 Funding
Geographic Areas Used
(Millions)
Department of Labor – Employment and Training Administration
Adult Employment and Training
$ 861.5
States and Areas of Substantial
Activities (WIA, Title I, Chapter 5)
Unemployment (ASUs). An ASU is
a contiguous piece of geography
consisting of counties, cities, and/or
parts of each, with a population of at
least 10,000 and an unemployment
rate of at least 6.5 percent. (1) (2)
Youth Activities (WIA, Title I, Chapter
$ 924.1
States and ASUs. (1) (2) (3)
4)
User Agency/Program

Dislocated Worker Employment &
Training Activities (WIA, Title I,
Chapter 5)

Employment Service Grants to States
(Wagner-Peyser Act, Section 5)

$ 1,466.9

$ 703.6

States. (1) (2)

States. (1)

Reference Period

Allocation Formulas/Qualifying Criteria

Most recent program
year (July -June).

Funding based on the following proportions: 1/3 on relative number of
unemployed in ASUs, 1/3 on relative number of excess unemployed (i.e.,
number of unemployed in excess of 4.5 percent of labor force), and 1/3 on
relative number of economically disadvantaged adults, age 22-72. Not more
than 0.25% of funds allocated to outlying areas. (Additional
minimum/maximum provisions apply.)

Most recent program
year (July -June).

Funding based on the following proportions: 1/3 on relative number of
unemployed in ASUs, 1/3 on relative number of excess unemployed, and 1/3
on relative number of economically disadvantaged youth, age 16-21. Not
more than 0.25% of funds allocated to outlying areas. Up to 1.5% allocated
to Native American programs. (Additional minimum/maximum provisions
apply.)
Funding based on the following proportions: 1/3 on relative number of
unemployed, 1/3 on relative number of excess unemployed, and 1/3 on
relative number of individuals unemployed for 15 weeks or more. Not more
than 0.25% of funds allocated to outlying areas.

Most recent program
year (July -June) for
unemployed and excess
unemployed; most
recent calendar year for
unemployed 15+
weeks.
Most recent calendar
year.

State funding algorithm is based on the following proportions: 2/3 on
relative number of civilian labor force and 1/3 on relative number of
unemployed.
An area qualifies as a LSA when its average unemployment rate is 20
percent or more above the national rate (including Puerto Rico) for the
period, with the threshold being no lower than 6 percent and no higher than
10 percent.
State is eligible to pay EB if: (1) the seasonally adjusted total
unemployment rate (TUR) for the most recent 3-month period is at least 6.5
percent and at least 10 percent above the State TUR for the same 3-month
period in either of the 2 preceding years, or (2) the insured unemployment
rate (IUR) is at least 5 percent and at least 120 percent of the average IUR
for the same 13-week period in either of the 2 preceding years.

Labor Surplus Areas

(4)

Counties, cities over 25,000
population, and county balances. (1)

Most recent 2-calendar
year average.

Federal-State Extended Unemployment
Benefits (EB)

(5)

States. (1)

Most recent 3 months
for total unemployment
trigger (TUR) or most
recent 13 weeks for
insured unemployment
trigger (IUR).

Census tracts and non-metropolitan
counties.

Not specified.

An area can qualify if it is an underserved area, which is defined as an area
comprised of census tracts with the following distress criteria: (i) a census
tract where the unemployment remains high (50 percent or more above the
nation’s unemployment rate) and (ii) a census tract where a high rate of
poverty persists.

Most recent 3-calendar
year average.

Funding is based on an estimate of the number of veterans seeking
employment in a State as a portion of the number of veterans seeking
employment nationwide.

Youthbuild Program

$47.0

Department of Labor – Veterans’ Employment and Training Service
Jobs for Veterans Act of 2002
$ 168.9
States. (1)

August 2010

LAUS Program Manual 1-17

ADMINISTRATIVE USES OF LOCAL AREA UNEMPLOYMENT STATISTICS
User Agency/Program
Department of Agriculture
The Emergency Food Assistance
Program (TEFAP)
Department of Agriculture (cont.)
Welfare Reform Act—Waivers to
Supplemental Nutrition Assistance
Program (SNAP) Time Limits for AbleBodied Adults Without Dependents
(ABAWD)

2009 Funding
(Millions)
$247.9

$38,601.0
(6)

Geographic Areas Used

Ten-month average of
most recent OctoberJuly period.

Farm commodities and funds are allocated based on the following
proportions: 3/5 on relative number of persons in households below the
poverty line and 2/5 on relative number of unemployed persons.

States, metropolitan areas (MAs),
counties, cities, Indian reservations,
and specially designated areas (e.g.,
census tracts). (1)

Generally 12-month
periods, but no less
than 3 months for
unemployment rate.
Not specified for
insufficient jobs
criterion.

Waivers are granted to areas with: (1) an unemployment rate over 10
percent for the latest 12-month (or 3-month) period or (2) insufficient jobs.

Most recent 24-month
average.

An area qualifies if: (1) the unemployment rate is at least one percentage
point above the national rate, (2) the per capita income is 80 percent or less
of the national average per capita income, or (3) there is a special need, as
determined by EDA, arising from actual or threatened severe unemployment
or economic adjustment problems resulting from severe short -term or longterm changes in economic conditions.
Same qualifying criteria used in the Public Works Program.

Same geographic areas used in the
Public Works Program.

Most recent 24-month
average.

Department of Defense – Defense Logistics Agency
Procurement Technical Assistance (PTA)
$12.0

States, counties, and cities. (1)

Most recent 24-month
average.

An area qualifies for assistance if: (1) the unemployment rate is at least one
percentage point above the national average for the most recent 24-month
period or (2) the per capita income is 80 percent or less of the State average.

Department of Health and Human Services
Temporary Assistance to Needy Families
(TANF)—Contingency Fund Drawdown

$272.9
(7)

States, District of Columbia, and
Puerto Rico.

Most recent 3-month
average.

$17,059.0
(8)

States, District of Columbia, and
Puerto Rico. (3)

Not available.

States and the District of Columbia can access funds if they are determined
to be "needy," based on a seasonally adjusted unemployment rate that is at
least 6.5 percent for the 3-month period and at least 110 percent of the rate
for the corresponding period in either of the 2 preceding calendar years; or if
the number of food stamp recipients increases at least 10 percent during the
3-month period. (TANF automatically gives block grants—with an upper
limit of $71 million—to Puerto Rico.)
In transitioning from welfare to work, individuals are granted up to 6 weeks
for which a job search or participation in a workfare program will be counted
as work. This time limit is extended to 12 weeks if the State unemployment
rate is at least 50 percent above the national rate. (TANF automatically
gives block grants—with an upper limit of $71 million—to Puerto Rico.)

TANF—Exemption from Benefit
Limitation

$35.3

Allocation Formulas/Qualifying Criteria

States. (1) (2) (3)

Department of Commerce – Economic Development Administration
Public Works Program
$129.3
Areas defined by geographic/political
boundaries, e.g., States, cities,
counties, Indian reservations. (1) (2)
(3)

Economic Adjustment (Title 9)

Reference Period

Department of Homeland Security – Federal Emergency Management Agency
Emergency Food and Shelter Program
$200.0
Counties and cities. (1) (2)

August 2010

Most recent 12-month
average.

Jurisdictions qualify for FEMA funding if they meet one of the following
criteria: (1) 13,000 or more unemployed with a jobless rate of 4.3 percent or
more, (2) 300-12,999 unemployed with a jobless rate of at least 6.3 percent,
or (3) 300 or more unemployed with a poverty rate of at least 11 percent.

LAUS Program Manual 1-18

ADMINISTRATIVE USES OF LOCAL AREA UNEMPLOYMENT STATISTICS
2009 Funding
Geographic Areas Used
(Millions)
Department of Homeland Security – U.S. Citizenship and Immigration Services
Immigration Act of 1990
(9)
MAs and counties, cities, and subEmployment Creation Visas
areas within MAs.
User Agency/Program

Department of the Treasury
Riegle Community Development and
Regulatory Improvement Act of 1994—
Bank Enterprise Awards

Riegle Community Development and
Regulatory Improvement Act of 1994—
Small and Emerging CDFI Assistance
Component (Technical Assistance)
North American Development Bank
(NADBank) Community Adjustment and
Investment Program (CAIP)
Appalachian Regional Commission
Area Development Program
Distressed Counties Grants

Small Business Administration
Historically Underutilized Business
Zones (HUBZones)

Total Appropriations

Reference Period

Allocation Formulas/Qualifying Criteria

Most recent calendar
year or 12-month
average.

Visas are granted for lower investment amounts in rural areas or areas with
an unemployment rate at least 50 percent above the national average.

An institution may qualify if part or all of its service area: (1) is located
within one unit of general local government, (2) has a contiguous boundary,
(3) (a) has a population of 4,000 or more, if in a metropolitan area, or (b) has
a population of 1,000 or more, if outside of a metropolitan area, or (c) is
entirely within an Indian reservation, (4) has a poverty rate of at least 30
percent, and (5) has an unemployment rate at least 1.5 times the national
rate.
Same qualifying criteria used for the Bank Enterprise Award.

$20.0

MAs, counties, cities, and possible
sub-areas (e.g., census tracts). (1) (2)
(3)

Most recent 12-month
period before
announcement of
application period.

$2.4

Same geographic areas used for the
Bank Enterprise Awards.

Same reference period
used for the Bank
Enterprise Awards.

$22.5
(10)

Communities (discrete geographical
areas) i.e., counties, towns, or cities.

Most recent 12-month
average.

Eligibility for CAIP financing is based on: (1) significant job loss connected
to the passage of NAFTA and (2) an unemployment rate that is at least one
percentage point above the national rate over the same time period.

$75.0
(11)

410 counties in 13 states (all of WV
and parts of AL, GA, KY, MD, MS,
NY, NC, OH, PA, SC, TN, VA);
currently 81 counties qualify.

Most recent 3-year
average.

An area qualifies as distressed if it ranks in the worst 10 percent of the nation
according to ARC’s index-based County Economic status Classification
System, which compares each county with national averages of: (1) the
unemployment rate, (2) per capita market income, and (3) the poverty rate.

(4)

Census tracts, non-metropolitan
counties, or Indian reservations. (1)

Most recent annual
average for
unemployment rate.

An area qualifies if it is: (1) a "qualified" census tract (as defined in the
1986 IRS code), (2) a non-metropolitan county with (a) median household
income less than 80% of the statewide non-metropolitan median or (b) an
unemployment rate at least 140% of the statewide average or the national
average, or (3) within the boundaries of an Indian reservation, or (4) a
military base closed under the Defense Base Realignment and Closure Act of
1990.

$60,849.3

NOTE: The term “cities” also includes townships and boroughs in selected states for various programs.
(1) The District of Columbia and Puerto Rico are treated as states.
(2) Outlying areas include the U.S. Virgin Islands, Guam, American Samoa, Northern Marianas Islands, Marshall Islands, Micronesia, and Palau.
(3) Native American Program includes American Indians, Native Hawaiians, and Alaska Natives.
(4) Program does not allocate funds, but gives preference to firms in bidding on federal procurement.
(5) Under regular state extended benefits, monies are not appropriated, but are drawn from the Unemployment Insurance Trust Fund. If the 3 -month average TUR is at least 8%, and at least 10% above the TUR for the same 3 - month period in either of the 2
preceding years, the State enters a "high unemployment period" during which 20 weeks of EB are payable.
(6) Dollar amount is full cost of Food Stamp Program. Soup Kitchen and Food Bank funding was merged into the Welfare Reform Act of 1996, and, though the program may continue to receive donations, there is no separate funding.
(7) Under the Welfare Reform Act, a Contingency Fund of State Welfare Programs was established, with a $2 billion limit.
(8) Dollar amount is the full cost of the TANF program.
(9) Under the Act, at least 3,000 visas are distributed to eligible immigrant entrepreneurs who establish a new commercial enterprise in a targeted employment area (rural area or area with high unemployment).
(10) Dollar amount is the total amount of capital available to finance adjustment assistance.
(11) At least half of this program funding is allocated to counties classified as distressed.
June 26, 2009

August 2010

LAUS Program Manual 1-19

Department of Labor:
Employment and Training Administration,
Workforce Investment Act, Title 1, Chapter 5: Adult
Employment and Training Activities
Program Objectives: To provide job training and related assistance to
economically disadvantaged individuals and others who face significant
employment barriers. The ultimate goal of the Act is to move trainees into
permanent, self-sustaining employment. This legislation authorizes
training and services for the economically disadvantaged and others who
face significant employment barriers. Training is afforded through grants
to States for local training and employment programs. States are
responsible for further allocating funds to their Service Delivery Areas
(SDAs) and for overseeing the planning and operation of local programs.
Program services include an assessment of an unemployed individual’s
needs and abilities and a strategy of services such as classroom training,
on-the-job training, job-search assistance, work experience, counseling,
basic skills training, and support services.

Employment and Training Administration,
Workforce Investment Act, Title 1, Chapter 5: Youth
Activities
Program Objectives: To serve eligible low- income youth, ages 14-21, who
face barriers to employment. Funds for youth services are allocated to
State and local areas based on a formula distribution. Service strategies,
developed by workforce providers, prepare youth for employment and/or
post-secondary education through strong linkages between academic and
occupational learning. Local communities provide youth activities and
services in partnership with the WIA One-Stop Career Center System and
under the direction of local Workforce Investment Boards.

Employment and Training Administration, Workforce
Investment Act, Title 1, Chapter 5, Dislocated Workers
Program Objectives: To assist dislocated workers to obtain unsubsidized
employment through training and related employment services using a
decentralized system of State and local programs. The Workforce
Investment Act (WIA) provides funds to States and local substate
grantees. The Act authorizes employment and training help for dislocated
workers. Workers who lose their jobs in mass layoffs or plant closings and
others who have been laid off and are unlikely to return to their jobs can
take advantage of early intervention programs, occupational skill training,
August 2010

LAUS Program Manual 1-20

job search assistance, support services, and relocation assistance.

Employment and Training Administration,
Employment Service Grants to States
Program Objectives: The employment service is available to all those
legally authorized to work in the United States in order to assist millions
of job seekers and employers and, in some areas, provide job training and
related services. The Federal Government, through the Employment and
Training Administration, provides general direction, funding, and
oversight, and also assists the States with programs of test development,
occupational analysis, and maintenance of an occupational classification
system. The State employment security agencies operate 1,800 local
Employment Service offices. In accordance with their needs, States may
provide specialized assistance to such groups as youth ages 16-22, women,
older workers, persons with disabilities, rural residents and workers, and
the economically disadvantaged.
Public employment service assistance, including employability
assessment and referral to training if necessary, is free to job seekers.
Most of the service’s appropriations come from the trust funds collected
under the Federal Unemployment Tax Act (FUTA), with a small portion
coming from general revenues.

Employment and Training Administration, Labor
Surplus Areas
Program Objectives: The purpose in classifying labor surplus areas is to
put the Federal Government’s procurement contracts into areas of high
unemployment. Employers located in these labor surplus areas are
eligible for preference in bidding on Federal procurement contracts to
direct government funds into areas where people are in the most severe
economic need.

Employment and Training Administration, FederalState Extended Unemployment Compensation
Program and Special Temporary Programs
Program Objectives: Unemployment compensation is designed to provide
benefits to most workers out of work due to no fault of their own for
periods between jobs. Most States pay a maximum of 26 weeks of regular
benefits, except for two States-- Massachusetts and Washington --which
pay up to 30 weeks of benefits. In periods of very high unemployment in
individual States, benefits are payable for as many as 13 additional weeks,
up to a maximum of 39 weeks. In 14 States, an additional 7 weeks of
benefits are available to unemployed workers, depending on the
unemployment rate. Reflecting the high unemployment during the 2008August 2010

LAUS Program Manual 1-21

2010 recession, the Unemployment Compensation Act Extension Act of
2008 was enacted to provide up to 53 additional weeks of benefits to
unemployed workers.

Employment and Training Administration, Workforce
Investment Act, Title I, Subtitle D, Section 173A:
Youthbuild Program
Program Objectives: To provide disadvantaged youth with: the education
and employment skills necessary to achieve economic self sufficiency in
occupations in high demand and postsecondary educatio n and training
opportunities; opportunities for meaningful work and service to their
communities; and opportunities to develop employment and leadership
skills and a commitment to community development among youth in lowincome communities. As part of their programming, YouthBuild grantees
will tap the energies and talents of disadvantaged youth to increase the
supply of permanent affordable housing for homeless individuals and lowincome families and to assist youth develop the leadership, learning, and
high-demand occupational skills needed to succeed in today's global
economy.

Veterans’ Employment and Training Service, Jobs for
Veterans Act of 2002
Program Objectives: To establish priority of service for veterans in
Department of Labor job training programs. The Act calls for priority of
service to be implemented by all “qualified job training programs,”
defined as “any workforce preparation, development or delivery program
or service that is directly funded, in whole or in part, by the Department of
Labor.

August 2010

LAUS Program Manual 1-22

Department of Agriculture:

Temporary Emergency Food
Assistance Program
Program Objectives: To make funds available to States for storage and
distribution costs incurred by nonprofit eligible recipient agencies in
providing nutrition assistance in emergency situations and to aid needy
people. TEFAP was created to reduce excess USDA inventories of
surplus commodities in storage, especially dairy products such as cheese,
and to supplement the diets of low-income households at a time of high
unemployment. Each State designates one agency to administer TEFAP.
Once USDA commodities are made available to the States, State officials
are responsible for determining the eligibility of organizations to receive
the commodities and for entering into agreements regarding allocation and
distribution. In addition, States are responsible for determining the types
and amounts of each commodity to be made available to organizations
within the State.
The Personal Responsibility and Work Opportunity Reconciliation Act of
1996 provided for the absorption of the Soup Kitchens/Food Banks
Program into TEFAP and requires the Secretary to use $100 million yearly
from the Food Stamp account to purchase commodities for TEFAP during
Fiscal Years 1997 through 2002.

Welfare Reform Act—Waivers to Supplemental
Nutrition Assistance Program
Program Objectives: The Personal Responsibility and Work Opportunity
Reconciliation Act of 1996 limits receipt of Food Stamp benefits to 3
months in a 3-year period for able-bodied adults who are not working,
participating in a work program for 20 hours or more each week, or in
workfare. States may request a waiver of this provision in areas with an
unemployment rate above 10 percent, or for those residing in an area that
has an insufficient number of jobs to provide employment for individuals.
In addition, waiver of this provision may also occur in recognition of the
challenges that low-skilled workers may face in finding and keeping
permanent employment. In some areas, including parts of rural America,
the number of unemployed persons and the number of job seekers may be
far larger than the number of vacant jobs. This may be especially so for
persons with limited skills and minimal work history.
August 2010

LAUS Program Manual 1-23

Department of Commerce:
Economic Development Administration,
Public Works Program
Program Objectives: To assist States and local areas in the development
and implementation of strategies designed to arrest and reverse the
problems associated with long-term economic deterioration. Grants are
provided to help distressed communities attract new industry, encourage
business expansion, diversify local economies, and generate long-term,
private sector jobs.
Among the types of projects funded are water and sewer facilities
primarily serving industry and commerce; access roads to industrial parks
or sites; port improvements; and business incubator facilities. Proposed
projects must be located within an EDA-designated Redevelopment Area
(RA) or Economic Development Center. Projects in other areas of an
EDA-designated Economic Development District are also eligible if they
will directly benefit a RA within the District. Projects must be consistent
with an approved Overall Economic Development Program (OEDP). An
applicant may be a State, political subdivision of a State, Indian tribe,
special-purpose unit of government, or a public or private nonprofit
organization or an association representing the RA or part thereof.

Economic DevelopmentAadministration,
Economic Adjustment
Program Objectives: The Economic Adjustment Program helps States and
local areas design and implement strategies for facilitating adjustment to
changes in their economic situation that are causing or threaten to cause
serious structural damage to the underlying economic base. Such changes
may occur suddenly (Sudden and Severe Economic Dislocation) or over
time (Long-Term Economic Deterioration) and result from industrial or
corporate restructuring, new Federal laws or requirements, reductions in
defense expenditures, depletion of natural resources, or natural disasters.
Strategy grants provide the recipient with the resources to organize and
carry out a planning process resulting in an adjustment strategy tailored to
the particular economic problems and opportunities of the impacted
area(s). Implementation grants may be used to support one or more
activities identified in an adjustment strategy approved, though not
necessarily funded, by EDA. Implementation activities may include, but
are not limited to: the creation or expansion of strategically targeted
business development and financing programs including grants for
revolving loan funds, infrastructure improvements, organizational
development, and market or industry research and analysis.
August 2010

LAUS Program Manual 1-24

Department of Defense:

Defense Logistics Agency—
Procurement Technical Assistance
Program Objectives: To provide funding assistance to civil jurisdictions
and nonprofit agencies working with small and disadvantaged businesses.
The purpose of the Procurement Technical Assistance (PTA) Cooperative
Agreement Program is to (1) generate employment and improve the
general economy of a locality by assisting business firms in obtaining and
performing under Federal, State, and local government contracts; (2)
increase Department of Defense assistance for eligible entities furnishing
PTA to business entities; and (3) assist eligible entities in the payment of
the costs of establishing and carrying out new PTA programs and
maintaining existing PTA programs.

Department of Health
and Human Services:
Temporary Assistance for Needy Families—
Contingency Fund Drawdown and Exemption from
Benefit Limitation
Program Objectives: Temporary Assistance for Needy Families (TANF)
was established under the Personal Responsibility and Work Opportunity
Reconciliation Act of 1996 to replace Aid to Families with Dependent
Children (AFDC), Job Opportunities and Basic Services (JOBS), and
Emergency Assistance (EA) programs. In order to receive the new block
grants under TANF, States must submit a State TANF plan outlining how
they intend to conduct a program that provides assistance to needy
families with children and provide parents with job preparation, work, and
support services to enable them to leave the program and become selfsufficient. States must submit plans every two years and may submit
amendments to keep the plan current whenever they wish to make changes
in the administration or operation of the program. In addition to State
plans, federally recognized Indian Tribes and approved Alaskan Native
entities are also eligible to submit TANF plans to the Secretary of Health
and Human Services.
August 2010

LAUS Program Manual 1-25

Department of Homeland
Security:
Federal Emergency
Management Agency:
Emergency Food and Shelter Program
Program Objectives: To help meet the needs of hungry and homeless
people in the U.S. and its territories by allocating Federal funds to the
neediest areas, ensuring quick response, fostering public and private
cooperation, ensuring local decision- making, and maintaining minimal
reporting. The program began in 1983 to help meet the needs of hungry
and homeless people throughout the United States and its territories by
allocating federal funds for the provision of food and shelter to those
impacted by natural disasters or emergencies. The program is governed
by a national board composed of representatives of the American Red
Cross; Catholic Charities, USA; Council of Jewish Federations; the
National Council of the Churches of Christ in the USA; the Salvation
Army; and the United Way of America. The national board awards funds
to jurisdictions based upon a formula. Once an award is made, local
boards decide which agencies are to receive funds, and then those agencies
are paid directly by the national board.
Funds are used to provide the following, as determined by
the local board in funded jurisdictions: (1) food, in the form
of served meals or groceries; (2) lodging in a mass shelter
or hotel; (3) one month’s rent or mortgage payment; (4) one
month’s utility bill; (5) minimal repairs to allow a mass
feeding or sheltering facility to function during the program
year; and (6) equipment necessary to feed or shelter people,
up to $300 per item.

U.S. Citizenship and Immigration Services,
Immigration Act of 1990 (IMMACT)
Program Objectives: To make 10,000 visas available each fiscal year to
qualified immigrants seeking to enter the U.S. for the purpose of engaging
in a new commercial enterprise. The new commercial enterprise may take
any lawful business form and must both benefit the U.S. economy and
create full-time employment for not fewer than 10 U.S. citizens, lawful
permanent residents, or other immigrants lawfully authorized to be
employed.
To encourage the establishment of new enterprises in areas which would
most benefit from employment creation, 3,000 of the employment creation
visas are reserved for qualified aliens who have made investments in
August 2010

LAUS Program Manual 1-26

“targeted employment areas.” Such areas are defined to include rural
areas and areas which have experienced high unemployment. A rural area
is defined as any area other than an area within a metropolitan statistical
area (MSA) or within the outer boundary of any town having a population
of 20,000 or more. An area of high unemployment under the Act is
defined as a non-rural area with an average unemployment rate of 150
percent of the national average in the previous calendar year.
Alternatively, a letter from an authorized body of the government may
certify that the area has been designated a high unemployment area.

Department of the Treasury:

Riegle Community Development and Regulatory
Improvement Act of 1994, Bank Enterprise Awards
Program and Small and Emerging CDFI Assistance
Component
Program Objectives: To promote the formation and expansion of
Community Development Financial Institutions (CDFIs); promote
community lending and investment activities by banks and thrifts; enhance
the liquidity of community lending products; and enhance the capacity of
CDFIs, banks and thrifts to engage in community lending and investment
activities. The Bank Enterprise Award Program is intended to encourage
banks and thrifts to invest in and support community development
financial institutions and to inc rease the lending and services provided in
distressed communities by traditional financial institutions.

North American Development Bank:

(NADBank), Community Adjustment and
Investment Program (CAIP)
Program Objectives: To finance community adjustment and investment
efforts throughout the United States and Mexico. U.S. appropriations of
$225 million will be leveraged into financing for community adjustment
projects that will provide significant benefits for U.S. citizens and
businesses. The NADBank's U.S. community adjustment window will
operate nationwide to offer financing directly through existing federal
credit programs to assist communities and businesses adjust to the new
trade environment created by NAFTA.
August 2010

LAUS Program Manual 1-27

Appalachian Regional Commission:

Area Development Program Distressed
Counties Grants
Program Objectives: The Commission was established to assist in the
long-term development of the chronically depressed region. Its main
objectives are the creation of new jobs and preparation of the people in
the region to compete for jobs wherever they choose to work and live.
The non-highway program focuses on the creation of new jobs and
private investments and special help for the region’s poorest or distressed
counties. New jobs and private investment are encouraged by grants
supporting education, water and sewer services for industrial and
commercial needs, housing, small business development, health care,
development of na tural resources, and research on topics directly related
to the region’s economic development.

Small Business Administration:

Historically Underutilized Business Zones
(HUBZones)
Program Objectives: To encourage economic development and create
jobs in urban and rural communities by providing contracting
preferences to small businesses located in and hiring employees from
historically underutilized business zones. A firm may be determined to
be a qualified HUBZone small business if it is located in a historically
underutilized business zone, it is owned and controlled by one or more
U.S. citizens, and at least 35 percent of its employees reside in a
HUBZone.
Under the program, three types of contracts exist: (1) A competitive
contract, in which at least two qualified small businesses are expected to
submit offers, and at least one of which will be at a fair market price; (2) a
sole source contract , and
(3) an open competition award, in which a qualified HUBZone small
business receives a price preference over another non-HUBZone bidder
that is other than small.
August 2010

LAUS Program Manual 1-28

The American Recovery and Reinvestment Act of 2009
The American Recovery and Reinvestment Act of 2009 (ARRA), also known as the “Recovery Act” or “Economic Stimulus
Act”, was designed to stimulate spending and limit economic downturn. It contains funds for both existing programs and new
programs. The ARRA allocates approximately $60,849.3 million to States based in whole or in part on LAUS estimates. The
details of the ARRA programs are described in the table below and on the Bureau’s website at
http://www.bls.gov/lau/lauarra.pdf.

ADMINISTRATIVE USES OF LOCAL AREA UNEMPLOYM ENT STATISTICS—American Recovery and Reinvestment Act Allocations
Increase
from FY 2009
Geographic Areas Used
Funding
(Millions)
Department of Labor – Employment and Training Administration
User Agency/Program

Adult Employment and Training
Activities (WIA, Title I, Chapter 5)

+$ 500.0

Reference Period

Allocation Formulas/Qualifying Criteria

States and Areas of Substantial
Unemployment (ASUs). An ASU is a
contiguous piece of geography consisting of
counties, cities, and/or parts of each, with a
population of at least 10,000 and an
unemployment rate of at least 6.5 percent.
(1) (2)

Average 12-months
ending June 30.

Funding based on the following proportions: 1/3 on relative number of unemployed in
ASUs, 1/3 on relative number of excess unemployed (i.e., number of unemployed in
excess of 4.5 percent of labor force), and 1/3 on relative number of economically
disadvantaged adults, age 22-72. Not more than 0.25% of funds allocated to out lying
areas. (Additional minimum/maximum provisions apply.)

Youth Activities (WIA, Title I,
Chapter 4)

+$ 1,200.0

States and ASUs. (1) (2) (3)

Average 12-months
ending June 30.

Funding based on the following proportions: 1/3 on relative number of unemployed in
ASUs, 1/3 on relative number of excess unemployed, and 1/3 on relative number of
economically disadvantaged youth, age 16-24. Not more than 0.25% of funds allocated
to outlying areas. Up to 1.5% allocated to Native American programs. (Additio nal
minimum/maximum provisions apply.)

Dislocated Worker Employment &
Training Activities (WIA, Title I,
Chapter 5)

+$ 1,250.0

States. (1) (2)

Average 12-months
ending Dec. 31.

Funding based on the following proportions: 1/3 on relative number of unemployed, 1/3
on relative number of excess unemployed, and 1/3 on relative number of individuals
unemployed for 15 weeks or more. Not more than 0.25% of funds allocated to outlying
areas.

+$400.0

States. (1)

Average 12-months
ending Dec. 31.

State funding algorithm is based on the following proportions: 2/3 on relative number
of civilian labor force and 1/3 on relative number of unemployed.

(4)

States. (1)

Most recent 3-month
average TUR or a 13week average IUR.

Expanded emergency unemployment compensation to 20 wks nationwide and created
2 nd tier of EUC08 for people in states with high unemployment rates. Tier 2 is available
for States with a 3-mo. avg. seasonally adjusted unemployment rate of at least 6% or a
13-wk avg. IUR of at least 4% and provides up to 13 more weeks of EUC08 benefits
(for a total of 33 wks).

Employment Service Grants to States
(Wagner-Peyser Act, Section 5)
Unemployment Compensation—
Emergency Unemployment
Compensation (EUC08)

August 2010

LAUS Program Manual 1-29

User Agency/Program

Increase
from FY 2009
Funding
(Millions)

Geographic Areas Used

Reference Period

Allocation Formulas/Qualifying Criteria

Department of Agriculture
The Emergency Food Assistance
Program (TEFAP)

+$150.0

States. (1) (2) (3)

October 2007 –
December 2008

Farm commodities and funds are allocated based on the following proportions: 3/5 on
relative number of persons in households below the poverty line and 2/5 on relative
number of unemployed persons.

Areas defined by geographic/political
boundaries, e.g., States, cities, counties,
Indian reservations. (1) (2) (3)

Most recent 24-month
period.

An area qualifies if: (1) the unemployment rate is at least one percentage point above
the national rate, (2) the per capita income is 80 percent or less of the national average
per capita income, or (3) there is a special need, as determined by EDA, arising from
actual or threatened severe unemployment or economic adjustment problems resulting
from severe short-term or long-term changes in economic conditions. [New legislation
authorizes ACS as the primary source of unemployment data, with BLS as the
secondary source.]

States

Most recent 3-month
average between Oct.
2008 and Dec. 2010.

States may receive additional assistance under the FMAP based on high unemployment.
Using the State’s lowest quarterly average unemployment rate beginning with January
2006 as its base quarter, if the State’s rate increased by at least 1.5 percentage points
over the base quarter it will receive an additional increase in it federal match rate. There
are 3 tiers of increasing federal assistance: Tier 1, 1.5-2.5 pts reduces the state’s share
by 5.5 %, Tier 2, 2.5-3.5 pts. reduces the state’s share by 8.5 %; Tier 3 3.5 pts or more
reduces the state’s share by 11.5%.

Latest annual average
unemployment rate for
the area.

Jurisdictions qualify for FEMA funding if they meet one of the following criteria: (1)
13,000 or more unemployed with a jobless rate of 5.0 percent or more, (2) 300-12,999
unemployed with a jobless rate of at least 7.0 percent, or (3) 300 or more unemployed
with a poverty rate of at least 11 percent.

Department of Commerce – Economic Development Administration
Economic Adjustment (Title 9)

+$150.0

Department of Health and Human Services--Medicaid
Federal Medical Assistance
Percentage (FMAP)

+87,000.0
(5)

Department of Homeland Security – Federal Emergency Management Agency
Emergency Food and Shelter Program
Counties and cities. (1) (2)
+100.0

Department of Justice
COPS-- Office of Community
Oriented Policing Services
Department of Transportation
Federal Highway Administration

August 2010

+1,000.0

Counties and municipalities

Jan 2008-Jan. 2009.

Unemployment rates along with population, poverty rates, foreclosure rates and fiscal
problems within the geographic area are used to determine qualification for funding.

+$27,500
(6)

Priority given to projects located in
economically distressed areas as defined by
section 301 of the Public Works and
Economic Development Act of 1965, as
amended (42 U.S.C. 3161)

Most recent 24-month
period.

An area qualifies if: (1) the unemployment rate is at least one percentage point above
the national rate, (2) the per capita income is 80 percent or less of the national average,
or (3) there is a special need, as determined by EDA.

LAUS Program Manual 1-30

User Agency/Program
Department of the Treasury
Community Development Financial
Institution Fund (CDFI)

Increase
from FY 2009
Funding
(Millions)
+$100.0

Internal Revenue Service—Recovery
Zone Development Bonds
Internal Revenue Service—Recovery
Zone Facility Bonds

+$10,000.0

Total Appropriations

$144,350.0

+$15,000.0

Geographic Areas Used

Reference Period

Same geographic areas used for the Bank
Enterprise Awards.

Most recent 12-month
period before
announcement of
application period.

States, counties, and large (>100,000
population) municipalities
States, counties, and large (>100,000
population) municipalities

Dec. 2007-Dec. 08
Dec. 2007-Dec. 08

Allocation Formulas/Qualifying Criteria

An institution may qualify if part or all of its service area: (1) is located within one unit
of general local government, (2) has a contiguous boundary, (3) (a) has a population of
4,000 or more, if in a metropolitan area, or (b) has a population of 1,000 or more, if
outside of a metropolitan area, or (c) is entirely within an Indian reservation, (4) has a
poverty rate of at least 30 percent, and (5) has an unemployment rate at least 1.5 times
the national rate.
Allocations to states, then substate areas based on Dec. 2007-Dec. 2008 employment
decline.
Allocations to states, then substate areas based on Dec. 2007-Dec. 2008 employment
decline.

NOTE: The term “cities” also includes townships and boroughs in selected states for various programs.
(1)
(2)
(3)
(4)
(5)
(6)

The District of Columbia and Puerto Rico are treated as states.
Outlying areas include the U.S. Virgin Islands, Guam, American Samoa, Northern Marianas Islands, Marshall Islands, Micronesia, and Palau.
Native American Program includes American Indians, Native Hawaiians, and Alaska Natives.
Under Emergency Unemployment Compensation, monies are not appropriated, but are drawn from the Federal Unemployment Account under Title 12 of the Social Security Act..
Dollar amount is full cost of allotment for Medicaid.
Dollar amount is full cost of allotment for Federal Highway Administration Highway Infrastructure Investment

August 2010

LAUS Program Manual 1-31

Department of Labor:

Employment and Training Administration,
Emergency Unemployment Compensation
Program Objectives: Emergency Unemployment Compensation (EUC08)
is a 100 percent federally funded program that provides benefits to
individuals who have exhausted regular State benefits. The EUC program
was created on June 30, 2008, and has since been modified several times.
EUC08 initially made up to 13 additional weeks of benefits available to
unemployed individuals who had already collected all regular State
benefits for which they were eligible and who meet the eligibility
requirements. The program was expanded to 20 weeks nationwide. A
second tier of benefits was created for individuals in States with high
unemployment rate providing 14 additional weeks of benefits. A third tier
was added providing 13 more weeks of benefits to individuals who have
exhausted their second tier benefits and a fourth tier provides another 6
weeks of benefits for individuals who have exhausted their third tier.

Department of Health and Human
Services:
Federal Medical Assistance Percentage (FMAP)
Program Objectives: The Federal Medical Assistance Percentages
(FMAPs) are used in determining the amount of Federal matching funds
for State expenditures for assistance payments for certain social services,
and State medical and medical insurance expenditures. The FMAPs are
calculated annually and are based upon a formula which compares
individual state income to the continental United States income in order to
determine ratios the federal government will utilize in assisting each State.
The American Recovery and Reinvestment Act of 2009 (ARRA) provides
additional funding under the FMAP based on high unemployment. For the
recession adjustment period (October 1, 2008, through December 31,
2010), the ARRA provides $87 billion in additional Medicaid funding
based on temporary increases in States’ Federal medical assistance
percentages (FMAP).
August 2010

LAUS Program Manual 1-32

Department of Justice:
Office of Community Oriented Policing Services
Program Objectives: The Office of Community Oriented Policing
Services (the COPS Office) is the component of the U.S. Department of
Justice responsible for advancing the practice of community policing by
the nation's state, local, territory, and tribal law enforcement agencies
through information and grant resources.
The American Recovery and Reinvestment Act provides $1 billion under
the COPS Hiring Recovery Program (CHRP) in grant funding for the
hiring and rehiring of additional career law enforcement officers.

Department of Transportation:
Federal Highway Administration
Program Objectives: The Highway Infrastructure Investment
appropriation of the American Recovery and Reinvestment Act enables
the Federal Highway Administration (FHWA) to give priority to projects
that are located in economically distressed areas.
An area is economically distressed if it has a per capita income of 80
percent or less of the national average or has an unemployment rate that is,
for the most recent 24- month period for which data are available, at least 1
percent greater than the national average unemployment rate.

August 2010

LAUS Program Manual 1-33

Department of the Treasury:
Internal Revenue Service,
Recovery Zone Development Bonds and
Recovery Zone Facility Bonds
Program Objectives: The American Recovery and Reinvestment Act
(ARRA) authorized State and local governments to issue Recovery Zone
Bonds. The ARRA imposes a national bond volume cap of $10 billion for
Recovery Zone Economic Development Bonds and $15 billion for
Recovery Zone Facility Bonds. The volume cap for Recovery Zone Bonds
is allocated among the States and counties and large municipalities within
the States based on relative declines in employment in 2008.
In general, Recovery Zone Economic Development Bonds may be used to
finance certain “qualified economic development purposes” and Recovery
Zone Facility Bonds may be used to finance certain “recovery zone
property,” both as described further herein, generally for use within
designated “recovery zones,” as described below.
The ARRA defines the term “recovery zone” as (1) any area designated by
the issuer as having significant poverty, unemployment, rate of home
foreclosures, or general distress; (2) any area designated by the issuer as
economically distressed by reason of the closure or realignment of a
military installation pursuant to the Defense Base Closure and
Realignment Act of 1990; and (3) any area for which a designation as an
empowerment zone or renewal community is in effect as of the effective
date of ARRA, which effective date is February 17, 2009.

August 2010

LAUS Program Manual 1-34

Office of Management and Budget Statistical
Policy Directive Number 11
Standard Data Source for Statistical Estimates
of Labor Force and Unemployment
Accurate, consistent, publicly available estimates of the labor force and of
unemployment in the Nation, the States, and local areas are needed for
use in the formulation, implementation, and evaluation of public policy.

1. Source of Data
Federal executive branch departments, agencies, and establishments
(hereinafter Federal executive branch agency) shall use the most current
national, State, or local area labor force or unemployment data published
by the Bureau of Labor Statistics, United States Department of Labor,
with respect to all program purposes, including the determination of
eligibility for and/or the allocation of Federal resources, requiring the use
of such data unless otherwise directed by statute. In order to maintain
equity among local areas, comparable data series are to be used for all
program purposes. Further, unless otherwise required by statute, data
adjusted for seasonal variation shall be used for all program purposes as
soon as the Bureau of Labor Statistics shall have published such data for
local areas being examined for the program purpose then under
consideration.
No Federal executive branch agency shall begin or continue collecting or
using State or local area labor force or unemployment data other than that
published by the Bureau of Labor Statistics, without the written approval
of the Secretary of Commerce. This does not preclude the collection of
labor force and unemployment data by the Bureau of the Census, United
States Department of Commerce, for the Bureau of Labor Statistics or in
its conduct of a periodic or other census or statistical survey, and the
publication or other distribut ion thereof.

2. Data Consistency
With respect to any month, a consistent reference time period shall be
used for all national, State, and local area labor force and unemployment
data. The data for each State and area, to the extent technically feasible,
shall be conceptually consistent with the data for the Nation as a whole
and the State totals shall sum, within a range of acceptable sampling error,
to the national total.

August 2010

LAUS Program Manual 1-35

3. Data Publication
The Bureau of Labor Statistics, in accordance with the provisions of
Directive No. 4, Prompt Compilation and Release of Statistical
Information, shall establish a monthly release date or dates for all
regularly published labor force and unemployment data and shall provide
the release date schedule to the Office of Federal Statistical Policy and
Standards for publication in the Statistical Reporter.
The monthly publication or publications by the Bureau of Labor Statistics
shall contain data for the Nation as a whole, and for each State and each
local area for which the Bureau of Labor Statistics has agreed to publish
data. No agreement between the Bureau of Labor Statistics and other
Federal executive branch agencies shall be used to limit the number or
types of areas for which data are developed and/or published by the
Bureau of Labor Statistics. The data published by area shall at a minimum
provide the current estimates before seasonal adjustment, and as soon as
possible, and to the extent technically feasible, shall also provide the
estimate adjusted for seasonality.

4. Notification of Data Need
Federal executive branch agencies requiring State and local area labor
force or unemployment data shall notify the Commissioner, Bureau of
Labor Statistics, United States Department of Labor, of their need for such
data. The notification shall include info rmation about the purpose for
which the data are needed and the specification(s) (i.e., statistical
reliability, geographic and other) for the data.
Any Federal executive branch agency required by legislation to use labor
force or unemployment data other than that directed by this Directive and
any Federal executive branch agency notified by the Commissioner,
Bureau of Labor Statistics that the needed data cannot be provided
according to specification shall notify the Director, Office of Federal
Statistical Policy and Standards, Department of Commerce of that fact.
The notification shall include identification of the program(s) affected,
legislation implemented by those programs, data specifications, and a
report on consultations with the Bureau of Labor Statistics in respect to
such data.

5. Definitions
a. Labor Force and Unemployment Data. The term labor force and
unemployment data is defined to include all counts or estimates of the
total labor force, the civilian labor force, total employment, total civilian
employment, total unemployment, and total unemployment rates. The
term excludes data, obtained solely from administrative records of the
unemployment insurance system, pertaining to counts of covered
August 2010

LAUS Program Manual 1-36

employment, the insured unemployed, and to the insured unemployment
rate.
b. Current Data. For the purposes of this Directive, the term current data
means the most current, complete data published by the Bureau of Labor
Statistics.
c. Local Area. A local area, for purposes of this Directive, is any
geopolitical unit of the United States of America and any combination or
part of any such unit or units.

August 2010

LAUS Program Manual 1-37

2

Inputs to LAUS Estimation
The Current Population Survey

Introduction

T

he Current Population Survey (CPS) is a monthly survey of about 60,000
households conducted by the Bureau of the Census for the Bureau of
Labor Statistics. The survey has been conducted since the late 1940’s.

The CPS is the primary source of information on the labor force characteristics of
the U.S. population. The sample is scientifically selected to represent the civilian
noninstitutional population: that is, all persons aged 16 and over, residing in the
fifty states and the District of Columbia, not on active duty in the Armed Forces,
and not inmates of institutions. The sample is a State-based design, with onefourth of the households changed each month to avoid placing too heavy a burden
on the households selected for the sample. Households are interviewed for four
months, are out of the sample for the next eight, then return to the sample for the
same four calendar months a year later. Respondents are interviewed to obtain
information about the employment status of each member of the household 15
years of age and older in the reference week that includes the 12th of the month.
The responses are used to publish data on those aged 16 and over. The sample
provides estimates for the nation as a whole and for individual states and other
geographic areas that are also used as inputs to LAUS model-based estimation.
Estimates obtained from the CPS include employment, unemployment, earnings,
hours of work, and other indicators. They are available by a variety of
demographic characteristics including age, sex, race, marital status, and
educational attainment. They are also available by occupation, industry, and class
of worker. Supplemental questions to produce estimates on a variety of topics
including school enrollment, income, previous work experience, health, employee
benefits, and work schedules are also often added to the regular CPS
questionnaire.

August 2010

LAUS Program Manual 2-1

CPS data are used by government policymakers and legislators as important
indicators of our nation's economic situation and for planning and evaluating
many government programs. They are also used by the press, students,
academics, and the general public.
Background
The Current Population Survey grew out of a program set up to provide direct
measurement of monthly unemployment, a problem that became especially
pressing during the Economic Depression of the 1930’s.
The Enumerative Check Census, taken as a part of the 1937 unemployment
registration, was the first attempt to estimate unemployment on a nationwide basis
using probability sampling. In addition, during the latter half of the 1930s, the
research staff of the Work Projects Administration (WPA -- known prior to 1939
as the Works Progress Administration) began developing techniques for
measuring unemployment -- first on a local-area basis and then nationally. This
research and experience led to the Sample Survey of Unemployment, which the
WPA began as a monthly activity in March 1940.
In August 1942, responsibility for the Sample Survey of Unemployment was
transferred to the Bureau of the Census, and its title was changed to The Monthly
Report on the Labor Force. In 1948, the survey was renamed as the Current
Population Survey. BLS assumed responsibility for its publication and analysis in
1959.
The CPS is the oldest continuous household survey in the world. It has been
regularly revised and updated to keep pace with statistical and technological
advances. Improvements in the identification of households covered in the
sample, sample design, methodology, and estimation procedures have been paired
with modifications to the questionnaire and interview process to ensure increased
reliability and efficiency. In 1957, the Bureau of the Census began seasonally
adjusting selected CPS data series with its X-11 model. In January 1989 the X-11
model was updated to the X-11 Auto-Regressive Integrated Moving Average
(ARIMA) method, and updated to the X-12 ARIMA method in January 2003.
In response to a need for more data at the subnational level, the 1970s saw a series
of State supplementary sample expansions to the CPS which, at that time,
employed a national sample design. While the expansions provided reliable CPS
annual average benchmarks in all States, they were recognized as inefficient ways
of developing State estimates. In 1985, the national-based design was changed to
a State-based sampling design. This design required that annual average State
estimates fall within specified levels of reliability, while not adversely affecting
the reliability of national estimates.

August 2010

LAUS Program Manual 2-2

Important technological advances in data collection have also been implemented.
In 1994, computer assisted telephone interviewing (CATI) and computer assisted
personal interviewing (CAPI), along with a new questionnaire design, were
phased in to aid in the collection and reliability of the data.
In September 2000, the Census Bureau began augmenting the monthly CPS
sample in 31 states and the District of Columbia, as one part of the Census
Bureau’s plan to meet the requirements to produce certain estimates for the State
Children’s Health Insurance Program (SCHIP). States were identified for sample
supplementation based on the standard error of their March estimate of lowincome children without health insurance. The additional 10,000 households
were added to the sample over a 3- month period. Thus, starting with July 2001
data, official labor force estimates from the CPS and Local Area Unemployment
Statistics (LAUS) program reflected the expansion of the monthly CPS sample
from about 50,000 to about 60,000 eligible households.

Survey Process
The CPS survey process consists of three main phases: sampling, data collection,
and estimation.
Sampling involves (1) the determination, stratification, and selection of a sample
of Primary Sampling Units (PSUs) and (2) the selection of sample households
within those PSUs.
Data collection involves interviewers asking households about activities during
the reference week, which contains the 12th day of the month. A questionnaire is
completed for each household member 15 years of age and over. BLS
determines the labor force status for each household member for the month using
the information captured in the questionnaires.
Estimation is the process of taking sample data and making estimates for the
population as a whole. Estimation involves a number of steps including data
editing and imputation, basic weighting, non- interview adjustment, ratio
adjustment, compositing of estimates, and seasonal adjustment.

August 2010

LAUS Program Manual 2-3

CPS Labor Concepts and Definitions
In the CPS, persons are classified as “employed,” “unemployed,” or “not in the
labor force.” These classifications are mutually exclusive and based on a person's
labor force status during the survey reference week (the week including the 12th
of the month). Each person is classified according to the activities he/she engaged
in during the reference week, as defined by the set of questions in the survey.
Respondents are never asked specifically if they are unemployed, nor are they
given any opportunity to decide their own labor force status. Interviewers do not
determine a person’s labor force status. They simply ask the series of questions
and record the answers. (See Table 2-1 for the questions used to classify an
individual as employed or unemployment.)
Because the CPS is a household-based survey, it counts each person only once, at
their place of residence, even if they hold more than one job. It thus produces an
unduplicated count of employed and unemployed persons. In contrast, the
Current Employment Statistics (CES) survey is establishment-based and designed
to produce counts of the number of jobs in the economy: Persons holding more
than one job will be counted more than once, depending on which establishments
were in the survey sample. Since the LAUS program uses CPS concepts that
reflect resident employed, adjustment must be made to make the CES data reflect
a residency basis. (See section on Dynamic Residency Ratios.)
Labor Force
Labor force information is obtained after the household and demographic
information has been collected. One of the primary purposes of the labor force
information is to classify individuals as employed, unemployed, or not in the
labor force. Other information collected includes hours worked, occupation, and
industry and related aspects of the working population. It should be noted that the
major labor force categories are defined hierarchically and, thus, are mutually
exclusive. Employed supersedes unemployed which supersedes not in the labor
force. For example, individuals who are classified as employed, even if they
worked less than full time, are not asked the questions about having looked for
work, and hence cannot also be classified as unemployed. Similarly, an
individual who is classified as unemployed is not asked the questions used to
determine one’s primary non- labor market activity. For instance, retired persons
who are currently working are classified as employed, even though they have
retired from their previous jobs. Consequently, they are not asked the questions
about their previous employment nor can they be classified as retired.
The concepts and definitions underlying the collection and estimation of the labor
force data are presented below.
Reference week: The CPS labor force questions ask about labor market activities
for one week each month. This week is referred to as the “reference week.” The
August 2010

LAUS Program Manual 2-4

reference week is defined as the 7-day period, Sunday through Saturday that
includes the 12th of the month. (On occasion, the reference week in November
and December may be week including the 5th of the month, to facilitate data
collection during the holiday period.)
Civilian Noninstitutional Population (CNP): This includes all persons 16 years
of age and older residing in the 50 States and the District of Columbia who are not
inmates of institutions (e.g. penal and mental facilities, and homes for the aged)
and who are not on active duty in the Armed Forces. This is the base population
used in the calculation of labor force statistics.
Employed persons . This includes all persons who, during the reference week,
(a) did any work at all (at least 1 hour) as paid employees; worked in their own
business, profession, or on their own farm; or who worked 15 hours or more as
unpaid workers in an enterprise operated by a member of the family, or (b) were
not working, but had jobs or businesses from which they were temporarily absent
because of vacation, illness, bad weather, childcare problems, maternity or
paternity leave, labor-management dispute, job training, or other family or
personal reasons, whether or not they were paid for the time off or were seeking
other jobs.
Each employed person is counted only once, even if he or she holds more than
one job. Multiple jobholders are counted in the job at which they worked the
greatest number of hours during the reference week. (See the discussion of
multiple jobholders below.)
Included in the total are employed citizens of foreign countries who are residing
in the United States and do not live on the premises of an embassy. Excluded are
persons whose only activity consisted of work around their own home (such as
housework, painting, repairing, etc.) or volunteer work for religious, charitable, or
similar organizations.
The initial survey question, asked only once for each household, inquires whether
anyone in the household has a business or farm. Subsequent questions are asked
for each household member to determine whether any of them did any work for
pay (or profit if there is a household business) during the reference week. If no
work for pay or profit was performed and a family business exists, respondents
are asked whether they did any unpaid work in the family business or farm. (See
Table 2-0 for the questions used to classify an individual as employed.)
Multiple jobholders . Persons who, during the reference week, had either two or
more jobs as a wage and salary worker, were self-employed and also held one or
more wage and salary jobs, or worked as an unpaid family worker and also held
one or more wage and salary jobs. A person employed only in private households
(cleaner, gardener, baby-sitter, etc.) is not counted as a multiple jobholder, even if
that person works for more than one employer. Working for several employers is
August 2010

LAUS Program Manual 2-5

considered an inherent characteristic of private household work. Also excluded
are self-employed persons with multiple unincorporated businesses and persons
with multiple jobs as unpaid family workers.
Since 1994, CPS respondents have been asked questions each month to identify
multiple jobholders. First, all employed persons are asked “Last week, did you
have more than one job (or business, if one exists), including part-time, evening,
or weekend work?” Those who answer “yes” are then asked, “Altogether, how
many jobs (or businesses) did you have?” Prior to 1994, this information had
only been available through periodic CPS supplements.
Hours of work: Beginning with the CPS redesign in January 1994, both actual
and usual hours of work have been collected. Prior to the redesign, only actual
hours were requested for all employed individuals.
Published data on hours of work relate to the actual number of hours spent “at
work” during the refe rence week. For example, persons who normally work 40
hours a week, but were off on the Veterans’ Day holiday, would be reported as
working 32 hours, even though they wee paid for the holiday. For persons
working in more than one job, the published figures relate to the number of hours
worked on all jobs during the week.
Data on persons “at work” exclude employed persons who were absent from their
jobs during the entire reference week for reasons such as vacation, illness, or
industrial dispute. Data also are available on usual hours worked by all employed
persons, including those who were absent from their jobs during the reference
week.
At work part time for economic reasons. Sometimes referred to as involuntary
part-time work, this category refers to individuals who gave an economic reason
for working 1 to 34 hours during the reference week. Economic reasons include
slack work or unfavorable business conditions, inability to find full-time work,
and seasonal declines in demand. Those who usually work part time must also
indicate that they want and are available to work full time to be classified as being
part time for economic reasons.
At work part time for noneconomic reasons. This group includes those persons
who usually work part time and were at work 1 to 34 hours during the reference
week for a noneconomic reason. Noneconomic reasons include illness or other
medical limitation, childcare problems or other family or personal obligations,
school or training, retirement or Social Security limits on earnings, and being in a
job where full- time work is less than 35 hours. The group also includes those
who gave an economic reason for usually working 1 to 34 hours but said they do
not want to work full time or were unavailable for such work.

August 2010

LAUS Program Manual 2-6

Usual full- or part-time status. In order to differentiate a person’s normal
schedule from his/her activity during the reference week, persons are also
classified according to their usual full- or part-time statuses. In this context, fulltime workers are those who usually work 35 hours or more (at all jobs combined).
This group includes some individuals who worked less than 35 hours in the
reference week—for either economic or noneconomic reasons—as well as those
who are temporarily absent from work. Similarly, part-time workers are those
who usually work less than 35 hours per week (at all jobs), regardless of the
number of hours worked in the reference week. This may include some
individuals who actually worked more than 34 hours in the reference week, as
well as those who were temporarily absent from work. The full- time labor force
consists of employed persons who usually work full time and unemployed
persons who are either looking for full-time work or are on layoff from full-time
jobs. The part-time labor force consists of employed persons who usually work
part time and unemployed persons who are seeking or are on layoff from parttime jobs.
Prior to 1994, persons who worked full time during the reference week were not
asked about their usual hours. Rather, it was assumed that they usually worked
full time, and hence they were classified as full- time workers.
Occupation, industry, and class-of-worker. For the employed, this information
applies to the job held in the reference week. A person with two or mo re jobs is
classified according to the job at which he or she worked the greatest number of
hours. The unemployed are classified according to their last jobs.
The class-of-worker classification assigns workers to one of the following
categories: wage and salary workers, self- employed workers, and unpaid family
workers. Wage and salary workers are those who receive wages, salary,
commissions, tips, or pay in kind from a private employer or from a government
unit. The class-of-worker question also includes separate response categories for
“private for profit company” and “nonprofit organization” to further classify
private wage and salary workers.
Self-employed persons are those who work for profit or fees in their own
businesses, professions, trades, or farms. Only the unincorporated self- employed
are included in the self-employed category since those whose business are
incorporated technically are wage and salary workers because they are paid
employees of a corporation.
Unpaid family workers are persons working without pay for 15 hours a week or
more on a farm or in a business operated by a member of the household to whom
they are related by birth or marriage.
Occupation, industry, and class-of-worker on second job. The occupation,
industry, and class-of-worker information for individuals’ second jobs is collected
August 2010

LAUS Program Manual 2-7

in order to obtain a more accurate measure of multiple jobholders, to obtain more
detailed information about their employment characteristics, and to provide
information necessary for comparing estimates of number of employees in the
CPS and in the BLS establishment survey (the Current Employment Statistics).
For the majority of multiple jobholders, occupation, industry, and class-of-worker
data for their second jobs are collected only from a quarter of the sample—those
in their fourth or eighth monthly interviews. However, for those classified as
“self employed unincorporated” on their main jobs, class-of-worker of the second
job is collected each month. This is done because individuals who are selfemployed unincorporated on both of their jobs are not considered multiple
jobholders.
Earnings. Information on what people earn at their main jobs is collected only
for those who are receiving their fourth or eighth monthly interviews. This means
that earnings questions are asked of only one-fourth of the survey respondents.
Respondents are asked to report their usual earnings before taxes and other
deductions and to include any overtime pay, commissions, or tips usually
received. The term “usual” is perceived by the respondent. If the respondent asks
for a definition of “usual”, however, interviewers are instructed to define the term
as more than half the weeks worked during the past 4 or 5 months. Respondents
may report earnings in the time period they prefer—for example, hourly, weekly,
biweekly, monthly, or annually. (Based on additional information collected in the
interview, earnings reported on a basis other than weekly are converted to a
weekly amount in later processing.) Data are collected for wage and salary
workers, excluding the self-employed who respond that their businesses were
incorporated. These earnings data are used to construct estimates of the
distribution of usual weekly earnings and median earnings. Individuals who do
not report their earnings on an hourly basis are asked if they are, in fact, paid at an
hourly rate and if so, what the hourly rate is. The earnings of those who reported
hourly and those who are paid at an hourly rate are used to analyze the
characteristics of hourly workers, for example, those who are paid the minimum
wage.
Unemployed persons. All persons who were not employed during the reference
week but were available for work (excluding temporary illness) and had made
specific efforts to find employment some time during the 4-week period ending
with the reference week are classified as une mployed. Individuals who were
waiting to be recalled to a job from which they had been laid off need not have
been looking for work to be classified as unemployed.
A relatively minor change was incorporated into the definition of unemployment
with the implementation of the 1994 redesign. Under the formal definition,
persons who volunteered that they were waiting to start a job within 30 days (a
very small group numerically) were classified as unemployed, whether or not they
were actively looking for work. Under the new definition, by contrast, people
August 2010

LAUS Program Manual 2-8

waiting to start a new job must have actively looked for a job within the last 4
weeks in order to be counted as unemployed. Otherwise, they are classified as not
in the labor force
As the definition indicates, there are two ways people may be classified as
unemployed. They are either looking for work (job seekers) or they have been
temporarily separated from a job (persons on layoff). Job seekers must have
engaged in active job search during the above mentioned 4-week period in order
to be classified as unemployed. Active methods are defined as job search
methods that have the potential to result in a job offer without any further action
on the part of the job seeker. Examples of active search methods include going to
any employer directly or to a public or private employment agency, seeking
assistance from friends or relatives, placing or answering ads, or using some other
active method. Other active methods include being on a union or professional
register, obtaining assistance from a community organization, or waiting at a
designated labor pickup point. Passive methods, which do not qualify as job
search, include reading (as opposed to answering or placing) “help wanted” ads
and taking a job training course. The response categories for active and passive
methods are clearly delineated in separately labeled columns on the interviewers’
computer screens.
Job search methods are identified by the following questions: “Have you been
doing anything to find work during the last 4 weeks?” and “What are all the
things you have done to find work during the last 4 weeks?” To ensure that
respondents report all of the methods of job search used, interviewers ask
“Anything else?” after the initial or subsequent job search method is reported.
Persons on “layoff” are defined as those who have been separated from a job to
which they are waiting to be recalled (i.e., their layoff status is temporary). In
order to measure layoffs accurately, the questionnaire determines whether people
reported to be on layoff did in fact have an expectation of recall; that is, whether
they had been given an indication that they would be recalled within the next 6
months. As previously mentioned, persons on layoff need not be actively seeking
work to be classified as unemployed. (See Table 2-0 for the questions used to
classify an individual as unemployed.)
Reasons for unemployment. Unemployed individuals are categorized according
to their status at the time they became unemployed. The categories are:
1) Job losers: a group comprised of (a) persons on temporary layoff from a
job to which they expect to be recalled and (b) permanent job losers,
whose employment ended involuntarily and who began looking for work.
2) Job leavers: persons who quit or otherwise terminated their employment
voluntarily and began looking for work.
3) Persons who completed temporary jobs: persons who began looking for
work after their job ended.
August 2010

LAUS Program Manual 2-9

4) Reentrants: persons who previously worked but were out of the labor
force prior to beginning their job search.
5) New entrants: persons who never worked before and who are entering the
labor force for the first time.
Each of these five categories of unemployed can be expressed as a proportion of
the entire civilian labor force or as a proportion of the total unemployed.
Prior to 1994, new entrants were defined as job seekers who had never worked at
a full-time job lasting 2 weeks or longer; reentrants were defined as job seekers
who had held a full-time job for at least 2 weeks and had then spent some time out
of the labor force prior to their most recent period of job search. These definitions
have been modified to encompass any type of job, not just a full-time job of at
least 2 weeks duration.
Duration of unemployment. The duration of unemployment is expressed in
weeks. For individuals who are classified as unemployed because they are
looking for work, the duration of unemployment is the length of time (through the
current reference week) that they have been looking for work. For persons on
layoff, the duration of unemployment is the number of full weeks (through the
reference week) they have been on layoff.
Not in the labor force. Included in this group are all persons in the civilian
noninstitutional population who are neither employed nor unemployed.
Information is collected on their desire for and availability to take a job at the
time of the CPS interview, job search activity in the prior year, and reason for not
looking in the 4-week period prior to the survey week. This group includes
discouraged workers, defined as persons not in the labor force who want and are
available for a job and who have looked for work sometime in the past 12 months
(or since the end of their last job if they held one within the past 12 mo nths), but
are not currently looking, because they believe there are no jobs available or there
are none for which they would qualify. (Specifically, the main reason identified
by discouraged workers for not recently looking for work is one of the following:
Believes no work available in line of work or area; could not find any work; lacks
necessary schooling, training, skills, or experience; employers think too young or
too old; or other types of discrimination.)
Data on a larger group of persons outside the labor force, one that includes
discouraged workers as well as persons who desire work but give other reasons
for not searching (such as childcare problems, family responsibilities, school, or
transportation problems) are also published regularly. This group is made up of
persons who want a job, are available for work, and have looked for work within
the past year. This group is generally described as having some marginal
attachment to the labor force.

August 2010

LAUS Program Manual 2-10

Prior to January 1994, questions about the desire for work among those who were
not in the labor force were asked only of a quarter of the sample. Since 1994,
these questions have been asked of the full CPS sample. Consequently, since
1994, estimates of the number of discouraged workers as well as those with a
marginal attachment to the labor force are published monthly rather than just
quarterly.
Estimates of the number of employed and unemployed are used to construct a
variety of measures. They include:
•
•
•
•

Labor force: The labor force consists of the all persons 16 years of age
and older classified as employed or unemployed in accordance with the
criteria described above.
Unemployment rate: The unemployment rate represents the number of
unemployed as a percentage of the labor force.
Labor force participation rate: The labor force participation rate is the
proportion of the age-eligible population that is in the labor force.
Employment-population ratio: The employment-population ratio
represents the proportion of the age-eligible population that is employed.

The CPS Survey Questionnaire for Employed and Unemployed
1. Does anyone in this household have a business or a farm?
2. LAST WEEK, did you do ANY work for (either) pay (or profit)?
(If 1 is “yes” and 2 is “no”, ask 3. If 1 is “no” and 2 is “no”, ask 4.)
3. LAST WEEK, did you do any unpaid work in the family business or farm?
(If 2 and 3 are both “no”, ask 4.)
4. LAST WEEK, (in addition to the business) did you have a job, either full or
part time? Include any job from which you were temporarily absent.
( If 4 is “no”, ask 5.)
5. LAST WEEK, were you on layoff from a job?
(If 5 is “yes”, ask 6. If 5 is “no”, ask 8.)
6. Has your employer given you a date to return to work?
(If “no”, ask 7.)

August 2010

LAUS Program Manual 2-11

7. Have you been given any indication that you will be recalled to work within
the next 6 months?
(If 7 is “no”, ask 8.)
8. Have you been doing anything to find work during the last 4 weeks?
(If “yes”, ask 9).
9. What are all of the things you have done to find work during the last 4 weeks?
Individuals are classified as employed if they say “yes” to questions 2, 3 (and work 15
hours or more in the reference week or receive profits from the business/farm), or 4.
Individuals who are available to work are classified as unemployed if they say “yes” to 5
and either 6 or 7, or if they say “yes” to 8 and provide a job search method that could
have brought them into contact with a potential employer in 9.

August 2010

LAUS Program Manual 2-12

Reliability of CPS Estimates
The two types of errors possible in an estimate based on a sample survey are
sampling and nonsampling.
Nonsampling error arises from errors in the collection and processing of data.
These errors -- including response variability, response bias, other types of bias,
and processing error -- occur in complete censuses as well as sample surveys. In
some instances, nonsampling error can be more tightly controlled. For example,
in a well- conducted survey, it is feasible to collect and process the data more
skillfully. Reinterview programs are often used to measure response variability
and response bias. However, estimation of other types of bias is very difficult and
often adequate measures of bias cannot be made.
Sampling error occurs because only a sample of the population has been
surveyed. Standard errors can be estimated when the probability of selection of
each member of a population can be specified. These standard errors can be used
to compute confidence intervals that indicate the range within which the true
population values likely lie.
Nonsampling Error
The full extent of nonsampling error is unknown, but special studies have been
conducted to quantify some sources of nonsampling error in the CPS. The effect
of the error has been found to be small on estimates of change, such as month-tomonth change ; however, estimates of monthly levels are generally more severely
affected.
Some specific types of nonsampling errors affecting the CPS include response
error, nonresponse error, error in independent population controls, processing
error, and coverage error.
• Response Error. This error arises when survey respondents’ answers are
incomplete or inconsistent with reality. It includes the inability to obtain
information about all persons in the sample, differences in the interpretation of
questions, inability or unwillingness of respondents to provide correct
information, and inability to recall information. These errors are studied by
means of a reinterview program: A random sample of each interviewer’s work is
inspected through reinterview at regular intervals. This program is used to
estimate various sources of error as well as to evaluate and control the work of the
interviewer. Results indicate that the data published from the CPS are subject to
moderate systematic biases.
• Nonresponse Error. This error arises in situations when respondents fail to
answer some or all of the questions. In a typical month, about 7-8 percent of
occupied sample households are not interviewed because residents are not at
August 2010

LAUS Program Manual 2-13

home, refuse to cooperate, or are unavailable. Therefore, sample weights are
adjusted to account for households not interviewed. To the extent that
interviewed households differ from those not interviewed, the estimates are
biased. Similarly, for a relatively few households, some questions are left
unanswered, either because respondents were unable or unwilling to answer or
because of interviewer error. Entries for omitted items are usually imputed on the
basis of the distributions of these items for persons of similar demographic
characteristics.
• Independent Population Controls. These are used to account for population
changes in intercensal years. They are extrapolated using the 2000 Census as a
base, using data on births, deaths, and net migration. Although the use of
independent population estimates in the estimation procedure substantially
improves the statistical reliability of many CPS estimates, the independent
estimates are also subject to error in both the base and the change factors. Base
errors may arise because of under-enumeration of certain population groups,
errors in age reporting in the last census, or similar problems in the components of
population change (mortality, immigration, etc.) since that date. Also, errors in
estimated components of change since the last census affect the accuracy of
intercensal population estimates.
• Processing Error Although the CPS employs computer-assisted interviewing
and a quality control program on coding and all other phases of data processing,
some processing error is inevitable in large surveys. Net CPS processing error is
probably negligible relative to sampling error and other nonsampling errors.
• Coverage Error. Undercoverage in the CPS results from missed housing units
and missed persons within sample households. The CPS covers about 92 percent
of the decennial census population. It is known that the CPS undercoverage
varies with age, sex, race, and Hispanic origin. Generally undercoverage is larger
for men than for women and larger for non-whites than for whites.
Ratio adjustment to independent age-sex-race-origin population controls,
described later in the Estimation section, partially corrects for the biases due to
survey undercoverage. However, biases exist in the estimates to the extent that
missed persons in missed households or missed persons in interviewed
households have different characteristics than interviewed persons in the same
age-sex-race-origin group.
Sampling Error
When a sample is surveyed, estimates are derived for the whole population, based
on the sample data and current statistical theory. Since these results are estimates,
they differ from the true population values that they mean to represent. This is
sampling error, and the variability of this sampling error is measured by the
standard error of the estimate. A given survey design is considered to produce
August 2010

LAUS Program Manual 2-14

unbiased estimates if the average of the estimates from all possible samples would
yield, hypothetically, the true population value. If this is the case, the sample
estimate and its standard error can be used to construct approximate confidence
intervals – ranges of values – that include the true population value with known
probabilities. If the process of selecting a sample from the population were
repeated many times and an estimate and its standard error calculated for each
sample, then:
1.) Approximately 68 percent of the intervals from one standard error below the
estimate to one standard error above the estimate would include the true
population value.
2.) Approximately 90 percent of the intervals from 1.6 standard errors below the
estimate to 1.6 standard errors above the estimate would include the true
population value.
3.) Approximately 95 percent of the intervals from two standard errors below the
estimate to two standard errors above the estimate would include the true
population value.
Although the estimating methods used in the CPS do not produce unbiased
estimates, biases for most estimates are believed to be small enough so that these
confidence interval statements are approximately true. Generalized variance
function techniques are used to calculate sets of standard errors for various types
of labor force characteristics. Standard errors computed from these methods
reflect contributions from sampling errors and some kinds of nonsampling errors
and indicate the general magnitude of an estimate’s standard error rather than its
precise value. Standard error tables for national estimates are provided in the
monthly publication Employment and Earnings, with State and Regional
estimates provided on the LAUS website.
CPS Monthly and Annual Reliability Criterion
Data reliability is measured by calculating the coefficient of variation (CV) of the
unemployment level; the CV is defined as the standard error of the estimate
divided by the estimate itself. The CPS sample design takes into consideration
both national and State reliability. The sample design, including the SCHIP
expansion, maintains a 1.8 percent CV on national monthly estimates of
unemployment level. A 6-percent unemployment rate is assumed. This means a
month-to-month change in the unemployment rate must be at least 0.2 percent to
be considered statistically significant at a 90-percent confidence level.
For each of the 50 States and for the District of Columbia, the design maintains a
CV of at most 8 percent on the annual average estimate of unemployment level,
again assuming a 6-percent unemployment rate. Due to the national reliability
criterion, samples for the more populous States are substantially larger than the
August 2010

LAUS Program Manual 2-15

State design criterion requires. As a result, annual average unemployment
estimates for large States such as California, Florida, New York, and Texas, for
example, carry a CV of less than 5 percent.

August 2010

LAUS Program Manual 2-16

Sample Design
Introduction
The CPS sample design has undergone several changes throughout its history.
After each decennial census, the sample is redesigned and a new sample selected.
Also, occasionally the number of sample areas and of sample persons is changed.
Most changes are made to improve the efficiency of the sample design, increase
the reliability of the sample estimates, or control cost. The current CPS sample is
designed to produce reliable monthly unemployment estimates for the nation and
reliable annual average estimates for the 50 States and the District of Columbia.
In the first stage of sampling, the 754 sample areas, called Primary Sampling
Units (PSUs), are chosen via stratification (see below). In the second stage,
Ultimate Sampling Unit (USUs) clusters, composed of about four housing units
each, are selected. Sample sizes and sampling rates are determined by the
specified reliability requirements. While the best estimates of month-to-month
change would be obtained by surveying the same households each month,
indefinitely surveying a single sample of households would inevitably lead to
respondent ‘fatigue,’ increasing the probability of respondent refusals and errors.
Therefore, a sample rotation scheme is used, with chosen households interviewed
for eight out of sixteen months, and then leaving the sample.
Selection of Primary Sampling Units (First Stage of Sampling)
The entire area of the United States, consisting of 3,141 counties and independent
cities, is divided into 2,007 PSUs. Each PSU consists of a county or group of
contiguous counties and is defined within State boundaries.
Metropolitan areas within a State are used as a basis for forming many PSUs.
Outside of metropolitan areas, two or more counties are normally combined to
form PSUs except where the geographic area of the sample county is too large.
Combining counties to form a PSU provides greater heterogeneity; a typical PSU
includes urban and rural residents of both high and low economic levels, and
encompasses, to the extent feasible, diverse occupations and industries. Another
important consideration is that the PSU be sufficiently compact so that, with a
small sample spread throughout, it can be efficiently canvassed without undue
travel cost.
Stratification of Primary Sampling Units
The 2,007 PSUs are grouped into strata within each State. Then, one PSU is
selected from each stratum with the probability of selection proportional to the
population of the PSU. Nationally, there are a total of 428 PSUs in strata by
themselves. These strata are self-representing and generally are the most populous
PSUs in each State. The 326 remaining strata are formed by combining PSUs that
August 2010

LAUS Program Manual 2-17

are similar in such characteristics as unemployment, proportion of housing units
with three or more persons, number of persons employed in various industries,
and average monthly wages for various industries. The single PSU randomly
selected from each of these strata is non-self-representing because it represents
not only itself but all PSUs within the stratum. The probability of selecting a
particular PSU in a non-self-representing stratum is proportional to its 2000
population. For example, within a stratum, the chance that a PSU with a
population of 50,000 would be selected for the sample is twice that for a PSU
having a population of 25,000.
Selection of Households Using Census Data
Because the sample design is State-based, the sampling ratios differ by State and
depend on State population sizes as well as national and State reliability
requirements. The State sampling ratios range roughly from 1 in every 200
households to 1 in every 3,000 households. The sampling ratio used within a
sample PSU depends on the probability of selection of the PSU and the sampling
ratio for the State.
The 2000 within-PSU sample design uses census block level data from the 2000
decennial census. Normally, census blocks are bounded by streets and other
prominent physical features such as rivers or railroad tracks. County, minor civil
division, and census place limits also serve as block boundaries. In cities, blocks
can be bounded by four streets and be quite small in land area. In rural areas,
blocks can be several square miles in size.
For purposes of sample selection, census blocks are grouped into three strata:
“unit”, “group quarters”, and “area”. The unit stratum contains regular housing
units with addresses that are easy to locate (e.g., most single family homes,
townhouses, condominiums, apartment units, and mobile homes). The group
quarters stratum contains housing units where residents share common facilities
or receive formal or authorized care or custody. These two strata exist primarily
in urban and suburban areas. The area stratum contains blocks with addresses that
are more difficult to locate. Area blocks exist primarily in rural areas.
These strata are then sampled using sampling intervals which preserve each
individual State’s sampling ratio. To reduce the variability of the survey
estimates and to ensure that the within-PSU sample reflects the demographic and
socioeconomic characteristics of the PSU, blocks within the unit, group quarters,
and area strata are sorted using geographic and block-level data from the census.
Examples of the census variables used for sorting include proportion of minority
renter-occupied housing units, proportion of housing units with female
householders, and proportion of owner-occupied housing units.
By grouping, sorting, and systematically sampling blocks in these strata, the
sampling process insures that the ultimate sampling units (USUs) selected within
August 2010

LAUS Program Manual 2-18

the PSU reflect the demographic and socio-economic characteristics of the PSU
as a whole. This design reduces the within-PSU variance, compared to the
variances associated with a simple random sample of units within the PSU.
Most USUs are formed of four housing unit addresses that are in the same general
neighborhood or block. There is some variation in the number of addresses in a
USU, primarily because the number of addresses in a block may not be evenly
divisible by 4. The number of housing units found in a USU may be more or
fewer than the number of addresses for the following reasons: major building or
demolition within the area since the census, left over housing units that did not fit
into other USUs, and rezoning may have created more units where previously
there had been less. Special procedures are used in the group quarters strata for
identifying approximate housing unit equivalents and USUs. It is more efficient
in terms of cost (interviewer travel time and time spent locating housing units) to
sample compact clusters housing units (USUs), as opposed to sampling individual
housing units. Due to privacy concerns, housing units within a USU are not
adjacent to each other, thus making it more difficult to uncover the identity of
individual respondents.
The USUs for the CPS sample are systematically selected from sorted lists of
blocks and housing units prepared as part of the decennial census. It is both
unfair and impractical to have a single panel of housing units in the sample for an
entire decade that follows. Instead, USUs are rotated in-and-out of the CPS
sample on a fixed schedule; all housing units in a sampled USU come into the
sample and are retired from the sample at the same time. A particular USU is
rotated into the sample for 4 consecutive months, is temporarily rotated out for 8
months, is rotated back in for another 4 consecutive months, then is again rotated
out and permanently retired from the sample. It is advantageous to replace a USU
that is rotated out with a USU that is comprised of housing units from the same
general neighborhood This is accomplished by initially setting the systematic
sampling parameters in each State and stratum (unit, group quarters, area) so that
USUs are selected that include a sufficient number of households for a single
monthly sample for that State. Within a block stratum, for each sampled USU,
the ensuing 20 USUs are also sampled. These 21 USUs are called a hit string.
When a USU in a hit string is rotated out of the sample, it is replaced by rotating
in another USU from the same hit string. That is, when one set of housing units
drops out of the sample rotation, another set from the same neighborhood is ready
to take its place.
Units in the three strata described above all existed at the time of the 2000
decennial census. A sample of building permits, collected by the States in an
ongoing cooperative procedure, is included in the CPS to represent housing units
built after the decennial census. Adding these newly-built units keeps the sample
up-to-date and representative of the population. It also helps to keep the sample
size stable. Over the life of the sample, the addition of newly-built housing units
compensates for the loss of “old” units which may be abandoned, demolished, or
August 2010

LAUS Program Manual 2-19

converted to nonresidential use. It is common for State samples to slowly grow
over a decade, and systematic "sample maintenance reductions" are sometimes
needed to return State samples to their budgeted sizes.
CPS State Sample Sizes and Sampling Ratios
The CPS sample is selected from within the PSUs identified above. The CPS has
a State-based sample design which allocates the sample in such a way that each of
the States and the District of Columbia has the same minimum target reliability on
their annual average estimates. A national reliability criterion is also set.
Because the sample design is State-based, the sampling ratio differs by State and
depends on the various demographic characteristics of each State. The State
sampling ratios vary from approximately 1 in every 200 to 1 in every 3,000
households in each stratum of the State. The sampling ratio is occasionally
modified slightly to hold the size of the sample relatively constant given the
overall growth of the population (this is called “sample maintenance reduction”).
In determining sample size, a number of factors are taken into account including
population density, average household size, and variance in the unemployment
rate across areas in the State. The preliminary sample size estimate for
households is adjusted by a factor which increases the sample size to account for
the normal sample loss of eligible households that occurs due to household
vacancies, buildings demolished, etc.
The probability design of the CPS is self- weighting, meaning that each housing
unit in a State is given an equal chance of selection. The sampling ratio used
within a sample PSU depends on the probability of selection of the PSU and the
sampling ratio for each State. In a sample PSU with a probability of selection of
1 in 10 and a state sampling ratio of 3,000, a within-PSU sampling ratio of 1 in
300 achieves the desired overall ratio of 1 in 3,000 for the stratum.

August 2010

LAUS Program Manual 2-20

The Sample Rotation Design
The best estimates of month-to- month change would be obtained from 100percent sample overlap, surveying the same households every month. However,
indefinitely surveying a single sample of households would lead to respondent
“fatigue” or “exhaustion” and the increasing the probability of refusals and
respondent errors.
Therefore, part of the sample is changed each month. Each monthly sample is
divided into eight representative subsamples, or rotation groups. A given rotation
group is interviewed for a total of 8 months, divided into two equal periods. It is
in the sample for 4 consecutive months, leaves the sample during the following 8
months, and then returns for another 4 consecutive months. In any one month,
one of the eight rotation groups is in the first month of enumeration, another
rotation group is in the second month, and so on. Under this system, 75 percent
of the sample is common from month to month and 50 percent from year to year
for the same month. (See following chart.) This procedure provides a substantial
amount of month-to- month and year-to-year overlap in the sample, thus providing
better estimates of change and reducing discontinuities in the series of data
without burdening any specific group of households with an unduly long period of
inquiry. However, the overlap creates a correlated error which must be taken into
account in the State-estimation modeling process. (See Chapter 6.)

The rotation plan used for the CPS sample also introduces nonsampling error,
referred to as month- in-sample bias. This bias generally refers to the observed
phenomenon of rotation groups differing in responses, when theoretically they
should have approximately equal measurements. The 4-8-4 rotation pattern adds
an additional dimension to the month- in-sample bias because of factors related to
the large overlap of households from one month to the next and one year to the
next. Samples with large numbers of overlapping units should have a high degree
of consistency with regard to interview responses. However, in repeated CPS
interviews, the later interviews yield consistently higher or lower estimates than
earlier interviews. Unemployment data exhibit the most pronounced month- insample bias, with various subgroups, such as nonwhites and females, exhibiting

August 2010

LAUS Program Manual 2-21

more bias than the general population. Historically, the following national trends
have been fairly regularly demonstrated:
• The unemployment rate drops from month-in-sample one to month-in-sample
two and from month-in-sample five to month-in-sample six.
• There is an overall trend for the rate to decline from month-in-sample one to
month-in-sample eight.
• There is an increase, or surge, in the unemployment rate from month-in-sample
three to month-in-sample four and from month-in-sample seven to month-insample eight.
• There is general agreement that the month-in-sample four and eight “surges”
are attributable to the probing questions on discouraged workers asked in those
months. These questions apparently elicit information that changes previous
negative responses regarding the “looking for work” questions, to positive
responses. Thus, more unemployed persons are identified. This probing was
formerly done in months-in-sample one and five, and significant changes in
reported responses result when the shift was made to months-in-sample four and
eight.
It has also been observed that CPS sample responses vary from one month to the
next. There are a number of theories to explain this phenomenon. One suggests
that for a variety of reasons, sub-groups of potential respondents are successfully
interviewed at different rates. The degree of differential response can change
from one month- in-sample to another. For example, suppose that employed
persons living alone are harder to find than other persons in month- in-sample one
(since, often, no one is at home when the interviewer calls). If arrangements are
made with those contacted in month-in-sample one to retain them in month- insample two, then month- in-sample two could have a better representation of
employed persons living alone, relative to other persons, than month- in-sample
one.
Because month- in-sample bias is believed to exist in the CPS, it is controlled for
in the compositing portion of the estimation process covered later in this chapter.
When the two components of the composite estimate are combined, a month- insample bias adjustment is added to adjust for the relative bias associated with
month- in-sample.

August 2010

LAUS Program Manual 2-22

Data Collection
The housing units which belong to the selected USUs (Ultimate Sampling Units)
are called “designated” households. The list of designated households is a
preliminary list of potential addresses to be sampled. Nationally, there are
approximately 72,000 designated households on this list. This list of designated
household units is then refined by adding households found by reviewing building
permits and sub tracting housing units that have been demolished, converted to
business use, relocated, or are in the sample by mistake (i.e., units are
nonresidential). The result of this refining process is a list of “assigned”
households.
This group of assigned households undergoes further refinement when
interviewers canvas the areas removing vacant housing units, vacant sites for tents
or mobile homes, units occupied by persons with usual residence elsewhere, or
units converted to temporary nonresidential use. These are called “Type B”
noninterviews. An additional noninterview type, “Type C”, occurs when the CPS
collector finds a building demolished, converted to permanent nonresidential use,
or moved from a site.
The remaining households are called “eligible” households. There are
approximately 60,000 eligible households nationally.
CPS data are collected each month during the week containing the 19th day of the
month. Respondents are asked about their labor force activity for the entire
preceding week—the week containing the 12th . A week is defined as Sunday
through Saturday. The data are collected by approximately 1,500 interviewers.
Personal visits are preferred in the first month in which the household is in the
sample. In other months, the interview ge nerally is conducted by telephone.
Approximately 70 percent of the households in any given month are interviewed
by telephone. A portion of the households (10 percent) is interviewed via
computer-assisted telephone interviewing (CATI), from three centralized
telephone centers (located in Hagerstown, MD; Jeffersonville, IN; and Tucson,
AZ) by interviewers who also use a computerized questionnaire.
On the first visit, the interviewer prepares a roster of the household members and
completes a questionnaire for each person 15 years of age and older. The roster is
updated with each visit. The interviewer does not ask directly if the person is
employed, unemployed, or not in the labor force because of potential bias from
the different interpretations these terms might have. Instead, a series of questions
are asked that allow a basic assignment to one of these three categories to be
made. A Computer Assisted Personal Interview (CAPI) is conducted by the
interviewers during each visit. Each interviewer has a laptop computer with a
August 2010

LAUS Program Manual 2-23

computerized version of the CPS questionnaire. When the interviewer has
completed a day’s interviews, the data are transmitted to the Census Bureau’s
central computer in Washington, D.C. Once files are transmitted to the main
computer, they are deleted from the laptops.
Of the 60,000 eligible households, about 7 to 8 percent are not interviewed in a
given month due to temporary absence (vacation, for example) of the occupants,
other failures to make contact after repeated attempts, inability of persons
contacted to respond, unavailability for other reasons, and refusals to cooperate
(about half of the noninterviews). Information is obtained each month for about
110,000 persons 16 years of age or older.

Training and Quality Control
Because of the crucial role interviewers have in the household survey, a great
amount of time and effort is spent maintaining the quality of their work.
Interviewers are given intensive training, including classroom lectures,
discussion, practice, observation, home-study materials, and on-the-job training.
At least once a year, they convene for daylong training and review sessions, and,
also at least once a year, they are accompanied by a supervisor during a full day
of interviewing to determine how well they carry out their assignments.
The data collection technology and the questionnaire provide an opportunity to
build functions to assist and improve data quality into the system itself. For
instance, computer-aided interview technology guides an interviewer through
complex questions: previous answers are used to eliminate further questions that
would elicit extraneous or impossible answers. Built- in range checks for
responses alert interviewers to possible inaccuracies.
Quality control procedures for the CPS are extensive, with more than 20 percent
of the CPS budget spent on training and quality control. The procedures include
extensive data checking and editing of the raw data by Census staff. Using
information from the completed questionnaire as well as additional comments
provided by the interviewer in a “real-time” comments file stored in the computer,
the Census Bureau staff reviews and edits the information obtained for each
person in the sample, and, where possible, identify and correct omissions,
unintelligible entries, and other errors.
Quality control procedures also include monitoring “on line” CATI interviews by
Census Bureau supervisory staff; a system of reinterviews, where a selection of
the sample is interviewed again, and those responses are compared with initial
interview responses; and monthly feedback to the field staff on any errors,
omissions, or inconsistencies detected by the computer edits.

August 2010

LAUS Program Manual 2-24

Estimation Procedures
There are six main steps to the estimation process in the CPS; editing of raw data
and imputation, basic weighting, noninterview adjustment, ratio adjustment,
compositing estimates, and seasonal adjustment. This process takes the raw data
from the CPS interviews, edits it, weights it to represent the population as a
whole, adjusts the data for nonresponse and consistency with independently
derived population counts for demographic sub-groups, combines current
estimates with estimates for the prior month to reduce the variability of the data,
and adjusts for seasonality.
Data Editing and Imputation
Raw CPS data are corrected for inconsistencies or missing items to make them
suitable for use in estimation. This process is completed by Census staff at a
central location in Suitland, MD, and involves two steps: editing of the raw data
and imputing for missing or unacceptable data items. Editing involves identifying
and, where possible, correcting inconsistencies, omissions, illegible entries, and
other errors in the raw data. When the data are received at the national Census
Bureau, they are reviewed for completeness and consistency. Responses to
various survey questions are interpreted and combined to classify respondents as
employed, unemployed, or not in the labor force.
Imputation involves correcting for item nonresponse – the case in which
interviewed persons do not respond to all of the survey questions or their answers
to some questions are deleted during the editing process. The empty data cells are
filled using the “hot deck” method of imputation, which is based on the premise
that persons with similar characteristics provide data that are a good
approximation for the missing responses. In the “hot deck” method, data for all
interviewed persons are cross-classified by age/sex/race and geography. Missing
answers are imputed by using the data from the most recently processed record
for a person in the same age/sex/race/geography group.
Basic Weighting
The basic weighting procedure begins the process of inflating the sample data to
produce an estimate for the entire population. In the basic weighting procedure,
data from each sample person are weighted by the inverse of the probability of the
person being in the sample. This is roughly equal to the number of actual persons
the sample person represents. Thus, adding the basic weights of all sample
persons having a given characteristic yields a simple unbiased estimate of the
number of persons in the population possessing that characteristic.

August 2010

LAUS Program Manual 2-25

When a selected cluster of housing units is found to have many more units than
expected, field subsampling is carried out. Appropriate special weights, reflecting
the subsampling of the cluster, are then applied to the sample data.
Noninterview Adjustment
The weights for all interviewed households are adjusted to account for occupied
sample households for which no information was obtained because of absence,
impassable roads, refusals, or unavailability of the respondents for other reasons
(Type A noninterviews). This noninterview adjustment is made separately for
clusters of similar sample areas that are usually, but not necessarily, contained
within a State. Similarity of sample areas is based on Metropolitan Statistical
Area (MSA) status and size. Within each cluster, there is a further breakdown by
residence. Each MSA cluster is split by “central city” and “balance of the MSA”.
Each non-MSA cluster is split by “urban” and “rural” residence categories. The
proportion of sample households not interviewed averages between 7 and 8
percent, depending upon weather, vacations, etc.
Sample units found vacant, demolished, or converted to nonresidential use (Types
B and C noninterviews) are excluded from those counted for the numerator of this
ratio because such units are out of the scope of the survey. This means that the
weights are not adjusted upwards.
Ratio Adjustment
The distribution of the population selected for the sample may differ somewhat,
by chance, from that of the population as a whole in such characteristics as age,
race, sex, and State of residence. Because these characteristics are closely
correlated with labor force participation and other principal measurements made
from the sample, the survey estimates can be substantially improved when
weighted appropriately by the known distribution of these population
characteristics. This is accomplished through two stages of ratio adjustment, as
follows:
1.) First-stage ratio adjustment - The purpose of the first-stage ratio adjustment is
to reduce the contribution to variance that results from selecting a sample of PSUs
rather than drawing sample households from every PSU in the Nation. This
adjustment is made to the CPS weights in two race cells, black and nonblack; and
two age cells, 0-15 and 16+. It is applied only to PSUs that are non-selfrepresenting in States that have a substantial number of black households. The
first-stage ratio adjustment procedure corrects for differences that existed in each
State cell at the time of a decennial census between a) the race distribution of the
population in sample PSUs and b) the race distribution of all PSUs. Both a) and
b) above exclude self-representing PSUs. This adjustment is not made to housing
units but to the individual household member record.

August 2010

LAUS Program Manual 2-26

The first stage ratio adjustment factors do not depend on response data and remain
the same from month to month during the entire intercensal period. The factors
change when a new sample of PSUs is drawn after a decennial census. The factors
also change if the non-self- representing PSU composition of a State changes for
any other reason.
2.) Second-stage ratio adjustment - The second-stage ratio adjustment procedure
substantially reduces the variance of the estimates and corrects, to some extent,
for CPS undercoverage at the national level. The CPS sample weights are
adjusted to ensure that sample-based estimates of population match national
independent population controls. Each month, independent estimates of various
civilian noninstitutional population distributions at the national level are produced
based on the decennial census and birth and death data from several sources.
Since those characteristics are correlated with labor force status and other items of
interest, weighted CPS sample estimates are forced to agree with the known
distributions of selected population characteristics.
Beginning in 2003, the second-stage ratio adjustment (also known as “raking”)
consists of coverage steps 0A and 0B, followed by three basic iterative steps.
California and New York are split into substate areas (Los Angeles- Long BeachGlendale Metropolitan Division, New York City, and the respective balances of
states). The coverage steps are then applied to these, the remaining 48 States, and
the District of Columbia, primarily to improve the efficiency of adjustments for
subpopulations prone to undercoverage (0A) and to account for variations in
race/gender/age differences between States (0B).
Next, a three step, iterative process is applied to adjust the sample weights of the
estimates.
i.
ii.
iii.

State step—6 gender x age cells defined for 53 States/areas.
Ethnicity step—26 Hispanic and 26 non-Hispanic gender x age
cells.
Race step—34 white-only, 26 black-only, and 26 Asian-only and
residual gender x age cells.

Composite Estimation
The last step in the preparation of most CPS estimates makes use of a composite
estimation procedure. Statistical theory states that the estimate of a quantity can
be improved if two (or more) estimates of that quantity, obtained by different
methods, are combined. This technique is called compositing.
The composite estimate consists of a weighted average of two estimates:
1.) The second-stage ratio estimate based on the entire sample from the
current month.
August 2010

LAUS Program Manual 2-27

2.) A composite estimate for the previous month, adjusted by an estimate
of the month-to-month change based on the six rotation groups common to
both months.
In addition, a bias adjustment term is added to the weighted average to account
for relative bias associated with month-in-sample estimates. This month- insample bias is exhibited by unemployment estimates for persons in their first and
fifth months in the CPS being generally higher than estimates obtained for the
other months.
These composite estimates are then used as controls in the composite weighting
procedure. Both employment and unemployment are controlled in each defined
cell, and not-in- labor force (NILF) is controlled as a residual. This is an iterative
process, similar to that used for second-stage weighting:
i.
ii.
iii.

State step—a single CPS 16+ cell is used for all 53 States/areas.
Ethnicity step—10 Hispanic and 10 non-Hispanic gender x age
cells.
Race step—22 white-only, 14 black-only, and 10 Asian-only and
residual gender x age cells.

The composite estimate results in a reduction in the sampling error beyond that
which is achieved after the two stages of ratio adjustment by taking advantage of
the sample overlap of the survey. For some items, the reduction is substantial.
The resultant gains in reliability are greatest in estimates of month-to- month
change, although gains are also usually obtained for estimates of level in a given
month, change from year to year, and change over other interval of time.
Population Controls
The independent population controls to the CPS are prepared by projecting
forward the resident population from the 2000 Decennial Census. They are
derived by updating demographic census data with information that accounts for
births, deaths, and net migration. Subtracting the estimated numbers of resident
Armed Forces personnel and institutionalized persons reduces the resident
population to the civilian noninstitutional population.
Unlike the LAUS estimates, the monthly national CPS estimates are not revised
on an annual basis. (See Chapter 10 Annual Processing.) However, the monthly
CPS estimates at the State level that are input to the model estimation are revised
each year to reflect the latest population estimates.
Population estimates are released each year by the Census Bureau in a revised
time series for the period following the decennial census. For most years,
revisions to previous years’ estimates tend to be minimal. Revisions can occur for

August 2010

LAUS Program Manual 2-28

two basic reasons: revisions to input data for population estimates, and
methodological changes.
At the beginning of each year, new CPS population controls are introduced for
use in Division, State, and substate model estimation. These controls typically
reflect both new data for the most recent year and revisions to data for earlier
years. After being re-controlled, CPS estimates are then used in model reestimation. Once the model estimates are re-estimated, smoothed, and adjusted to
new the Division control totals, they become the official series.
State CPS labor force levels are adjusted by a simple ratio of the new estimate of
civilian noninstitutional 16 years old and over population to the old estimate of
the same component. Since the same ratio is applied to both employment and
unemployment within a State, there is no impact on percentages, such as
unemployment rates and employment/population ratios. The re-controlled
estimates are generally an improvement over those provided during last year’s
benchmarking.
The availability of the monthly re-controlled data is anno unced in a memorandum
to States at the beginning of each year and is accessible by States via the Extract
module in the STARS application.
Seasonal Adjustment
Over the course of a year, the size of the nation’s labor force and the levels of
employment and unemployment undergo sharp fluctuations due to such seasonal
events as changes in weather, reduced or expanded production, harvests, major
holidays, and the opening and closing of schools. The effect of such seasonal
variation can be very large; seasona l fluctuations may account for as much as 95
percent of the month-to-month changes in unemployment.
Because these seasonal events follow a more or less regular pattern each year,
their influence on statistical trends can be eliminated by adjusting the statistics
from month to month. These adjustments make nonseasonal developments, such
as declines in economic activity or increases in the participation of women in the
labor force, easier to spot.
Seasonal adjustment involves using past data to approximate seasonal patterns.
The seasonally adjusted series therefore have a broader margin for error than the
original data series. They are subject to the same errors as the original series plus
the uncertainties of the seasonal adjustment process. Adjusted series are,
however, useful in analyzing nonseasonal economic and social trends.
Beginning in January 2003, BLS started using the X-12-ARIMA (AutoRegressive Integrated Moving Average) seasonal adjustment program to
seasonally adjust national labor force data from the CPS. This program replaced
August 2010

LAUS Program Manual 2-29

the X-11 ARIMA program which had been used since January 1980. Beginning
in January 2004, BLS converted to the use of concurrent seasonal adjustment to
produce seasonally-adjusted labor force estimates. Concurrent seasonal
adjustment uses all available monthly estimates, including those for the current
month, in developing seasonal factors. Previously, seasonal factors for the CPS
data had been projected twice a year. As a result of this change in methodology,
BLS no longer publishes seasonal factors for the labor force series.
All national labor force and unemployment rate statistics, as well as the major
employment and unemployment estimates, are computed by aggregating
independently adjusted series. For example, for each of the major labor force
components—employment and unemployment—data for four sex-age groups
(men and women under and over 20 years of age) are separately adjusted for
seasonal variation and are then added to derive seasonally-adjusted total figures.
The seasonally-adjusted figure for the labor force is a sum of four seasonallyadjusted civilian employment components and four seasonally-adjusted
unemployment components. The total for unemployment is the sum of the four
unemployment components, and the unemployment rate is derived by dividing the
resultant estimate by the estimate of the labor force. Because of the independent
seasonal adjustment of various series, components will not necessarily add to
totals.
(See Chapter 6 for a discussion of model-based smoothed seasonal adjustment.)

August 2010

LAUS Program Manual 2-30

Publications and Uses of CPS Estimates
INTRODUCTION
Information collected in the Current Population Survey (CPS) is made available
by both the Bureau of Labor Statistics and the Census Bureau through broad
publication programs which include news releases, periodicals, and reports. This
section lists many of the different types of products currently available from the
survey and describes the forms in which they are available. This section also
provides examples of how the data are used for analysis.
BUREAU OF LABOR STATISTICS
Each month, national employment and unemployment data
are published initially in The Employment Situation news
release about 2 weeks after data collection is completed.
The release includes a narrative summary and analysis of
the major employment and unemployment developments
together with tables containing statistics for the principal
data series. The news release is also available electronically
on the Internet and can be accessed at
http://www.bls.gov/news.release/empsit.toc.htm
Subsequently, more detailed statistics are available in the web-only monthly
publication Employment and Earnings Online. The detailed tables provide
information on the labor force, employment, and unemployment by a number of
characteristics, such as age, sex, race, marital status, industry, and occupation.
Estimates of the labor force status and detailed characteristics of selected
population groups not published on a monthly basis, such as Vietnam-era veterans
and Hispanics are published every quarter. Data are also published quarterly on
usual median weekly earnings classified by a variety of characteristics. In
addition, the January issue of Employment and Earnings Online provides annual
averages on employment and earnings by detailed occupational categories, union
affiliation, and employee absences.
About 25,000 of the monthly labor force data series plus quarterly and annual
averages are maintained in LABSTAT, the BLS public database, on the Internet.
They can be accessed from http://www.bls.gov/cps/#data. In most cases, these
data are available from the inception of the series through the current month.
Approximately 250 of the most important estimates from the CPS are presented
monthly and quarterly on a seasonally adjusted basis. The CPS is also used for a
program of special inquiries to obtain detailed information from particular
segments or for particular characteristics of the population and labor force. About
four such special surveys are made each year. The inquiries are repeated annually
in the same month for some topics, including the earnings and total incomes of
individuals and families (published by the Census Bureau); the extent of work
August 2010

LAUS Program Manual 2-31

experience of the population during the calendar year; the marital and family
characteristics of workers; the employment of school-age youth, high school
graduates and dropouts, and recent college graduates; and the educational
attainment of workers. Surveys are also made periodically on subjects such as
contingent workers, job tenure, displaced workers, and disabled veterans.
Generally, the persons who provide information for the monthly CPS questions
also answer the supplemental questions. Occasionally, the kind of information
sought in the special surveys requires the respondent to be the person about whom
the questions are asked. The results of these special surveys are first published as
news releases and subsequently in the Monthly Labor Review or BLS reports.
In addition to the regularly tabulated statistics described above, special data can
be generated through the use of the CPS individual (micro) record files. These
files contain records of the responses to the survey questionnaire for all
individuals in the survey. While the microdata can be used simply to create
additional cross-sectional detail, an important feature of their use is the ability to
match the records of specific individuals at different points in time during their
participation in the survey. (The actual identities of these individuals are
protected on all versions of the files made available to noncensus staff.) By
matching these records, data files can be created which lend themselves to some
limited longitudinal analysis and the investigation of short run labor market
dynamics. An example is the statistics on gross labor force flows, which indicate
how many persons move among the labor force status categories each month.
Microdata files are available for all months since January 1976 and for various
months in prior years. These data are made available on magnetic tape, CDROM, or diskette.
Annual averages from the CPS for the four census regions and nine divisions, the
50 states and the District of Columbia, 50 large metropolitan areas, and 17 central
cities are published annually in Geographic Profile of Employment and
Unemployment. Data are provided on the employed and unemployed by selected
demographic and economic characteristics.
Table 2– 1 provides a summary of the CPS data products available from BLS.

August 2010

LAUS Program Manual 2-32

Table 2-1. Bureau of Labor Statistics Data Products from the CPS
Product
College Enrollment
and Work Activity of
High School
Graduates

Description

Periodicity

News Releases
An analysis of the college enrollment and work activity of the
prior year’s high school graduates by a variety of
characteristics

Source

Annual

October CPS
supplement

Contingent and
Alternative
Employment
Arrangements

An analysis of workers with ‘‘contingent’’ employment arrangements (lasting less than 1 year) and alternative arrangements
including temporary and contract employment by a variety of
characteristics

Biennial

January CPS
supplement

Displaced Workers

An analysis of workers who lost jobs in the prior 3 years due to
plant or business closings, position abolishment, or other
reasons by a variety of characteristics

Biennial

February CPS
supplement

Employment
Situation of VietnamEra

An analysis of the work activity and disability status of persons
who served in the Armed Forces during the Vietnam era

Biennial

September CPS
supplement

An analysis of employee tenure by industry and a variety of
demographic characteristics

Biennial

February CPS
supplement

An analysis of state and regional employment and
unemployment

Annual

CPS annual
averages

Seasonally adjusted and unadjusted data on the Nation’s
employed and unemployed workers by a variety of
characteristics

Monthly

Monthly CPS

Union Membership

An analysis of the union affiliation and earnings of the Nation’s
employed workers by a variety of characteristics

Annual

Monthly CPS;
outgoing rotation groups

Usual Weekly
Earnings of Wage
and Salary Workers

Median usual weekly earnings of full-and part-time w age and
salary workers by a variety of characteristics

Quarterly

Monthly CPS;
outgoing rotation groups

Work Experience of
the Population

An examination of the employment and unemployment
experience of the population during the entire preceding
calendar year by a variety of characteristics

Annual

March CPS
supplement

Annual

March CPS
supplement

Annual

CPS annual
averages

Veterans Job Tenure
of American Workers
State and Regional
Unemployment The
Employment
Situation

A Profile of the
Working Poor
Geographic
Profile of Employment and
Unemployment
Issues in Labor
Statistics

August 2010

Other Publications
An annual report on workers whose families are in poverty by
work experience and various characteristics
An annual publication of employment and unemployment data
for regions, states, and metropolitan areas by a variety of
characteristics
Brief analysis of important and timely labor market issues

Occasional

CPS; other
surveys and
programs

LAUS Program Manual 2-33

Uses of Unpublished Tabulations
Unpublished tabulations include the national and State CPS rotation group data
and the State demographic and economic file commonly know as the DEMECON
tables. These tabulations are used to analyze the movements in the CPS data and
provide a better understanding of the current estimates.
Rotation Groups Analysis. These data are available monthly from the CPS and
represent the raw data obtained from the monthly sample. As discussed earlier in
this chapter, a given rotation group is interviewed for a total of 8 months, divided
into two equal periods. It is in the sample for 4 consecutive months, leaves the
sample during the following 8 months, and then returns for another 4 consecutive
months. (See section 2-15 on Sample Rotation Design.)
Table 2-2 is an example of the monthly rotation group data for the U.S. The new
groups that are being introduced to the sample are the first and fifth groups for the
current month. The groups that are leaving the sample are the fourth and eighth
groups.
Table 2-2. CPS Rotation Group Data
CPS DATA BY ROTATION GROUP
Civilian
noninstitutional
population

Total

Civilian labor force
Unemployed
Employed
Level
Rate

Not in
Labor
force

Sample Counts
Total 16 years and over

3,089

1,675

1,622

53

3.2

786

First month in sample

355

180

171

9

5.0

102

Second month in sample

395

231

228

3

1.3

90

Third month in sample

419

217

210

7

3.2

104

Fourth month in sample

373

191

182

9

4.7

100

Fifth month in sample

420

231

223

8

3.5

93

Sixth month in sample

386

200

193

7

3.5

106

Seventh month in sample

365

207

200

7

3.4

98

Eighth month in sample

376

218

215

3

1.4

93

Incoming rotations

775

411

394

17

4.1

195

Outgoing rotations

749

409

397

12

2.9

193

Rotations common to month before

2,314

1,264

1,228

36

2.8

591

Rotations common to month after

2,340

1,266

1,225

41

3.2

593

Rotations common to year before

1,547

856

831

25

2.9

390

Rotations common to year after

1,542

819

791

28

3.4

396

August 2010

LAUS Program Manual 2-34

Table 2-3 provides an example of the State spreadsheet used in the rotation
analysis. This spreadsheet examines three characteristic of the monthly 4-8-4
sample rotation including how groups entering and leaving the sample affect the
monthly estimates, the labor force status of the groups remaining in the sample,
and the net effect of the entire rotation change.
The comparison of the in-coming groups for the current month to the out-going
groups of the previous months allows the analyst to determine a change in the
employment or unemployment level is due to an economic change in the groups
currently in the sample or if it is caused by the groups coming into or leaving the
sample.
To calculate the in-coming verses the out-going rotation change, add group 1 and
group 5 of the current month and subtract group 4 and group 8 of the previous
month.
In verses out rotation change = (group1 + group5)t – (group4 + group8)t-1

For example, the in-coming verses the out-going rotation change for employment
in Table 2-5 is an increase of 33.8 and is derived from the sum of 311.6 (group 1)
and 399.8 (group 5) for the current month less the sum of 333.9 (group 4) and
343.7 (group 8) of the previous month.
The common rotation change is the sum of the differences between the current
month and the previous month. Each group of the prior month is move up to the
next consecutive group in for the current month. There are no differences for
groups 1 and 5 since these are incoming groups for the current month and did not
exist in the prior month.
Common rotation change = (group2t – group1t-1 ) + (group3t – group2t-1 ) +
(group4t – group3t-1 ) + (group6t – group5t-1 ) + (group7t – group6t-1 ) +
(group8t – group7t-1 )

In the employment example in Table 2-3, the common rotation change is -12.5
and is the sum of 11.4, -0.4, 9.4, -28.2, -4.6 and -0.1.
The net rotation change is the in-coming verses the out- going rotation change less
the common rotation change.
Net rotat ion change = (in-coming – out-going) – common rotation change

In our example the 12.5 over-the- month decrease in the employment levels of the
common rotation groups for the current month is offset by the 33.8 higher
employment levels of the incoming groups resulting in a net change increase of
21.3.

August 2010

LAUS Program Manual 2-35

Table 2-3. CPS Rotation Group Analysis
Employment
Current
Group
Prior Month
Group
Change
Month
311.6
1
1
381.8
393.2
2
11.4
2
367.7
367.3
3
-0.4
3
321.5
330.9
4
9.4
4
333.9
399.8
5
5
368.6
340.4
6
-28.2
6
357.4
352.8
7
-4.6
7
391.7
391.6
8
-0.1
8
343.7
in vs. out rotation change
33.8
common rotation change
-12.5
net rotation change
21.3

Unemployment
Group
1
2
3
4
5
6
7
8

Prior Month
4.3
9.9
19.0
15.7
11.6
12.9
5.9
17.4

Current
Group
Change
Month
18.9
1
7.0
2
2.7
12.6
3
2.7
15.7
4
-3.3
15.3
5
12.1
6
0.5
14.3
7
1.4
5.8
8
-0.1
in vs. out rotation change
1.1
common rotation change
3.9
net rotation change
5.0

Unemployment Rate
Group

August 2010

Prior Month

1
2
3
4
5
6
7

1.1
2.6
5.6
4.5
3.1
3.5
1.5

8

4.8

Current
Month
5.7
1.8
3.3
4.5
3.7
3.4
3.9
1.5

Group
1
2
3
4
5
6
7
8

Change
0.7
0.7
-1.1
0.3
0.4
0.0

LAUS Program Manual 2-36

DEMECON Tables These tables contain monthly and quarterly demographic and
economic data from the CPS for States and regions including the District of
Columbia, New York City and the Los Angeles - Long Beach metropolitan area.
The following tables appear in the DEMECON file:
a. Employment status of the civilian non- institutional population by sex, age,
race, and Hispanic origin
b. Civilians not in the labor force by sex and age
c. Unemployed persons by sex, age, race, Hispanic origin, and reason for
unemployment
d. Unemployed persons by sex, age, race, Hispanic origin, and duration of
unemployment
e. Full- and part-time status of the civilian non- institutional population by
sex, age, race, Hispanic origin
Since the data are available by age, race and sex, they are useful to gain insight on
demographic groups that may be experiencing changes contributing to the monthto-month variations in the CPS estimates.
However, the data contained in DEMECON file are unpublished and generally do
not meet the BLS publication standards for accuracy and reliability. Tables B-2
through B-5 in the Geographic Profiles bulletin provide generalized sampling
error information for CPS annual average data. To obtain approximate error
measures for the monthly estimates in this package, double the sampling errors in
those tables.

Caution should be used in drawing inferences from these data and they should not
be released unless they are specifically requested. When providing data to users,
a statement similar to the following should be used: The data provided are
unofficial, unpublished, data from the Bureau of Labor Statistics and do not meet
BLS publication standards for accuracy and reliability. If you publish or cite
these data please refer to them as such.

August 2010

LAUS Program Manual 2-37

U.S. CENSUS BUREAU
The U.S. Census Bureau has been analyzing data from the
Current Population Survey and reporting the results to the
public for over five decades. The reports provide information
on a recurring basis about a wide variety of social,
demographic, and economic topics. In addition, special
reports on many subjects have also been produced. Most of
these reports have appeared in 1 of 3 series issued by the
Census Bureau: P-20, Population Characteristics; P-23, Special Studies; and P-60,
Consumer Income. Many of the reports are based on data collected as part of the
March demographic supplement to the CPS. However, other reports use data
from supplements collected in other months (as noted in the listing below). A full
inventory of these reports as well as other related products is documented in:
Subject Index to Current Population Reports and Other Population Report Series,
CPR P23-192, which is available from the Government Printing Office, or the
Census Bureau. Most reports have been issued in paper form; more recently,
some have been made available on the Internet at http://www.census.gov.
Generally, reports are announced by press release, and are released to the public
via the Census Bureau Public Information Office.
Census Bureau Report Series
P-20, Population Characteristics. Regularly recurring reports in this series
include topics such as geographic mobility, educational attainment, school
enrollment (October supplement), marital status, households and families,
Hispanic origin, the Black population, fertility (June supplement), voter
registration and participation (November supplement), and the foreign-born
population.
P-23, Special Studies. Information pertaining to special topics, including one-time
data collections, as well as research on methods and concepts are produced in this
series. Examples of topics include computer ownership and usage, child support
and alimony, ancestry, language, and marriage and divorce trends.
P-60, Consumer Income. Regularly recurring reports in this series include
information concerning families, individuals, and households at various income
and poverty levels, shown by a variety of demographic characteristics. Other
reports focus on health insurance coverage and other noncash benefits.
In addition to the population data routinely reported from the CPS, Housing
Vacancy Survey (HVS) data are collected from a sample of vacant housing units
in the Current Population Survey (CPS) sample. Using these data, quarterly and
annual statistics are produced on rental vacancy rates and home ownership rates
for the United States, the four census regions, location inside and outside
metropolitan areas (MAs), the 50 states and the District of Columbia, and the 75
August 2010

LAUS Program Manual 2-38

largest metropolitan areas. Information is also made available on national home
ownership rates by age of householder, family type, race, and Hispanic origin. A
press release is issued each quarter as well as quarterly and annual data tables on
the Internet.
Supplement Data Files
Public use microdata files containing supplement data are available from the
Census Bureau. These files contain the full battery of basic labor force and
demographic data along with the supplement data. A standard documentation
package containing a record layout, source and accuracy statement, and other
relevant information is included with each file. (The actual identities of the
individuals surveyed are protected on all versions of the files made available to
noncensus staff.) These files can be purchased through the Customer Services
Branch of the Census Bureau and are available in either tape or CD-ROM format.
The CPS homepage is the other source for obtaining these files. The CPS
homepage can be accessed at http://www.bls.census.gov/cps/cpsmain.htm.

August 2010

LAUS Program Manual 2-39

3

Inputs to LAUS Estimation:
The Unemployment Insurance
System

T

he Federal-State Unemployment Insurance Program provides
unemployment benefits to eligible workers who are unemployed through
no fault of their own (as determined under State law), and meet other
eligibility requirements of State law. Unemployment insurance payments
(benefits) are intended to provide temporary financial assistance to unemployed
workers who meet the requirements of State law.
Each State administers a separate unemployment insurance program within
guidelines established by Federal law. The State law under which unemployment
insurance claims are established determines eligibility for unemployment
insurance, benefit amounts and the length of time benefits. In the majority of
States, benefit funding is based solely on a tax imposed on employers.
The Federal-State Unemployment Insurance (UI) system was established in 1935
as part of the Social Security Act. It was intended by its founders to serve both as
a counter-cyclical economic stabilizer for the economy and as a central part of the
nation’s economic security system for workers with a strong attachment to the
labor force who are temporarily laid off or permanently lose their jobs. The
program is funded through mandatory payroll taxes paid by employers.
Statistics from the UI systems are the only current measure of unemployment at
the substate level available at the county (and in some States, city) level. They
are a key input to the unemployment models used to estimate unemployment for
the 50 States, the District of Columbia, Los Angeles-Long Beach, the balance of
California, New York City and the balance of New York. Claims data from the
UI systems are inputs to the Handbook method for estimating labor market area
(LMA) unemployment and their use in the claims-based unemployment
disaggregation yields more accurate sub-LMA estimates than are obtained from
decennial census-based approaches.
While these statistics are biased for estimating total unemployment between
States in so far as they reflect the particular State's UI law, the statistics have the
advantage of being current and, with proper coding and tabulation, are consistent
among areas within States.
August 2010

LAUS Program Manual 3-1

Federal Role
The UI system is based on a dual program of federal and State statutes. Much of
the federal program is implemented through the Federal Unemployment Tax Act
(FUTA). Each State administers a separate UI program with guidelines
established by Federal statute. The States also determine the eligibility, the
benefit amount and the length of time that benefits are paid.
A combination of federal and State taxes is levied on employers
to support the UI program. The proceeds from the
unemployment taxes are deposited into the Unemployment
Trust Fund. Each State has a separate account in the Fund to
which deposits are made.
Federal UI Programs
In addition to the regular State UI program which cover the bulk of nonfarm
workers, separate Federal UI programs exist for specific types of workers.
Railroad workers receive unemployment insurance benefits through the Railroad
Retirement Board (RRB). Federal employees are covered through
Unemployment Compensation for Federal Employees (UCFE). Former military
personnel are covered through the Unemployment Compensation for ExServicemen (UCX).
Special benefits programs exist for specific situations as well. During periods of
high unemployment, programs are activated that extend the time period that
individuals can receive benefits. Workers who lose their jobs as a result of the
nation's trade policies may receive special benefits after they exhaust their regular
UI benefit through the Trade Adjustment Assistance program. Workers who lose
their jobs due to a natural disaster may qualify for benefits through the Disaster
Unemployment Assistance program.
Railroad Retirement Board (RRB)
The Railroad Unemployment Insurance Act provides two
kinds of benefits for railroad employees: unemployment
benefits and sickness benefits. Benefit payments are
based on biweekly claims filed with the Railroad
Retirement Board, the Federal agency responsible for
administering the Railroad Unemployment Insurance Act.
The funds to pay unemployment and sickness benefits are provided by payroll
taxes on railroad employers only. Railroad employees do not pay unemployment
insurance taxes.
Claims filed through the Railroad Unemployment Insurance Act for
unemployment during the reference period including the 12th of the month are
used in the calculation of monthly LAUS substate estimates.

August 2010

LAUS Program Manual 3-2

Unemployment Compensation for Federal Employees (UCFE)
Unemployment Compensation for Federal Employees is the benefit program for
unemployed federal employees. Funding comes from the Federal Government
and is distributed through State agencies. Federal wages are not reported to a
State unemployment compensation agency until a claim is filed. The claimant's
federal wages will be "assigned" to the State of the last duty station or the State of
residency if the duty station was outside the U.S or if covered work was done in
the State after leaving federal service. This is also the case if the employer was
the Federal Emergency Management Agency (FEMA), since this is the only
Federal agency that does not report wages to the last duty station.
UCFE claims filed for unemployment in the reference period including the 12th of
the month are used in the calculation of monthly Statewide and substate LAUS
estimates.
Unemployment Compensation for Ex-Service members (UCX)
Unemployment Compensation for Ex-Service members is the
benefit program for ex-military personnel to provide weekly
income to meet basic needs while searching for employment.
Those who were on active duty with a branch of the U.S.
military or active duty in reserve status as a member of a
National Guard or Reserve component continuously for 90 or
more days may be entitled to unemployment benefits based on that service. The
military wages are assigned to the State where they first file a new claim after the
separation from active duty.
UCX benefits are paid under the same conditions as benefits based on other
employment. However, since LAUS measures the civilian labor force, UCX
claims data are not used in the calculation of LAUS estimates.
Extended Benefits (EB) and Special Temporary Programs
Two types of UI programs grant additional weeks of unemployment
compensation to individuals who have depleted their UI benefits under the regular
UI program during economic downturns. One is the permanent program that is
funded by both the State and the Federal governments, known as State Extended
Benefits (EB), and the others are temporary programs that are financed by the
Federal government and enacted for periods of higher unemployment.
In addition, a number of States have solely State-financed programs for extending
the potential duration of benefits during periods of high unemployment, for
claimants in approved training who exhaust benefits, or for a variety of other
reasons.
Federal temporary extended benefits programs have been used periodically during
economic downturns since the late 1950s. The first temporary program,
Temporary Unemployment Compensation (TUC), was effective for the period of

August 2010

LAUS Program Manual 3-3

June 1958 to June 1959. Since then, temporary extended assistance programs
have occurred seven times under different titles. These include
•
•
•
•
•
•
•

Temporary Extended Unemployment Compensation (TEUC), April 1961 to
June 1962;
Temporary Compensation (TC), January 1972 to March 1973;
Federal Supplemental Benefits (FSB), January 1975 to January 1978;
Federal Supplemental Compensation (FSC), September 1982 to June 1985;
Emergency Unemployment Compensation (EUC) November 1991 to April
1994;
Temporary Extended Unemployment Compensation (TEUC); March 2002 to
March 2004; and
Emergency Unemployment Compensation (EUC08), July 2008 to January 3,
2012.
All Special Extended Benefit Programs

Name

Effective Dates

Weeks of Benefits

Temporary Unemployment
Compensation (TUC)
Temporary Extended
Unemployment Compensation
(TEUC)
Temporary Compensation(TC)
Federal Supplemental Benefits
(FSB)
Federal Supplemental
Compensation (FSC)
Emergency Unemployment
Compensation (EUC)
Temporary Extended
Unemployment Compensation
(TEUC)
Emergency Unemployment
Compensation (EUC 08)

06/58 – 06/59

Up to 13

04/61 – 06/62

Up to 13

01/72 –03/73
01/75 –01/78

Up to 13
Up to 13 or 26

09/82 –06/85

Up to 8, 10, 12, or 14

11/91 –04/94

Up to 7, 13, 20, 26 or 33

03/02 –03/04

Up to 13 or 26

07/08 –1/12

Up to 20 , 34, 47, or 53

Source: Employment and Training Administration, Office of Workforce Security

State Extended Benefits (EB): The EB program has been a permanent part of the
Federal-State UI system since 1970 and is available to workers who have
exhausted regular unemployment insurance benefits during periods of high
unemployment as defined by the trigger “on” criteria below.
The basic EB program provides up to 13 additional weeks of benefits when a
State is experiencing high unemployment. A number of States have also enacted
voluntary programs to pay up to 7 additional weeks (20 weeks maximum) of
extended benefits during periods of extremely high unemployment which is
defined as a State 3-month average smoothed seasonally adjusted unemployment
rate greater than 8 percent.
A State may trigger "on" for extended benefits for a week if certain criteria are
met based on the State’s insured unemployment rate (the ratio of individuals

February 2011

LAUS Program Manual 3-4

collecting benefits to UI covered employment) or alternatively its total
unemployment rate (the LAUS smoothed seasonally adjusted rate). There are
three separate sets of criteria that can be used to determine if the EB program is
activated in a State. One is mandatory and the other two are optional and enacted
by State law.
For the mandatory trigger, a State must pay up to 13 weeks of EB if the insured
unemployment rate (IUR) for the previous 13 weeks is at least 5 percent and is
120 percent of the average of the rates for the corresponding 13-week period in
each of the 2 previous years. This comparison is called a “look-back.” Prior to
the December 2007-June 2009 recession, ten States had only the mandatory
trigger in their State laws. These included Delaware, Florida, Georgia, Iowa,
Kentucky, Massachusetts, North Dakota, South Dakota, Utah, and Wyoming.
Since then five States remain that have only the mandatory trigger in their State
laws; these are Iowa, North Dakota, South Dakota, Utah, and Wyoming.
State law may allow for the use of one of the following alternative triggers
instead. However, if a State does not use an alternative trigger, the mandatory
method must be used to determine trigger status.
One of the optional triggers allows a State to pay up to 13 weeks of EB if the IUR
for the previous 13 weeks is at least 6 percent, regardless of the experience in the
previous years. The majority of States have specified the use of this trigger.
The other optional trigger allows a State to pay up to 13 weeks of EB if the
average total unemployment rate (TUR), smoothed seasonally adjusted, for the
most recent 3 months is at least 6.5 percent and is 110 percent of the rate for the
corresponding 3-month period in either of the 2 previous years. If such rate is at
least 8.0 percent and is 110 percent of the rate for the corresponding 3-month
period in either of the 2 previous years, the duration increases from 13 to 20
weeks. Prior to the December 2007-June 2009 recession, eleven States including
Alaska, Connecticut, Kansas, New Hampshire, New Jersey, New Mexico, North
Carolina, Oregon, Rhode Island, Vermont, and Washington have permanent laws
enacting the TUR options.
Summary of State Extended Benefits Trigger Options Prior to the
December 2007-June 2009 Recession
Trigger Option

Number
of States

States*

Potential
Number of
Weeks
13

5% IUR
with look back

52

All States

6% IUR
without look
back

40

AL, AK, AZ, AR, CA, CO, CT, DC, HI,
ID, IL, IN, KS, LA, ME, MD, MI, MN,
MS, MO, MT, NE, NV, NJ, NM, NY,
NC, OH, OK, OR, PA, PR, RI, SC, TN,
TX, VT, WA, WV, WI

13

TUR
(6.5% and 8%)

11

AK, CT, KS, NH, NJ, NM, NC, OR, RI,
VT, WA

20

* States include DC and PR.

Source: Employment and Training Administration, Office of Workforce Security

February 2011

LAUS Program Manual 3-5

A State triggers “off” extended benefits if, for the period consisting of the
reference week and the immediately preceding twelve weeks, the requirements of
the selected method are not satisfied.
Temporary Changes: The Assistance for Unemployed Workers and Struggling
Families Act (Public Law No. 111-5) enacted February 17, 2009, encouraged States
experiencing high unemployment to invoke the program’s optional total
unemployment rate (TUR) trigger by providing that 100 percent of the benefit costs
of EB would be federally funded for a specified period. This resulted in 27 States
enacting laws allowing for the use of the optional TUR trigger. Among these States
are Alabama, Arizona, California, Colorado, District of Columbia, Delaware, Florida,
Georgia, Idaho, Illinois, Indiana, Kentucky, Massachusetts, Maine, Michigan,
Minnesota, Missouri, New York, Nevada, Ohio, Pennsylvania, South Carolina,
Tennessee, Texas, Virginia, Wisconsin, and West Virginia.
The federal funding also applied to States triggering “on” under other EB triggers and
was available to States that already have the TUR trigger in their laws. It covered the
costs of EB for weeks beginning after February 17, 2009, and before January 1, 2010
and continued to be 100 percent federally funded for the benefit costs of weeks of
unemployment ending before June 1, 2010.
The Tax Relief, Unemployment Insurance Reauthorization, and Job Creation Act of
2010 (Public Law 111-312), enacted December 17, 2010, temporarily extended the
100 percent federal funding for EB. States are reimbursed for 100 percent of the
benefit costs of EB for weeks of unemployment beginning after February 17, 2009,
and before January 4, 2012. The EB will continue to be 100 percent federally funded
for the benefit costs of weeks of unemployment ending before June 11, 2012.
Public Law 111-312 also permits States to amend their EB laws to temporarily
modify the provisions concerning “on” and “off” indicators based on the insured
unemployment rate and the total unemployment rate. Specifically, it permits states to
make determinations of whether there is an “on” or “off” indicator by comparing
current unemployment rates to the unemployment rates for the corresponding period
in the three preceding years. (Under permanent EB law, the look-back is to
unemployment rates during the last two years.) This modification to the look-back
provisions will enable many states to remain “on” EB much longer. The new EB
indicator provisions are effective with respect to compensation for weeks of
unemployment beginning after the date of enactment of the Act (or, if later, the date
established pursuant to State law) and ending on or before December 31, 2011.
Once the 100 percent Federal funding was no longer available States that temporarily
enacted the TUR most likely reverted to their original UI laws.

February 2011

LAUS Program Manual 3-6

State Extended Benefits Trigger Options Summary as a Result of the
Assistance for Unemployed Workers and Struggling Families Act and
The Tax Relief, Unemployment Insurance Reauthorization, and Job
Creation Act of 2010 (as of February 20, 2011)
Trigger Option

Number of
States

States*

5% IUR
with look back

52

All States

6% IUR
without look
back

40

TUR
(6.5% and 8%)

38

AL, AK, AZ, AR, CA, CO, CT, DC, HI,
ID, IL, IN, KS, LA, ME, MD, MA, MN,
MS, MO, MT, NE, NV, NJ, NM, NY,
NC, OH, OK, OR, PA, PR, RI, SC, TN,
TX, VT, VA, WV, WI
AK, AL, AZ, CA, CO, CT, DC, DE, FL,
GA, ID, IL, IN, KS, KY, MA, ME, MI,
MN, MO, NC, NH, NJ, NM, NV, NY,
OH, OR, PA, RI, SC, TN, TX, VA, VT,
WA, WI, WV

Potential
Number of
Weeks
13

13

20

* States include DC and PR.

Source: Employment and Training Administration, Office of Workforce Security

Emergency Unemployment Compensation 2008 (EUC08): The EUC08 program
is the latest temporary federal program providing extended benefits. It was
created on June 30, 2008, by the Supplemental Appropriations Act of 2008 (P.L.
110-252). It made up to 13 additional weeks of benefits available to unemployed
individuals who had already collected all regular State benefits for which they
were eligible and who met the eligibility requirements.
EUC08 was payable to individuals who (1) had exhausted all rights to regular
compensation with respect to a benefit year that ended on or after May 1, 2007;
(2) had no rights to regular compensation or extended benefits (EB); and (3) were
not receiving compensation under the unemployment compensation law of
Canada. To qualify for EUC08, individuals must have had employment of 20
weeks of work, or the equivalent in wages, in their base periods. Continuing
eligibility was determined under the requirements of the State law.
On November 21, 2008, the President signed the Unemployment Compensation
Extension Act of 2008. This bill expanded Emergency Unemployment
Compensation (EUC08) to 20 weeks nationwide, and created a second tier of
EUC08 for individuals in States with high unemployment rates.
The second tier of benefits was available for States with a three-month smoothed
seasonally adjusted average unemployment rate of at least 6 percent and provided
up to an additional 14 weeks of EUC08 benefits.
A third tier, which provided 13 more weeks of benefits to individuals who had
exhausted their second tier benefits, was added for States experiencing a 3-month
seasonally-adjusted TUR greater than 6 percent or a 13-week average IUR greater
than 4 percent.A fourth tier provided another 6 weeks of benefits for individuals
who had exhausted their third tier benefits in States that have a total

February 2011

LAUS Program Manual 3-7

unemployment rate of at least 8.5 percent or a 13-week average IUR greater than
6 percent.
UI Program
Regular UI
EB

Maximum
Weeks of
Benefits
26
13

Additional EB*
(potentially 14 States)
EUC08 1st tier
EUC08 2nd tier

20
14

EUC08 3rd tier

13

EUC08 4th tier

6

Total

7

Trigger
(Some States may allow more weeks, see below.)
IUR ≥ 5%, and IUR ≥ 6%, or 3 mo average State
smoothed seasonally adjusted TUR ≥ 6.5%
3 mo average State smoothed seasonally
adjusted TUR ≥ 8%, enacted by State law
Exhausted all regular UI benefits
Exhausted all regular EUC 1st tier benefits with
3-month average TUR ≥ 6%
Exhausted all regular EUC 2nd tier benefits with
3-month average TUR ≥ 6% or 13 week average
IUR ≥ 4%
Exhausted all regular EUC 3rd tier benefits in States
with TUR ≥ 8.5% or 13 week average IUR ≥ 6%

92 99*

EB and Special Temporary Programs and LAUS Estimation: Claim counts from
the EB and special temporary programs are not used in LAUS estimation. The
LAUS substate methodology already includes an estimate of unemployed
exhaustees — those persons who have exhausted the total benefit award under
State unemployment insurance within their benefit year and are still jobless. This
component is developed by using area-specific counts of individuals who received
final payments under State UI and survival rates based on the duration of jobless
spells in the Current Population Survey. (See Chapter 7.)
Claims data from permanent and temporary extensions of unemployment
compensation have not been used directly in LAUS estimation primarily because
of their temporary nature. Although individuals exhaust benefits continuously
throughout the year, insured claims under permanent or temporary extended
benefit programs for these people will be taken only when an extended UI benefit
period is established, either by trigger or by special legislation. The
administrative and methodological steps needed to incorporate such extended
benefit program data into LAUS estimation is disproportionately complex, given
the trigger on/trigger off nature of the program and the intricacies of the
temporary program. Likewise, data from the regular EB and the four levels of
EUC08 are not to be incorporated into LAUS estimation. LAUS uses the estimate
of unemployed exhaustees to represent those persons who claimed their total
benefit award within the benefit year and are still jobless.

February 2011

LAUS Program Manual 3-8

Trade Readjustment Allowances (TRA)
Trade Readjustment Allowances are benefits to persons whose jobs were affected
by foreign imports. Benefits are provided through the Federal Trade and the
North American Free Trade Agreement.
The Federal Trade Act provides special benefits under the Trade Adjustment
Assistance (TAA) program to those who were laid off or had hours reduced
because their employer was adversely affected by increased imports from other
countries.
The North American Free Trade Agreement (NAFTA) provides special benefits
under the NAFTA Transitional Adjustment Assistance (NAFTA-TAA) program.
Individuals who were laid off, or had hours reduced because their employer was
adversely affected by increased imports from Mexico or Canada, or because their
employer shifted production to either of these countries may qualify for benefits.
These benefits include paid training for a new job and financial help in searching
for work in other areas or relocation to an area where jobs are more plentiful.
Individuals can qualify for these special benefits only after their regular
unemployment compensation is exhausted. Thus claims data from the TAA and
the NAFTA-TAA programs are not directly counted in LAUS estimates.
Individuals who have exhausted their benefits in the regular UI program are
already accounted for through final payment counts. See UI Claims Data for
LAUS Estimation later in this chapter or Labor Market Area Unemployment in
Chapter 7 for more information on final payments.
Disaster Unemployment Assistance (DUA)
Section 410 of the Robert T. Stafford Disaster Relief and Emergency Assistance
Amendments of 1988 created a program for the payment of unemployment
assistance to individuals whose unemployment is the direct result of a major
disaster as declared by the President of the United States.
Suffering a monetary loss due to damage of property or crops does not
automatically entitle claimants to DUA. Individuals may qualify for DUA if: they
worked in or were scheduled to begin work in a county declared as a federal
disaster county and they cannot work as a direct result of a disaster. In addition,
the work that cannot be performed must be their primary source of income and
livelihood; and they must not qualify for regular unemployment insurance from
any State.
This includes workers who suffer a loss or interruption of work as a direct result
of a major disaster, and, self-employed individuals, including farmers and day
care providers who lost or suffered a substantial reduction or interruption of selfemployment activities as a direct result of a major disaster.
Although individuals may receive DUA benefits, they are not counted as
unemployed. The definition of employed persons under the CPS includes all
those who were not working but who had jobs or businesses from which they
August 2010

LAUS Program Manual 3-9

were temporarily absent because of bad weather, whether they were paid for the
time off or were seeking other jobs.
Birth and Adoption Unemployment Compensation (BAA–UC)
On August 14, 2000, the Department of Labor published a ruling that
allowed States to implement an experiment to provide Unemployment
Compensation to employees on approved leave following the birth or
adoption of a child. The experiment was to test whether partial wage
replacement would strengthen new parents’ connection to the
workforce, as some studies indicated. State participation was voluntary
and States had wide latitude in developing their experiments.
On November 10, 2003, the Department of Labor repealed the Birth and Adoption
Unemployment Compensation (BAA–UC) regulations for the following reasons.
The UI system is designed to provide temporary wage insurance for individuals
who are unemployed due to lack of suitable work. This would generally not be
the case for parents who would avail themselves of BAA–UC. Such parents
would be out of work because they both initiated their separation from the
workforce and are currently unavailable for work; they would have effectively
withdrawn from the labor market for a period of time. As a result, BAA–UC paid
to these individuals would be a payment for voluntarily taking time off work
rather than payment due to lack of suitable work. As such, it would be paid leave,
which was not envisioned in the design of the UC program.
Although no State had enacted BAA–UC legislation, individuals who would have
received benefits under this experimental program would not have met the CPS
criteria for unemployed and therefore would not have been counted in the
monthly LAUS estimates.
Self-Employment Assistance (SEA)
Self-Employment Assistance offers dislocated workers the opportunity for early
re-employment. The program is designed to encourage and enable unemployed
workers to create their own jobs by starting their own small businesses. Under
the program, States can pay a self-employed allowance, instead of regular
unemployment insurance benefits, to help unemployed workers while they are
establishing businesses and becoming self-employed. Participants receive weekly
allowances while they are getting their businesses started. To participate in the
program an individual must be eligible for unemployment compensation, have
been permanently laid off from his/her previous job and identified through the
State’s UI profiling system as likely to exhaust his/her benefits, and must
participate in self-employment activities including entrepreneurial training and
business counseling.
The following State have provisions for SEA: California, Delaware, Louisiana
(law in place but no active program), Maine, Maryland, New Jersey, New York
(expired 12/7/2011), Oregon, Pennsylvania, and Washington (expires 7/1/2012).

August 2010

LAUS Program Manual 3-10

Since individuals participating in a State SEA program are exempt from the State
laws relating to availability for work, search for work, and refusal to accept work,
SEA claims are not included in the claims counts for LAUS estimation.
Short-Time Compensation (STC)
The STC program, commonly known as work-sharing, provides partial UI
benefits to individuals whose work hours are reduced from full-time to part-time
on the same job. STC allows an employer, faced with potential layoffs because of
reduced workload, to reduce the number of regularly scheduled hours of work for
all employees rather than incur layoffs. Benefits are payable to workers for the
hours of work lost, as a proportion of the benefit amount for a full week of
unemployment.
The STC program currently has eighteen States participating: Arizona, Arkansas,
California, Connecticut, Florida, Iowa, Kansas, Louisiana, Maryland,
Massachusetts, Minnesota, Missouri, New York, Oregon, Rhode Island, Texas,
Vermont and Washington.
Workers are not obligated to meet the State’s regular UI requirements of
availability for work, actively seeking work, or refusal to accept work, but must
be available for their normal workweek. Thus, any claim records associated with
STC are not included in the LAUS estimation.
Office of Workforce Security Responsibilities
The Office of Workforce Security (OWS) of the Employment and Training
Administration (ETA) oversees the UI system and works closely with State
employment security agencies. The OWS is responsible for Program
Development and Implementation, Performance Review, Legislation, Policy and
Research, Fiscal and Actuarial services, and Information Technology.
Each Thursday, the OWS releases the “Unemployment Insurance Weekly Claims
Report” at 8:30 AM. This release can be found on the Department of Labor’s
website at www.doleta.gov.
The seasonally adjusted weekly initial claims series is a leading economic
indicator. They are the most current data on the number of people filing for
unemployment insurance. The seasonally adjusted continued claims data provide
insight as to the duration of the claims initially filed.
Interstate Statistical Data Exchange
The OWS is responsible for administrating and maintaining the Interstate
Statistical Data Exchange, which operates on the Interstate Connection Network
(ICON). The ICON is a hub-oriented data communication network that enables
54 jurisdictions, including Canada, to exchange interstate wage and benefit
transactions through batch applications. The ICON hub is located in Orlando,
Florida and maintained by Affiliated Computer Services.
The reporting of initial claims and continued claims (referred to as weeks claimed
by ETA) information by the Liable State to the Agent/Residence State is not only
August 2010

LAUS Program Manual 3-11

vital for the efficient payment of interstate claims benefits, but it also has the
following important uses:
1) Interstate agent weeks claimed are needed for accurate counts of total
continued claims without earnings for LAUS estimation.
2) Interstate initial claim and weeks claimed information identifies interstate
claimants to the Agent/Residence State for purposes of providing reemployment assistance.
3) Interstate agent weeks claimed information is necessary to the
Agent/Residence State’s calculation of its insured unemployment rate that is
the trigger mechanism for the State’s Extended Benefit Program.
4) Interstate agent initial claims and weeks claimed are inputs to major economic
indicators which describe emerging and continuing unemployment conditions
in the Agent State.
The calculation of the State’s total unemployment rate (LAUS estimate), which is
the alternate trigger for the extended benefit program, includes the number of
residents that regularly commute across the State line to work in another State and
are unemployed and filing for benefits against another State. For this reason, the
reporting of commuter weeks claimed, for the survey week, is included in the data
reporting requirement of the Liable State to the Agent/Residence State.
Liable/Agent Data Transfer
The LADT record format was developed by the National Association of State
Workforce Agencies’ Interstate Benefit Committee in consultation with the
Unemployment Insurance (UI) Committee, the Labor Market Information
Committee and the U.S. Department of Labor’s Bureau of Labor Statistics and the
Employment and Training Administration.
The Liable/Agent Data Transfer (LADT) (Appendix/Table 3-1) is a batch
application that has a multi-purpose record format.
There are three types of records identified by the value in the Record Type field
(position 472 – Field number 62) on the LADT record layout:
1) Telephone Initial Claim (TIC);
2) Weeks Claimed (WC) (A “Weeks Claimed" record can be either an interstate
continued claim or a commuter claim);
3) Reopen/Transfer of Claim (Reopen/Transfer).
The origin of the record is identified by the value in the Liable State FIPS field
(Field number 18 - positions 184 – 185 on LADT record layout) and will be either
the alphabetic UI postal abbreviation or the numeric FIPS code.
The destination is determined by the value in the Agent State FIPS field (Field
number 20 - positions 190 – 191 on LADT record layout) and will be either the
alphabetic US postal abbreviation or the numeric FIPS code.

August 2010

LAUS Program Manual 3-12

A commuter claim is identified in the Commuter Identification Code field (Field
number 58 - position 412 on LADT record layout). An “X” in this field indicates
a claim filed by a commuter from the Residence State, while a blank space
indicates that it is not a commuter claim.
Commuter Claim data are included in the transmittal due the first Monday of each
month. The weekending dates on commuter weeks claimed records must be
xx/12/xx through xx/18/xx only. Each month’s commuter data report must
include detail data for the “current commuter reporting month” and “prior
commuter reporting month.” “Current commuter reporting month” is the most
recently completed month. “Prior commuter reporting month” is the month
proceeding the most recently completed month.
When each Liable State’s file is received at the Hub, the records are edited and
stored. As confirmation of receipt and processing of the file, the Liable State
immediately receives, or can request output of, three reports: 1) the Liable
Summary Report; 2) the LADT Error Report; and 3) the LADT Edit Counts
Report.
After close of business on Tuesday, the Hub database is updated with each Liable
State’s file. On Wednesday morning of each week, LADT data are distributed to
the destination Agent State(s). There are two reports distributed weekly and one
monthly:
1) Agent State Summary Report – Interstate Claims (Appendix/Table 3-2) 2) Agent State Detail Data Report (contains micro data of all records)
3) On the first Wednesday of each month, a third report, the Agent/Residence
Commuter Weeks Claimed Report is also distributed.
(See Appendix/Table 3-3)

August 2010

LAUS Program Manual 3-13

Weekly LADT Schedule
The Liable State’s detail data report is due at the Hub on
Monday of each week.

Monday
Commuter weeks claimed data must be included in the
transmission due the first Monday of the month only.
Hub processing of all data received takes place after the
close of business on Tuesday.

Tuesday

Wednesday

Hub processing of all data received takes place after the
close of business on Tuesday.
The Hub distributes Agent State detail data no later than
11:00 a.m. on Wednesday.

Additional information on the ETA, the OWS, and UI claims data can be found on the
Internet at www.doleta.gov.
Information Technology Support Center (ITSC)
In 1994, the ETA and the Maryland Department of Labor, Licensing, and Regulation
established the Information Technology Support Center (ITSC). The ITSC is a
collaboration of State employment security agencies, the Department of Labor (DOL),
and private sector partners. It promotes the appropriate application of information
technology and assists in providing States with more accurate, efficient, cost effective,
and timely service for unemployment insurance recipients.
Additional information on ITSC, UI programs, and statistics can be found on the Internet
at www.itsc.state.md.us.

August 2010

LAUS Program Manual 3-14

State UI Programs
Each State law, subject to federal requirements, establishes guidelines
determining employer coverage, individual employee eligibility, the amount and
duration of benefits paid for claims, and disqualification provisions. State UI
laws also determine the amount of payroll taxes used to fund regular UI benefits
that employers must pay. A summary of changes in individual State UI laws can
be found in each January issue of the Monthly Labor Review, published by BLS.
UI Coverage
Each State has determined its own laws regarding UI coverage, but they have
been greatly influenced by the federal government. The Federal Unemployment
Tax Act (FUTA) provides tax incentives that have ensured States conformity with
the minimum coverage standards set down in FUTA.
In general, a covered employer is defined under the FUTA as one who has a
quarterly payroll of $1500 in the calendar year or preceding calendar year, or one
worker in 20 weeks. While many States have chosen to expand coverage beyond
the FUTA standards, the notable exceptions and limitations are noted below.
Twenty-six weeks of regular UI benefits are provided by all States. (Two States
may provide more than 26 weeks of regular UI benefits. Iowa offers an additional
13 weeks of State financed regular UI benefits if individuals lost their jobs as a
result of the employer going out of business and Massachusetts offers 30 weeks of
regular UI benefits instead of the normal 26 weeks if the local unemployment rate
is greater than 5.1 percent).
Agriculture
For the majority of States, only employers with ten or more workers in twenty
weeks, or who paid $20,000 or more in wages in any quarter, are subject to
unemployment insurance laws. Farm owners/operators are excluded from
coverage in all States.
Domestic Service
Private households, social clubs, and college fraternities and sororities who
employ domestic help and pay wages of $1,000 or more in a quarter are subject to
unemployment insurance laws.
Nonprofit Organizations
Coverage is required for nonprofit organizations with four or more employees in
20 weeks. Almost half of the States, however, have elected more expansive
coverage, typically covering any organization with even one employee in twenty
weeks. Ministers employed by religious organizations to perform ministerial
duties are excluded from nonprofit coverage.
Self-employed Individuals and Unpaid Family Members
As defined by the unemployment insurance laws, employment is the hiring of
workers by others for wages. Self-employed individuals are therefore excluded,
August 2010

LAUS Program Manual 3-15

except in California, where they may elect to pay contributions for self-coverage.
Relatives are not covered unless they receive pay from the official business
payroll. However, the employment of minors by their parents, or parents by their
children, is excluded.
Railroads
Interstate railroad workers are covered by the Railroad Unemployment Insurance
Act administered by the Railroad Retirement Board. Workers on intrastate and
scenic railroads may also be covered.
State and Local Government Elected Officials and Others
All State and local government employees are covered under State UI laws with
the exception of elected officials, members of the judiciary, State national and air
national guardsmen, temporary emergency employees, and policy and advisory
positions.
Student Workers at Universities, Interns and Student Nurses
College and university students employed by the school at which they are
enrolled, such as work-study students, are excluded from coverage. Many States
also exclude the spouses of students who work at the university if the employment
is part of a program to provide financial assistance to the student. Student nurses
employed by hospitals as part of a training program are not covered. Similarly,
medical school graduates working as interns in hospitals are excluded from
coverage.
Armed Forces
Military personnel are excluded from State unemployment insurance coverage.
They are covered under a separate program, Unemployment Compensation for
Ex-Servicemen (UCX), but are not included in QCEW data. Civilian defense
workers, however, and all other federal employees covered under the
Unemployment Compensation for Federal Employees (UCFE) program are part
of the data reported to the QCEW program.
Agents on Commission
Insurance and real estate agents that are paid only by commission are excluded
from coverage in almost all of the States.

August 2010

LAUS Program Manual 3-16

UI Process
Just as each State has its own UI laws, each State has its own benefit payment
system for awarding UI benefits, providing documentation and fiscal control. The
benefit payment system is tied in with the taxation records system of the State.
Taxes levied from employers are deposited into the State's Unemployment Trust
Fund account.
Initial Claim
When an individual becomes unemployed, he or she must file an initial claim to
request determination for entitlement and eligibility to receive benefits.
Depending on the UI services available within each State, an unemployed person
may file an initial claim in person at an Employment Service office, over the
telephone, by mail, or via the Internet. There are three initial claim types: new,
additional and transitional.
The first claim in a benefit year filed to request benefits is
referred to as a new initial claim. A claim filed within the same
benefit year after intervening employment is call an additional
initial claim. A benefit year is the one-year period during
which an individual may receive UI benefits and is usually
related to the date of the individual's first spell of
unemployment and the filing of the claim.
A claim for benefits filed during the last week of a benefit year while a spell of
unemployment is ongoing and requesting an establishment of a new benefit year
and another eligibility determination is a transitional initial claim. A transitional
initial claim is an operational or administrative document facilitating the transition
from one benefit year to the next within a continuous spell of unemployment.
Therefore, it is excluded from the count of initial claims, since that count
represents new spells of unemployment.
Monetary Eligibility Determination
The new initial claim is evaluated, in accordance with the State’s laws, to
determine if the individual meets the monetary requirements necessary to
establish a benefit year and receive benefits and, if so, how much compensation
the individual is eligible to receive. Monetary eligibility for benefits is
determined by the amount of employment (in weeks or quarters) and wages
earned by the individual (in some combination of dollars and time worked) in a
specific base period.
A base period is a period of time prior to the benefit year (or period similar to a
benefit year) in which a claimant must have had a specified minimum amount of
insured (covered) work in order to qualify for benefits. Wages earned during this
period are used in determining the claimant’s weekly benefit amount (WBA) and
the claimant’s maximum total annual benefits. In the majority of States, the base
period is the first four quarters of the last five calendar quarters.

August 2010

LAUS Program Manual 3-17

There are two types of base periods, an individual and a uniform. The individual
base period varies as to the starting date for individual claimants, while a uniform
base period starts on the same calendar year for all claimants.
A benefit year usually consists of a 1-year period or a 52-week period during
which an individual may receive annual benefits. Nearly all States have what is
called an individual benefit year in which the beginning date is dependent on the
date the claim was filed.
Each State has its own formula for computing an individual’s WBA, total benefit
award, and duration of benefits. States may also elect to provide benefits to
dependents. See the annual ETA publication entitled Comparison of State
Unemployment Laws
(http://workforcesecurity.doleta.gov/unemploy/comparison.asp) for more details.
Claims that do not meet the State-specified requirements for the monetary
determination are denied benefits and result in a monetary disqualification.
Nonmonetary Determination
The new and additional initial claim is also subject to a nonmonetary
determination, in which a State determines whether individuals are eligible to
receive benefits based on the circumstances surrounding the loss of employment,
ability to work, availability for work, and activity in seeking work.
Each State has its own nonmonetary requirements for an unemployed individual
to receive benefits. All State laws provide that a claimant must have become
unemployed through no fault of his/her own and must be able and available to
work. The purpose of this is to provide benefits to individuals who are
unemployed primarily as a result of economic causes.
The nonmonetary determination is broken down into separation issues and
nonseparation issues.
Separation issues refer to situations surrounding the termination of the
employment relationship. This includes incidences where the individual
voluntarily quits without good cause, or voluntarily quits for personal reasons.
(For LAUS purposes, these individuals would be counted as unemployed even
though they do not qualify for UI benefits, as long as they are willing and able to
work, and are actively seeking employment.)
Nonseparation issues pertain to situations in which the individual's actions, the
type or seasonality of occupation, or income preclude eligibility. An individual
can be precluded from receiving benefits if they are not willing or able to seek
employment, or if they refuse suitable employment. Seasonal employees, such as
school personnel and professional athletes, are not eligible for benefits during the
time period between terms of employment. Disqualifying incomes includes
pensions, severance pay and other UI compensation, such as EB, TRA or DUA
benefits. Individuals disqualified because they were not able or willing to work
would not be counted as unemployed for LAUS purposes. Those disqualified for
receiving UI because they have income from pensions or severance pay would be

August 2010

LAUS Program Manual 3-18

counted as unemployed, as long as they are willing and able to work, and are
actively seeking employment and the amount of money earned does not exceed the
weekly benefit amount.
An individual may pass the monetary eligibility requirements but may not receive
benefits in the event of voluntary leaving without good cause, discharge or
suspension or misconduct, refusal of suitable work, labor dispute, and false
statements.
A claimant who does not meet the nonmonetary requirements and is denied
benefits is given a nonmonetary disqualification.
Nonmonetary Penalties
The circumstances of a nonmonetary disqualification may preclude the individual
from ultimately receiving benefits.
Separation Issues:
Voluntarily Leaving Work: Since the UI program is designed to compensate wage
loss due to lack of work, voluntarily leaving work without good cause is an
obvious reason for disqualification from benefits. All States have such
provisions. In most States, the disqualification lasts until the worker is again
employed and earns a specified amount of wages. However, in a few States the
disqualification is a fixed number of weeks and can be up to 12 weeks depending
on the reason why the individual needed to leave work. In addition, some States
may also reduce the individual’s benefit rights, usually equal in extent to the
weeks of benefit postponement.
Discharge for Misconduct: A number of States have a variable disqualification for
discharge for misconduct. In some States the range is small, for example, the
week of occurrence plus 3 to 7 weeks. In others, the range is large, 5 to 26 weeks.
Several States provide a fixed disqualification, and others disqualify for the
duration of the unemployment, or longer. Some States may reduce or cancel all
of the worker’s benefit rights and some provide for disqualification for
disciplinary suspensions.
Non-separation Issue:
Refusal of Suitable Work: Several States disqualify for a specified number of
weeks (3 to 20) any workers who refuse suitable work; others postpone benefits
for a variable number of weeks, with the maximum ranging from 1 to 12.
Monetary Penalties
The penalty for a monetary disqualification is in effect until the individual
becomes subsequently employed and earns a specified amount of wages to
become eligible.
Appeals
An individual whose claim has been denied for either
monetary or nonmonetary reasons may request an
opportunity for a fair hearing before a UI tribunal or UI
authority. A request can be made for a review by an

August 2010

LAUS Program Manual 3-19

appeals authority on the State's determination of the claim for benefits, the
employer's contribution rate, or a decision made by a lower appeals authority.
The employer for whom the individual worked during the wage qualifying base
period is charged with the liability for the claim payment. This employer may
also challenge the UI decision of the individual's eligibility to receive benefits.
Continued Claim
Once a claimant passes the monetary (for new initial claims only) and
nonmonetary eligibility requirements, the individual must satisfy mandatory
requirements for each week of unemployment for which he/she claim benefits.
These weekly requirements include actively seeking employment and being
available for work. Certification, or certifying, is the form and process by which
an individual attests to the facts that determine eligibility for a given week.
This certification process must be completed for each week that the individual
claims benefits. Most States establish a 52-week period (benefit year) during
which the individual may submit claims for benefits.
Individuals may also receive earnings from regular employment, or odd jobs,
while certifying for a week of unemployment. In such cases, these claims are
designated as continued claims with earnings. They are not used in LAUS
estimation since they do not meet the LAUS definition of unemployed. (Even one
hour of work results in the classification of the individual as “employed” in the
CPS.) Continued claimant counts without earnings due to employment are the
primary source of unemployment inputs for the Handbook method and an
important input into State models.
Final Payments
A final payment is the last continued claim for which an individual can receive
benefits within a benefit year. At this point, the individual has exhausted the
maximum benefits as was calculated at the time of filing and monetary
determination.
Final payment recipients, also referred to as UI exhaustees, are the primary input
to the monthly estimate of unemployed exhaustees of the Handbook method. (See
Chapter 7)
Benefits delivery arrangements
The State conducts UI activities under all of the following arrangements:
Intrastate Benefit Arrangements.
The State provides benefits to individuals who reside in and worked in that State.
The State also provides benefits to individuals who worked (and would continue
to seek work) in the State but reside in a border State. Intrastate claims filed in
the State where the claimant worked but does not reside are referred to as
commuter claims.

August 2010

LAUS Program Manual 3-20

Interstate Benefit Arrangements.
To encourage a claimant to move from a State where no suitable work is available
to one where there is a demand for the type of service that the claimant is able to
render, States have made agreements to protect the rights of workers who make
such moves. These arrangements permit the collection of benefits from the State
in which an individual has qualifying wages (liable State) even though the
claimant is not physically present in that State. The State in which the individual
is located may accept the claim, acting as Agent for the State that is liable for the
benefits claimed. The liable State may also accept the claim directly from the
claimant by telephone or electronic means. Determinations on eligibility,
disqualifications, and the amount and duration of benefits are made by the liable
State.
Wage-Combining Arrangements.
This arrangement permits workers to combine their wages and employment in
more than one State and file in a single State. This holds for situations where
there are insufficient wages and employment to qualify for benefits in any one
State and where, having sufficient wages and employment to qualify for benefits
in one State, the benefit amount would be increased by combining wages and
employment in other States. In addition, this arrangement permits workers having
sufficient wages and employment to qualify for benefits in more than one State, to
combine their wages in those and any other States in which they had wages and
employment in the base period of the liable State.

August 2010

LAUS Program Manual 3-21

UI Claims Process
The following chart illustrates the process of claims validation in a State.

August 2010

LAUS Program Manual 3-22

UI Claims Data for LAUS Estimation
Counts of individuals associated with continued claims are used in the
development of State and model-based area unemployment. Counts of
individuals associated with two types of claims, continued claims and final
payments, are used in the development of sub-state unemployment estimates.
Continued Claims
Continued claimant data for certifications to unemployment in the week including
the 12th of the month from the State UI program are used each month in the
development of State and area model-based total unemployment. The State UI
program data, along with data from the UCFE, and RRB programs, are used each
month to calculate monthly LAUS estimates for sub-state areas. Because the
unemployment measurement is limited to the labor force status of the civilian
population, claims data from the UCX program are excluded.
The continued claimant count is made up of persons who certified to a
compensated or non-compensated week of unemployment.
•

Compensated claims relate to those claimants who are receiving UI
benefits. A claimant who has worked during the week and received any
earnings while certifying for unemployment does not meet the criteria for
the CPS definition of unemployed and is omitted from the count used in
LAUS estimation even if the earnings do not result in a reduction in the
benefit award.

•

Noncompensated claims include individuals who are unemployed and are
certifying weekly for benefits but are not receiving compensation for any
of the following reasons:
o They are certifying during the waiting week,
o They are certifying while appealing a monetary or nonmonetary
disqualification. Or
o They are certifying while filing a pending claim.

These individuals are unemployed, are in the UI system, and are included in the
continued claimant count for LAUS estimation.
The LAUS program recognizes three types of continued claims data: intrastate,
commuters and interstate for both regular State UI and UCFE programs.
Intrastate claims
Intrastate claims are claims filed by unemployed persons in the State where they
live and worked.
Commuter claims
Commuter claims refers to claimants who worked and would continue to seek
work in one State while living in close proximity in another State. These

August 2010

LAUS Program Manual 3-23

claimants are treated as if they reside in the State of prior employment and file
intrastate claims in the State in which they had worked.
Interstate
Interstate refers to interstate claim filed by claimants who resided and became
unemployed in one State, and, during their spell of unemployment, moved to
another State and filed for UI benefits in the new State of residence. The State
where the claimant first became unemployed is still liable for that spell of
unemployment and the UI benefits that the claimant receives.
Interstate and commuter claims data are exchanged through the ICON system,
which enables States to transmit, and to retrieve appropriate claims data. The file
format used by ICON is called Liable-Agent Data Transfer (LADT).
Final Payments
Final payment recipients, also called UI benefit exhaustees, are continued
claimants who file for a week of unemployment that exhausts their total benefit
award.
Once claimants receive their final payment and leave the UI system, States are
unable to track them. The LAUS methodology estimates the number of UI
exhaustees who are still unemployed by applying a survival rate to the number of
final payments. This survival rate is developed from CPS data on duration of
unemployment. For further details regarding this process, see the Labor Market
Area Unemployment section of Chapter 7.

August 2010

LAUS Program Manual 3-24

Differences: UI Data versus the CPS
CPS data are used directly to produce official labor force estimates for the nation
and they are the key input to LAUS model based estimates of unemployment for
States and selected areas. Differences between the State UI count of continued
claimants without earnings and CPS unemployed result primarily from
differences in program coverage of the unemployed by the UI system.
Certain industries and occupations are excluded from UI coverage, including:
•

employees of certain nonprofit organizations;

•

insurance and real estate agents on commission;

•

agricultural workers on small farms and certain seasonal farm workers;

•

some domestic workers;

•

self-employed persons;

•

unpaid family workers;

•

some State and local government employees;

•

student nurses and interns in hospitals; and

•

railroad workers (covered under the RRB program).

Certain unemployed individuals may not be able to receive UI benefits, including:
•

people who have not worked long enough and therefore have insufficient
wages to establish eligibility for benefits under UI;

•

people who quit a job or were discharged for misconduct;

•

people who have exhausted their benefits and could not re-establish a
benefit year; and

•

people with no recent earnings, such as new entrants or reentrants to the
labor market.

The UI exclusions limit program coverage to unemployed individuals with recent
employment experience and exclude unemployed new entrants and reentrants to
the labor market. Also, much of agricultural unemployment is not represented in
UI statistics.

August 2010

LAUS Program Manual 3-25

BLS Standards for UI Data
LAUS labor force estimates produced for State and substate levels use the same
definitions and concepts as the CPS so that the resultant estimates are consistent
with the CPS and comparable within and across States. In order to further ensure
this, BLS sets standards for the UI statistics used in LAUS estimation
Standards for Continued Claims and Final Payments
Two insured unemployed counts, continued claimants and final payment
recipients, are used in the development of LAUS substate unemployment
estimates. For the model-based unemployment estimation procedure, only
continued claimants are used.
Continued claimants are persons certifying to a compensated or noncompensated
week of unemployment under the State UI and UCFE programs. Because
measurement is limited to the labor force status of the civilian population, the
UCX program is excluded. The continued claimant count includes intrastate
claimants, commuter claimants (based on State of residence), and interstate
claimants (based on State of residence). UCFE program data are not used in
developing unemployment estimates for modeled areas.
Persons receiving final payments are continued claimants certifying to a week of
unemployment which represents the last regular benefit payment in the benefit
year. Further benefits are not available until the beginning of a new benefit year.
The BLS standard of quality for these continued claims and final payment counts
is as follows:

August 2010

•

the counts reflect the State and county of residence of the unemployed;

•

the counts are unduplicated and based on the Social Security number and the
claimant’s week of unemployment for which the claimant certified;

•

the counts include both compensated and noncompensated claimants as
described above;

•

for continued claimants, the claimant’s week of certification is consistent with
the CPS reference week., i.e., the week including the 12th of the month(see
also the December reference week below);

•

for persons receiving final payments, the counts are weekly, based on the
week for which the claimant is certifying;

•

the counts exclude persons with any earnings due to employment, regardless
of their entitlement to full weekly UI benefits.

LAUS Program Manual 3-26

December Reference Week
Normally, the reference period is the week including the 12th of the month.
However, this is generally not the case for the month of December. In December,
the actual reference week used by the CPS and, thus, LAUS depends on the
number of business days between the 12th of the month and Christmas day. As a
result, the reference week is most often the week that includes the 5th of the
month, the week prior to the week that includes the 12th.
If the 12th of December falls on a Friday or Saturday, the reference period will be
the week including the 12th. However, if the 12th falls within Sunday-Thursday,
the reference period will be moved up to the week that includes the 5th. This
allows adequate time for CPS data collection and processing prior to the
Christmas holiday. Additionally, the change in the reference week is necessary
because CPS response rates fall substantially for the days immediately before the
Christmas holiday.
Standards for Initial Claims
While not part of direct LAUS estimation, initial claims are integral to the UI
process itself. Also, initial claims counts may be used in atypical or exception
procedures in the Handbook method to develop an estimate of those unemployed
who are eligible for UI but delay filing or never file for unemployment benefits.
(Estimates of delayed and never filers are no longer required in Handbook
estimation, but it is useful to define the standards for initial claims.)
An initial claim is a notice filed by an individual to request determination of
entitlement to and eligibility for compensation. A new initial claim is the first
claim filed by the claimant within the benefit year. An additional initial claim is a
second or subsequent claim filed by the claimant within the benefit year after an
intervening period of employment. The initial claims count representing new
spells of unemployment previously used in LAUS estimation included both new
and additional initial claims filed by individuals for State UI in the week
including the 19th of the month.
Reference Period
Unlike continued claims that relate to a certification period in the past, initial
claims do not refer to a reference period, but rather represent a point in time.
Information requested on the initial claim form typically includes the date of
filing, and the date of separation and the separating employer.
Excluded Groups
For purposes of defining spells of insured unemployment, the following types of
initial claims are excluded from consideration:
1) Invalid new initial claim where the individual is found to be monetarily
ineligible for UI.

August 2010

LAUS Program Manual 3-27

2) Transitional initial claim, where a new, unique spell of unemployment has not
occurred. Such an initial claim is filed by a claimant during a spell of
unemployment in the last week of his/her benefit year, requesting an eligibility
determination and establishment of a new benefit year. Because the claimant is in
a continuous spell of unemployment and is also filing a continued claim, such
transitional initial claims are excluded from the count representing new, emerging
unemployment.
3) Reopened claim, where a claimant reopens a continued spell of unemployment
after ceasing certifying to unemployment and withdrawing from the labor force.
If this atypical action does not reflect an intervening spell of employment, the
State may administratively reopen the claims series and allow the claimant to
resume filing continued claims. These claims are not to be considered initial
claims.
Standards for Residency Coding
With Federal funds allocated to areas below the State level, the use of claims data
by residence is imperative, not only as a determinant of the labor market area’s
total unemployment estimate, but also in the development of county and subcounty estimates. These estimates are created through a method called
disaggregation. For further details regarding this process see Chapter 9.
The residency requirement for claims data calls for the coding and tabulating of
claimants by State and county of residence. The geographic requirement applies
to counties within the State paying the benefits (or acting as agent State for
interstate claims) and to counties in contiguous States whose residents cross State
lines to file intrastate claims in the State holding their wages and paying the
benefits (commuter claimants). If State claims documents (either intrastate or
interstate) are preprinted with the State code, border State codes must be entered
for commuter claimants to insure correct residency information.
The UI Database Survey
Beginning in 1975 an effort was undertaken by BLS to improve the UI data used
in LAUS estimation. This effort started with a survey of the State UI database
systems. There were two primary reasons for the survey. The first was to
describe the methods States were using to obtain the UI statistics used in
developing LAUS estimates, and the second was to provide input to determining
the necessary modifications to State systems to achieve more uniform UI data
series. Based on this survey, the BLS standards for UI statistics were developed.
A plan of action was developed to eliminate inconsistencies in UI statistics both
between and within States as compared to official labor force concepts. Claimant
data that represented an unduplicated count of individuals by State and county of
residence for the appropriate reference period and with maximum adherence to
the CPS definition of unemployment in terms of any earnings due to employment
in the week of certification was the focus. These characteristics—unduplicated
count of persons, residency, reference period, and exclusion of persons with any
earnings—are essential elements for the UI claimant statistics used in
August 2010

LAUS Program Manual 3-28

unemployment estimation, and areas where improvement efforts have been
concentrated.
To obtain an unduplicated count of persons, Social Security numbers are used.
Also, the use of the week of certification prevents multiple counting because an
individual can certify only once to a week of unemployment.
Basing the UI statistics on claimant residency rather than more program-related
locations such as local office or place of employment ensures correspondence
with official labor force concepts. Proper residency coding affects all LAUS uses
of UI data.
The requirement that UI statistics relate to the week including the 12th of the
month also ensures correspondence with the official labor force estimates. The
exclusion of persons with any earnings makes the data consistent with the CPS,
where one hour of pay qualifies an individual as employed.
The initial UI Data Base Survey was a three-day-long meeting of State research,
State UI, and BLS staff. A detailed description of UI administrative and
operational processes was obtained.
Based on the results of the UI Data Base Survey, contracts were established with
all States to develop unique monthly statistical counts that reflect the BLS
standards for UI statistics and that are used for LAUS estimation purposes. These
State-developed tabulations have been replaced by tabulations developed by the
Program to Measure Insured Unemployed Statistics (PROMIS).
The UI Claims Review and Validation Project
In 2001 a simplified form of the UI Database Survey was introduced. Entitled the
UI Review and Validation Project, this examination of the data sources for UI
inputs into LAUS was designed to be less of a burden for the States to participate
in and to be conducted more frequently than the UI Database Survey. Due to the
ever-changing nature of technology, UI laws, and administrative procedures, it is
necessary to re-examine the data sources and procedures on a regular basis.
The project included reviewing and validating the quality of claimant data
extracted from the unemployment insurance (UI) system databases in the States.
In addition to the LAUS program, the MLS program requires UI statistics on
initial claims for the event trigger, continued claims for describing the ongoing
layoff, and final payments. Thus, the validation of UI data extract parameters and
specifications used in the State were integral to quality assurance of not only the
UI claims data inputs but also the LAUS estimates and the MLS statistics.
The validation procedure consisted of two separate stages, an initial examination
of the current procedures in place for obtaining UI claims data for use in LAUS
and MLS estimations and a follow up session to address any problems identified
in the initial examination.
The initial examination involved an on-site visit to the State agency by BLS staff.
The purpose of the visit was to interview the State staff who are responsible for
processing and extracting UI claims data for LAUS and MLS estimation and to
August 2010

LAUS Program Manual 3-29

conduct a detailed examination of the UI source files and the extract programs
used for LAUS and MLS purposes. A standard questionnaire form was used to
document the claimstaking method, the filenames where the claims are stored,
and to determine if the extract program was capturing all the necessary claims
data input.
Upon completion of the questionnaire, examination of the UI source files, and a
review of the extract program, a determination was made if the State's claims data
inputs for LAUS and MLS meet the BLS standards for UI claims data. If any
shortcomings are discovered, then a second stage is initiated to ensure that any
deficiencies will be corrected.
In the second stage, problems identified in the initial inquiry were addressed. The
BLS and the State staff determined solutions to ensure that the BLS data
standards are met. Solutions included the re-writing of UI claims extract
programs or other programs that process claims data for LAUS and MLS. In
addition, the BLS and State staff determined the necessary resources needed to
implement the corrections.
The validation process was an on-going project with each State being certified
and re-validated periodically as UI polices and technologies change. Two events
led to the elimination of this project. One was the limitation of travel to States
due to budget restraints starting in 2008. The other was the increasing number of
States implementing the PROMIS system.
The in-depth research and analysis required by States to gain approval to
officially use PROMIS has provided a much more comprehensive examination of
the participating State’s UI system and claims extract programs. In many cases
the set up of new UI database extract programs specifically for PROMIS has
revealed inaccuracies hidden in the legacy system’s extract program that went
unknown and were not uncovered by the validation project. The PROMIS
approval process has effectively removed the need for the UI Review and
Validation Project at the present time.
Program to Measure Insured Unemployment Statistics (PROMIS)
The PROMIS system is a stand-alone PC-based system that stores
all claimant information, including socioeconomic characteristics,
and generates the UI inputs to the LAUS and Mass Layoff Statistics
(MLS) programs. In addition to generating input files, PROMIS can
be used to develop tabulations at the State and area level of UI
claimants by socioeconomic characteristics. PROMIS operates as the
clearinghouse for multi-purpose input data, allowing flexibility to provide a more
complete picture of the unemployment situation at substate levels.
The PROMIS system is designed to provide States with increased quality
assurance and resource efficiency to develop monthly statistics. The
implementation of PROMIS by a State involves the creation of a new UI database
extract program. This enables the most up-to-date LAUS criteria to be
incorporated into the extract program and for the program to be written in a
August 2010

LAUS Program Manual 3-30

current programming language. The extracts used by most States for their legacy
systems were developed decades ago and were often not updated as needed.
These legacy extracts were written in programming languages that are now
antiquated and thus are confusing and unfamiliar to current State staff. A rigorous
examination of the PROMIS data and the legacy system data, which is also
required, has often exposed flaws in the legacy extract that went unnoticed.
The PROMIS data quality is further enhanced by the Residency Assignment
System, which corrects erroneous address information and assigns geocodes for
States, counties, cities and towns. (See the Residency Assignment System section
of this chapter on pages 3-29 to 3-47 for more information.) In addition, the use
of PROMIS facilitates the implementation of claims-based disaggregation by
providing claims data at the city and town level. (See Chapter 9 on
Disaggregation for more details.)
The following files are produced by the PROMIS system:
•
•
•
•
•
•
•

•
•

City Claims Input Files compiled by Counties and Cities
State UI and UCFE Continued Claims Less Earnings (batch IDs: M03,
M04, M05, M06, M07, M08) for the LAUS State System
State UI and UCFE Continued Claims Less Earnings (batch IDs: M03,
M04, M05, M06, M07, M08) compiled by Labor Market Area (LMA) for
New England States
Week 1 through Week 6 Final Payments (batch IDs: M10, M11, M12) for
the LAUS State System
Week 1 through Week 6 Final Payments (batch IDs: M10, M11, M12)
compiled by Labor Market Area for New England States
All UI and UCFE Continued Claims Less Earnings (batch IDs: M03, M04,
M05, M06, M07, M08), and Week 1 through Week 6 Final Payments
(batchM03, M04, M06, M07, M10, and M11
All UI and UCFE Continued Claims Less Earnings (batch IDs: M03, M04,
M05, M06, M07, M08), and Week 1 through Week 6 Final Payments
(batch IDs: M10, M11, M12) compiled by Labor Market Area (LMA) for
New England States, and one large file containing the output for batch IDs
M03, M04, M06, M07, M10, and M11
MicroData detail behind all LAUS output data (batch IDs: M03, M04,
M05, M06, M07, M08, M10, M11, M12, C06 if New England State, C07
if non-New England State)
12-month LSS output files for LAUS annual processing: one consolidated
file containing non-commuter output (batch IDs: M03, M04, M06, M07,
M10, M11); one file containing City Claims (batch IDs: C06 if New
England State or C07 if non-New England State); one consolidated file
with each of the commuter outputs (batch IDs: M05, M08, M12)

For more information, see the PROMIS system user’s guide available on the
LAUS/MLS Intranet under Operations and Manuals.

August 2010

LAUS Program Manual 3-31

Residency Assignment System
To assist States in correctly coding the residence of their UI
claims records, the LAUS National Office has made
available address correction and geocoding software called
the Residency Assignment System (RAS). RAS corrects
erroneous addresses, city, state and ZIP codes and assigns
FIPS codes for States, counties, incorporated places, minor
civil divisions (MCD) and census designated places (CDP).
It can also be customized to assign State specific city codes.
Overview
States provide the input file(s) and corresponding format file(s) that defines the
layout of the input file(s). States determine the layout of the output file and which
geocodes are to be assigned. Geocodes include State and county FIPS codes, FIPS
place codes, census designated place (CDP) codes, minor civil divisions (MCD)
codes, longitude and latitude, and census tracts and blocks. In addition, other
codes can be added such as LAUS codes or State specific codes. Job files are
created by national office staff to reflect the output file layout and the geocoding
requirements of the State.
States upload their input file via EUSweb to their State folder in either the LAUS
directory or the PROMIS directory (if the State is approved for PROMIS). Each
State can upload, download, rename or delete any of their files in their State
folder. The system will automatically process the input file as long as the file is
appropriately named (see the following section on input files).
A polling agent program searches each folder for a new input file and
automatically runs RAS when one is detected. Once the input file is processed,
RAS places the output files in the State folder and saves the input in the Archive
subfolder located within the State folder. The polling agent also notifies the State
via email when the file has been processed. If for any reason the same file has to
be processed again, the original input file can be moved or copied from the
archive folder back into the State folder, and the system will automatically
process it. The RAS output includes the processed file (STout.txt) with the correct
addresses and geocode assignment; the job summary report (ST.ajs), which
contains statistics on the input records and the level of assignment; and the error
report (ST.err), which lists each record that had errors along with explanations of
the errors.
Geocode Assignment
RAS is comprised of two databases, the Address-Level database and the Centroid
database, which verify and correct addresses in claims records and assigns
geocodes. Both databases can assign the following geocodes: the State and
county FIPS codes, longitude and latitude, and census tracts and blocks. However,
only the Address-Level database can assign the 5-digit FIPS place codes for

August 2010

LAUS Program Manual 3-32

incorporated places, census designated (CDP) places, and minor civil divisions
(MCD).
Geocode Assignment by Database Type
Address level database (rooftop)
Centroid database (ZIP code)
FIPS county codes
FIPS county codes
Census tracts
Census tracts
Latitude & longitude
Latitude & longitude
FIPS place codes
MCD codes
In order for RAS to assign a geocode from the Address-Level database, the input
record must be able to be corrected and matched to an entry in the Address-Level
database. This is called an exact Address-Level match or a roof top assignment.
If the input record does not contain enough detailed data in the record to make an
exact address level, RAS may be able to assign geocodes from the Centroid
database; however, as noted above the Centroid database can not assign MCD
codes or place codes.
Address-Level assignment reports a latitude and longitude set that is based on the
individual address. Assignment is to the exact individual address or not at all.
Thus an address failing to exactly match the Address-Level database will not be
assigned a FIPS place code or an MCD code.
Centroid assignment, on the other hand, reports a latitude and longitude set based
on the 9-digit ZIP Code. Its precision is at the block-face level. Latitude and
longitude may revert to the 7-digit or 5-digit ZIP Code level in areas where data
are less extensive.
The system will first attempt to process all records through the Address-Level
database. Records lacking the precise address information needed for an Addresslevel assignment will then be processed through the Centroid database which has
less stringent address requirements.
Limitations of the Address-Level Database
Certain place codes are not included in the Address-Level database. These include
place codes created after the 2000 Census, place codes for PO Boxes, and records
without exact address matches. Although a particular code may not be in the
database, the database does contain the geographic makeup of all places. For
more information, see the section entitled The Assignment of Codes Not in the
RAS Databases.)
Place codes created after the 2000 Census: Place codes created after the 2000
Census are not included in the Address-Level database. The RAS software uses
the Tele Atlas GeoCensus directory data. Tele Atlas supplies the majority of
today’s geographic information system (GIS) software packages with the
GeoCensus directory data. Although the database is updated continuously with
new geographic information, new place codes are only introduced at the time of
the decennial census.

August 2010

LAUS Program Manual 3-33

Although a particular code may not be in the database, the database does contain
the geographic makeup of all places in other forms. Census tracts and blocks can
be used to identify new places and assign the place code indirectly.
Records without an exact address match: States should note that if a record is not
an exact address match and thus not assigned a 5-digit FIPS place code or MCD
code, it signifies that RAS could not find an exact address match in the AddressLevel database. It does not necessarily indicate that the address in the record is
not located in that city.
Some or all of the Centroid level matches may be adequate to determine if the
address in a record is located in a certain city as they may be accurate at the city
street or block levels.
In situations where a specific city does not receive enough exact address matches
to assign place codes to support the expected level of unemployment, States can
opt to use a specific level of Centroid precision to assign a place code, such as a
match code of 1 (street segment), 4 (streets and blocks), or 5 (city blocks or entire
towns and villages).
PO Boxes: Records containing PO boxes for the address are not assigned MCD or
place codes unless they contain both mailing and residence address fields such as
in the LADT file. Otherwise, only the county FIPS code is assigned automatically
by the system. States must determine if they intend to count records with PO
boxes in the county of the Post Office or in the city of the Post Office. If the
records are to be counted in a city, the State must provide a list of the cities and
the corresponding place codes to be assigned.
For further information on RAS, see Appendix 3-4 for the RAS User’s Guide.

August 2010

LAUS Program Manual 3-34

APPENDIX TABLE 3-1
LIABLE/AGENT DATA TRANSFER RECORD
Required Fields by Record Type
Y = Required Field
N = Rule Number Listed at End of Layout
FLD
NBR

1

2

FIELD NAME

Social
Security
Number
Claimant’s
Name - First

FIELD
TYPE

BEGIN
COLUMN

FIELD
LENGTH

TIC

REOP/
TRAN

W/C IB

W/C COMM

N

1

9

Y

Y

Y

Y

A/N

10

12

Y

Y

DESCRIPTION

Enter Claimant’s
Social Security
Number
Enter the claimant’s
first name. First
position cannot be
blank. Enter at least
one alphabetic
character.
Claimant’s middle
initial.

3

Claimant’s
Name Middle Initial

A/N

22

1

Y

Y

4

Claimant’s
Name - Last

A/N

23

23

Y

Y

5

Mailing
Address Street
Mailing
Address - City
Mailing
Address State

A/N

46

30

Y

Y

Y

Y

Enter Claimant’s (Mailing) Street

A/N

76

19

Y

Y

Y

Y

A/N

95

2

Y

Y

Y

Y

Enter Claimant’s (Mailing) City
Enter Claimant’s (Mailing) State

Mailing
Address - Zip
Code
Residence
Addr - Street

A/N

97

9

Y

Y

Y

Y

Enter Claimant’s (Mailing) Zip Code

A/N

106

30

6

6

6

6

Enter Claimant’s (Residence) Street

Residence
Addr - City

A/N

136

19

6

6

6

6

Enter Claimant’s (Residence) City

6
7

8

9

10

August 2010

Enter the claimant’s
last name. First
position cannot be
blank. Enter at least
one alphabetic
character.

LAUS Program Manual 3-35

APPENDIX TABLE 3-1
LIABLE/AGENT DATA TRANSFER RECORD
Required Fields by Record Type
Y = Required Field
N = Rule Number Listed at End of Layout

(Continued)
FLD
NBR

FIELD NAME

FIELD
TYPE

BEGIN
COLUMN

FIELD
LENGTH

TIC

REOP/
TRAN

W/C IB

W/C COMM

DESCRIPTION

11

Residence
Addr State

A/N

155

2

6

6

6

6

Enter Claimant’s (Residence) State

12

Residence
Addr - Zip
Code

A/N

157

9

6

6

6

6

Enter Claimant’s (Residence) Zip
Code

13

Claimant’s
Telephone
Number

N

166

10

Y

Y

14

Year of
Birth

N

176

4

Y

Y

Y

Y

Claimant’s Year of
Birth - Format is
“CCYY”.

15

Sex

N

180

1

Y

Y

Y

Y

Enter the sex of the
claimant
1 = Male
2 = Female
3 = Unknown

16

Race

N

181

1

Y

Y

August 2010

Enter Area Code,
Exchange, and
Extension of the
Claimant’s
Telephone Number

Claimant’s Race
Code
1 = White
2 = Black
3 = Asian
4 = American
Indian/Alaskan
Native
5 = Native
Hawaiian/Other
Pacific Islander
6 = Information Not
Available

LAUS Program Manual 3-36

APPENDIX TABLE 3-1
LIABLE/AGENT DATA TRANSFER RECORD
Required Fields by Record Type
Y = Required Field
N = Rule Number Listed at End of Layout

(Continued)
FLD
NBR

FIELD NAME

FIELD
TYPE

BEGIN
COLUMN

FIELD
LENGTH

TIC

REOP/
TRAN

17

Education

N

182

2

18

Liable
State FIPS

A/N

184

2

Y

Y

19

Liable
State
Office/Call
Center
Number

N

186

4

Y

Y

August 2010

W/C IB

W/C COMM

DESCRIPTION

Highest Grade
Completed
01 - 12 Actual grade
completed (12 = GED)
13 = 1 year of college or
technical school
14 = 2 years of college
or Associate
degree/technical school
15 = 3 years of college
16 = 4 years of college
or undergraduate degree
17 = 1 year of post
graduate study
18 = 2 years of post
graduate study or
Masters degree
19 = Doctorate

Y

Y

Liable State FIPS Code.
The Liable State cannot
be the same as the
Agent State.

Enter number that
identifies the Liable
Interstate office/Liable
Call Center that handles
the claim.

LAUS Program Manual 3-37

APPENDIX TABLE 3-1
LIABLE/AGENT DATA TRANSFER RECORD
Required Fields by Record Type
Y = Required Field
N = Rule Number Listed at End of Layout

(Continued)
FLD
NBR

FIELD NAME

FIELD
TYPE

BEGIN
COLUMN

FIELD
LENGTH

TIC

REOP/
TRAN

W/C IB

W/C COMM

DESCRIPTION

21

Agent State
Local
Office/Call
Center
Number

N

192

4

Y

Y

Y

Y

Enter number that
identifies the Local
office/Call Center where
the claimant filed the
claim.

22

Residence
State FIPS

N

196

2

5

5

1

5

Residence State FIPS
Code. The Residence
State cannot be the
same as the Liable
State.

23

Residence
County
FIPS

N

198

3

Y

Y

2

2

Residence County FIPS
Code.

24

Residence
City/Town
FIPS

N

201

4

Y

Y

Y

Y

Residence City/Town
FIPS Code.

25

Date Claim
Taken

N

205

8

Y

Y

Enter the date the claim
was taken. Format is
“CCYYMMDD”.

26

Effective
Date of
Claim

N

213

8

Y

Y

27

Program
Type

N

221

1

Y

Y

Enter effective date of
the claim. Correlates
with today’s date,
backdate reason, and
Liable State. Format is
“CCYYMMDD”.
Enter the program type:
1 = UI
5 = UCFE
7 = UCX

August 2010

Y

Y

LAUS Program Manual 3-38

APPENDIX TABLE 3-1
LIABLE/AGENT DATA TRANSFER RECORD
Required Fields by Record Type
Y = Required Field
N = Rule Number Listed at End of Layout

(Continued)
FLD
NBR

FIELD NAME

FIELD
TYPE

BEGIN
COLUMN

FIELD
LENGTH

TIC

REOP/
TRAN

W/C IB

W/C COMM

28

Entitlement

N

222

1

Y

Y

Y

Y

29

SOC Code

N

223

4

3

3

3

3

30

Initial
Claim

N

227

1

Y

31

BYB

N

228

8

32

BYE

N

236

8

33

WBA

N

244

3

August 2010

DESCRIPTION

Enter the entitlement
type:
0 = Regular
1 = Extended Benefits
(EB)
2 = Federal Benefit
Extension
3 = Additional Benefits
(AB)
Enter at least the first 3
digits of the Claimant’s
Occupational
Classification (SOC)
Code (left justified)
followed by a zero, or
enter the first 4 digits of
the SOC.
Enter Status of Claim:
1 = New
2 = Additional
3 = Transitional
Benefit Year Beginning
Date. Format is
“CCYYMMDD”.
Benefit Year Ending
Date. Format is
“CCYYMMDD
Weekly Benefit Amount
(Include Dependents
Allowance)

LAUS Program Manual 3-39

APPENDIX TABLE 3-1
LIABLE/AGENT DATA TRANSFER RECORD
Required Fields by Record Type
Y = Required Field
N = Rule Number Listed at End of Layout

(Continued)
FLD
NBR

FIELD NAME

FIELD
TYPE

BEGIN
COLUMN

FIELD
LENGTH

TIC

REOP/
TRAN

W/C IB

W/C COMM

DESCRIPTION

34

MBA

N

247

5

Maximum Benefit
Amount (Include
Dependents Allowance)

35

Base Period
Wages 1st Qtr
Base Period
Wages 2nd Qtr
Base Period
Wages 3rd Qtr
Base Period
Wages 4th Qtr
Base Period
Wages 5th Qtr
Base Period
Wages Total

N

252

7

Enter BP Wages for the
1st Qtr

N

259

7

Enter BP Wages for the
2nd Qtr

N

266

7

Enter BP Wages for the
3rd Qtr

N

273

7

Enter BP Wages for the
4th Qtr

N

280

7

Enter BP Wages for the
5th Qtr

N

287

8

Enter Total BP Wages
for all quarters

NAICS (Employer
with Most
Wages)

N

295

6

Enter at least the first
four digits of the North
American Industry
Classification System
(NAICS) Code (left
justified), followed by
“00”, for the employer
with which the claimant
had the most wages, or
enter the 6-digit code.

36

37

38

39

40

41

August 2010

LAUS Program Manual 3-40

APPENDIX TABLE 3-1
LIABLE/AGENT DATA TRANSFER RECORD
Required Fields by Record Type
Y = Required Field
N = Rule Number Listed at End of Layout

(Continued)
FLD
NBR

FIELD
TYPE

BEGIN
COLUMN

FIELD
LENGTH

Last
Employer Name
Date
Employmen
t Began

A/N

301

30

N

331

8

Enter Date Employment
Began with Last
Employer. Format is
“CCYYMMDD

44

Date
Employmen
t Ended

N

339

8

Enter Date Employment
Ended with Last
Employer. Format is
“CCYYMMDD”.

45

NAICS –
(Last
Employer)

N

347

6

4

4

4

4

Enter at least the first
four digits of the North
American Industry
Classification System
(NAICS) Code (left
justified), followed by
“00”, for the claimant’s
Last Employer, or enter
the 6 digit code.

46

Last
Employer Ownership
Code

N

353

1

Y

Y

Y

Y

Valid entries are ‘1’
through ‘5’, default is
‘5’.
1 = Federal government
2 = State government
3 = Local government
4 = International or
Foreign
5 = Private

47

Separation

N

354

1

42

43

FIELD NAME

August 2010

TIC

REOP/
TRAN

W/C IB

W/C COMM

DESCRIPTION

Enter name of Last
Employer.

Separation:
1 = Permanent
2 = Temporary

LAUS Program Manual 3-41

APPENDIX TABLE 3-1
LIABLE/AGENT DATA TRANSFER RECORD
Required Fields by Record Type
Y = Required Field
N = Rule Number Listed at End of Layout

(Continued)
FLD
NBR

FIELD NAME

FIELD
TYPE

BEGIN
COLUMN

FIELD
LENGTH

48

Recall Date

N

355

8

49

Union

A/N

363

1

50

US
Citizenship
Alien
Registration
Number
Week
Ending
Date
Earnings
During
Week
Claimed
Date First
Payment
Issued
Exhaustee

A/N

364

1

A/N

365

20

N

385

8

A/N

393

1

N

394

8

A/N

402

1

51

52

53

54

55

56

Weeks
Compensat
ed

N

403

2

57

$ Amount
of Benefits
Paid

N

405

7

August 2010

TIC

Y

REOP/
TRAN

Y

W/C IB

W/C COMM

Y

Y

Y

Y

DESCRIPTION

Enter date claimant is to
return to work. If no
recall date, enter all
zeros, format is
“CCYYMMDD”.
Y = Yes
N = No
Y = Yes
N = No
Enter claimant’s Alien
Registration Number, if
applicable and available.
Week Ending Date of
week claimed, format is
“CCYYMMDD”.
X = Yes. Indicated that
claimant had earnings
during the week
claimed. Space = no
Enter the Date the First
Payment was Issued.
Format is “CCYYMMDD”.
X = Yes.
Complete only upon
exhaustion. Space = no
Enter the number of
weeks compensated
during the benefit year.
Enter the total amount
of benefits paid during
the benefit year.

LAUS Program Manual 3-42

APPENDIX TABLE 3-1
LIABLE/AGENT DATA TRANSFER RECORD
Required Fields by Record Type
Y = Required Field
N = Rule Number Listed at End of Layout

(Continued)
FLD
NBR

FIELD NAME

FIELD
TYPE

BEGIN
COLUMN

FIELD
LENGTH

TIC

REOP/
TRAN

W/C IB

W/C COMM

58

Commuter
Identificatio
n Code

A/N

412

1

59

Reopen
Claim/Trans
fer of Claim

N

413

1

60

Ethnic

N

414

1

61

Filler

A/N

415

57

62

Record
Type

A/N

472

1

Y

Y

Y

Y

Required Entry to
indicate Type of Record
1 = TIC (Telephone
Initial Claim)
2 = Weeks Claimed
3 = Reopen/Transfer

63

Process
Date

N

473

8

Y

Y

Y

Y

Enter the Date the claim
was Processed. Format
is “CCYYMMDD”.

August 2010

Y

DESCRIPTION

Y

Y

Y

X = Yes. Complete to
identify claims filed by
commuters from
Residence State. Space
= no
1 = Reopen claim.
Complete when there is
a break in claim series
not caused by
employment.
2 = Transfer of claim.
Complete when there is
a change in the
Residence/Agent State
with no break in claim
series.
0 = Neither

Enter claimant’s Ethnic
group. Valid values
are:
1 = Hispanic or Latino
2 = Non-Hispanic or
Latino
3 = Not Available
(for future use)

LAUS Program Manual 3-43

APPENDIX TABLE 3-1
LIABLE/AGENT DATA TRANSFER RECORD
Rule
Number
1

2

3

4

5

6

August 2010

Definition of Rule
On a Weeks Claimed (IB) – the Residence State FIPS can
be the same as the Liable State FIPS as long as the Agent
State FIPS is different.
Either Mailing Address or Field 23 (Residence County
FIPS) should be provided. For New England States,
either Mailing Address or Field 24 (Residence City/Town
FIPS) should be provided.
SOC should be provided if possible. If not provided, a
warning message will be returned. THE RECORD WILL BE
PROCESSED.
Last Employer – NAICS should be provided if possible. If
not provided, a warning message will be returned. THE
RECORD WILL BE PROCESSED.
On a Weeks Claimed – Commuter, the value in the
Residence State FIPS (field 20) will determine the
receiving state.
Residence Address will be completed when the Liable
State can provide a Residence Address that is different
from the Mailing Address.

LAUS Program Manual 3-44

APPENDIX TABLE 3-2
Agent Summary Report - Interstate Claims
01/03/00

LIABLE/AGENT DATA TRANSFER
PAGE 01
AGENT SUMMARY REPORT
STATE OF (Agent State Name)
FIPS ID IS: 00
AGENT STATE COUNTS FOR THE SAT. ENDING DATE 01/01/00

INITIAL CLAIMS
REG
EB
FSB
AB
TOTAL

UI
20
0
0
0
20

UCFE
0
0
0
0
0

UCX
0
0
0
0
0

TOTALS
20
0
0
0
20

REOPEN/TRANSFER
REG
EB
FSB
AB
TOTAL

UI
2
0
0
0
2

UCFE
0
0
0
0
0

UCX
0
0
0
0
0

TOTALS
2
0
0
0
2

UI
1358
0
0
0
1358

UCFE
24
0
0
0
24

UCX
11
0
0
0
11

TOTALS
1393
0
0
0
1393

WEEKS CLAIMED - IB
REG
EB
FSB
AB
TOTAL
01/03/00

LIABLE/AGENT DATA TRANSFER
PAGE 02
AGENT SUMMARY REPORT
STATE OF (Agent State Name)
FIPS ID IS: 00
AGENT STATE COUNTS FOR THE SATURDAY ENDING DATE 01/01/00
INITIAL CLAIMS
01 AL
ALABAMA
UI
UCFE
UCX
TOTALS
REG
0
0
0
0
EB
0
0
0
0
FSB
0
0
0
0
AB
0
0
0
0
TOTAL
0
0
0
0
02 AK
ALASKA
UI
UCFE
UCX
TOTALS
REG
1
0
0
1
EB
0
0
0
0
FSB
0
0
0
0
AB
0
0
0
0
TOTAL
1
0
0
1

August 2010

LAUS Program Manual 3-45

APPENDIX TABLE 3-3
Agent/Residence Summary Report - Commuter Claims
01/03/00

LIABLE/AGENT DATA TRANSFER
PAGE 01
AGENT/RESIDENCE SUMMARY
REPORT
FIPS ID IS: 00
STATE OF (Agent State Name)
CURRENT MONTH = 12/1999 PREVIOUS MONTH = 11/1999
COMMUTER CURRENT
UI
UCFE
UCX
TOTALS
TOTALS
REG
184
2
0
186
EB
0
0
0
0
FSB
0
0
0
0
AB
0
0
0
0
TOTAL
184
2
0
186
COMMUTER PREVIOUS
TOTALS
REG
EB
FSB
AB
TOTAL

UI

UCFE

UCX

TOTALS

442
0
0
0
442

0
0
0
0
0

1
0
0
0
1

443
0
0
0
443

01/03/00

LIABLE/AGENT DATA TRANSFER
PAGE 02
AGENT/RESIDENCE SUMMARY
REPORT
FIPS ID IS: 00
STATE OF (Agent State Name)
COMMUTER CURRENT
CURRENT = 12/1999 PREVIOUS = 11/1999
01 AL ALABAMA
UI
UCFE
UCX
TOTALS
REG
0
0
0
0
EB
0
0
0
0
FSB
0
0
0
0
AB
0
0
0
0
TOTAL
0
0
0
0
02 AK ALASKA
UI
UCFE
UCX
TOTALS
REG
0
0
0
0
EB
0
0
0
0
FSB
0
0
0
0
AB
0
0
0
0
TOTAL
0
0
0
0

August 2010

LAUS Program Manual 3-46

APPENDIX 3-4
Residency Assignment System User’s Guide
The Residency Adjustment System (RAS) is accessed using the EUSweb. State
LAUS technicians must have an EUSweb account. To obtain an account, State
technicians should contact their regional office. A server account consists of the
BLS assigned user ID along with a password and the necessary permissions to
access the BLS firewall and LAUS server.
States upload their inputs files through either the LAUS Program menu or, if the
State is participating in PROMIS, the PROMIS Program menu. RAS will
automatically process the files and return them to the appropriate State folder.
Email notification is automatically sent to State users to inform them that their file
has been processed and is ready to be downloaded.

Input Files
To use RAS, three files are needed for each claims file type to be processed; these
include the input file, the format file, and the definition file. States provide the
input file and the format file. The format file describes the layout of the input file
and must specifically identify the address fields. States are not limited to one file
type. For example, separate files can be processed for continued claims, initial
claims, LADT files or any other file a State may need to correct and geocode.
Corresponding format files must accompany each file type and individual RAS
job files must be setup for each file type.

August 2010

LAUS Program Manual 3-47

Input File
The input file is a UI claimant record file that may consist of interstate claims,
interstate claims and/or commuter claims. The data for the input file is obtained
from the State’s UI branch. The file must be submitted in a consistent format for
periodic processing since any changes in the input file layout requires that the
format file and associated jobs files be modified to reflect the new layout.
At a minimum, records must include street address, city, & zip. Other
information, such as local office ID, claim dates, and claimant information, may
remain in the file. The conversion process will not affect these fields. The input
file must be an ASCII file in fixed record length format. As previously
mentioned, for the system to automatically process the input file, it must be
named appropriately with the State’s alpha abbreviation. The following are
accepted file names and their possible contents:
Standard Input File Names
File Name

File Type

ST.txt generic file
STLADT.txt LADT file
STCOMM.txt commuter claims file
STCOM.txt commuter claims file variation
STHIS.txt generic history file
STANN.txt generic annual file
STPROMIS.txt weekly PROMIS input file
STLADTPROMIS.txt weekly PROMIS LADT input file
STINTRA.txt intrastate claims file
STREG.txt regular claims file
STCC.txt Continued claims file
STIC.txt initial claims file
ST = State alpha abbreviation

As long as the above criteria are met, the system can accept any file format. For
example, the complete LADT file can be submitted. Each record in the LADT file
contains 63 fields at a total of 473 bytes. The maximum number of fields per
record and the total number of bytes per record are limited together to 32,767
bytes.
RAS can also filter out unwanted records, such as claims with earnings and UCX
claims, prior to processing the input file. For large files, filtering out unwanted
records will speed up the RAS processing time.
In addition, large input files can be compressed before they are
uploaded to EUSweb using either WinZIP or PKZIP compression
utility software. Compressed files, or zipped files, use the same
naming conventions as above but must have the “ZIP” extension
instead of the “TXT” extension. The polling agent will automatically unzip any
files with the ZIP extension before sending them to RAS for processing. This
feature will significantly reduce the amount of uploading time for large input files
and eliminated EUSweb sessions from timing-out during the upload procedure.

August 2010

LAUS Program Manual 3-48

Formaat File
The fo
ormat file coontains a list of each dataa field, its length and datta type, and must
have the
t “FMT” extension.
e
Thhe file identiifies the fieldds and their locations wiithin
a reco
ord. Only thee address fiellds (street, city,
c
state, annd ZIP code)) need to be
speciffically identiified for RAS
S. It also identifies unneecessary fiellds that can be
b
ignoreed during proocessing or even
e
removeed before proocess beginss. Data that the
State does
d
not wannt to identifyy, such as Soocial Securitty numbers, should be
includ
ded in one laarge field aloong with otheer data. This way the poositions of thhe
sensitive data are not identifieed.
The fo
ollowing is an
a example of
o a format file.
f
In this case,
c
the Soccial Security
number can be hiddden in the “data”
“
field which
w
is 20 characters loong.
Formaat File Exam
mple
data,20,c
address,30,cc
city,19,c
state,2,c
zip,5,c
eor,2,b
d type is “c” for most of the fields indicating that
t the fieldd is a character
The data
type. Only the “eor”, or end of
o record, is a “b” for binnary type. Thhe eor indicaator
typicaally consists of two colum
mn lengths and
a is not visible unless using a text
editorr, such as Woord or TextP
Pad, which allows the vieewing of all spaces.
Althou
ugh the eor is
i takes up tw
wo column places,
p
it is usually
u
identtified as a single
symbo
ol, such as in Word orr in TextP
Pad. The eor indicator caan be viewedd by
the cliicking on thee following button
b
on thhe tool bar inn Word or TeextPad:
Definition File
RAS recognizes
r
a specific sett of input fields called PW
W fields. Foor RAS to proocess
the daata in a field,, that field must
m be mappped to one off the RAS-reecognized PW
W
fields. To map fieelds, a separaate file calledd a definitionn file, or DE
EF file, must be
createed.
The definition
d
filee also identiffies the type of input filee as ASCII. In this file and
a
the RA
AS as well, the
t input filee is referred to at the dattabase. In most
m cases thhe
definiition file wouuld initially contain one line stating "DATABAS
SE = ASCII"".
Durin
ng processingg, the individdual addresss fields and thheir associatted PW fieldds
will be added. Eaach definitionn file must have
h
the “DE
EF” extensioon. The definnition
file is created by the
t national office staff when
w
the RA
AS job files are
a setup.
Outp
put Files
Threee files are gennerated eachh time an inpput file is proocessed by RAS.
R
These
includ
de the outputt file with coorrected addrress fields annd assigned geocodes, annd
two reeport files thhat provide innformation on
o the geocooding precisiion and errorrs
associiated with thhe input file.
August 2010

L
LAUS
Progrram Manual 3-49

Output file
States have much flexibility in how their output files are produced. The output
can be as simple as having the same file layout as the input file only with
corrections. It can also contain additional fields assigned by the system. RAS can
overwrite existing address fields with corrected data or the corrected data can be
written to new fields in any location in a record. FIPS for States, counties,
incorporated places, CDPs, MCDs and State specific codes can be also be written
to any location in the record or appended to the end of each record. Several
output files can be produced from one input file. Multiple output files can be
designated by claim types, such as regular and commuter claims, or other criteria
depending on the needs of the State.
Once a file has been processed, the output file will have the same name as the
input file but will include the word “out” (ex. CTOUT.TXT). Multiple output
files would be named to reflect their contents.
Reports
Two reports are produced by the system, the Job Summary
Report, and the Error Report.
Job Summary Report: The Job Summary report provides
information regarding the job set up, the input data and the
results. This report shows the total number of records processed, the number of
records that were filtered out, the types of addresses, i.e. street, PO Box, Military,
etc., counts and percentages of the assigned and corrected address components
including the county FIPS, city names, FIPS places codes and ZIP codes, and the
error code listing and the number of records and the percentage of occurrences for
each error type.
The following pages show an example of the job summary report for the output
file.

August 2010

LAUS Program Manual 3-50

Job Summary Report
Job Summary

ACE 7.50c Page 5

21-Dec-2006
10:40:08am
Local Area Unemployment Statistics
Postal Code Summary Per File -----------------------------------------------File Name: P:\LAUS\ST\STOUT.TXT
Filter:
n/a
Records Passed by Filter:
22506
Postal
Assigned
Codes
Count
%
------ --------------ZIP
22495 99.95
ZIP+4
20204 89.77
DPBC
n/a
n/a
CART
n/a
n/a
LOT
n/a
n/a
LOT Or
n/a
n/a
County
22495 99.95
AGeo
16981 75.45
CGeo
5416 24.06
TaxIQ
0
0.00

Corrected
Count
%
--------------454
2.02
19734 87.68
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
0
0.00
0
0.00
0
0.00
0
0.00

Corrected
Components
--------------City
State
Trunc Addresses
Trunc Cities

Count
%
-------------696
3.09
4
0.02
554
2.46
0
0.00

Address Type Summary
(Percentages based on # of records assigned)
Count
%
Count
%
----------------------------------------------------------------------Street
16796 83.13
Firm
0
0.00
PO Box
1791
8.86
General Delivery
1
0.00
High Rise
1529
7.57
Military
0
0.00
Rural Route
87
0.43
Unique
0
0.00
CASS Qualitative Statistical Summary (QSS)
(Percentages based on # of records passing the filter)
Count
%
Count
%
----------------------------------------------------------------------High Rise Default
1063
4.72
Rural Route Default
16
0.07
High Rise Exact
470
2.09
Rural Route Exact
71
0.32
EWS Match
0
0.00
LACS Convertible
125
0.56
Error Code Summary ---------------------------------------------------------Error
----E101
E212
E213
E214
E216
E302
E412
E413
E420

Description
Count
------------------------------------------------------Last line is bad or missing
0
No city and bad ZIP
0
Bad city and no ZIP
0
Bad city and bad ZIP
0
Bad ZIP, can't determine which city match to select
0
No primary address line parsed
18
Street name not found in directory
382
Possible street name matches too close to choose
1
Primary range is missing
1524

August 2010

%
-----0.00
0.00
0.00
0.00
0.00
0.08
1.70
0.00
6.77

LAUS Program Manual 3-51

E421

Primary range is invalid for street/route/building

Job Summary

269

1.20

ACE 7.50c Page 6

21-Dec-2006
10:40:08am
Local Area Unemployment Statistics
----------------------------------------------------------------------------E422
Predirectional needed, input is wrong or missing
62
0.28
E423
Suffix needed, input is wrong or missing
16
0.07
E425
Suffix & directional needed, input wrong or missing
3
0.01
E427
Postdirectional needed, input is wrong or missing
21
0.09
E428
Bad ZIP, can't select an address match
0
0.00
E429
Bad city, can't select an address match
0
0.00
E430
Possible addr. line matches too close to choose one
0
0.00
E431
Urbanization needed, input is wrong or missing
0
0.00
E439
Exact match in EWS directory
0
0.00
E500
Other Error
2
0.01
E501
Foreign
0
0.00
E502
Input record entirely blank
0
0.00
E503
ZIP not in area covered by partial ZIP+4 Directory
0
0.00
E504
Overlapping ranges in ZIP+4 directory
3
0.01
E600
Marked by USPS as unsuitable for delivery of mail
1
0.00
-----------Total Error Codes:
2302
10.23
Geocensus Summary-----------------------------------------------------------Total Input Records:
Less 5 Digit Failure Records:
Net Input Records For Geo Processing:
Address Level Geo
Count
%
-----------------------------------(0) Matched
16981
81.01
(8) Non Matched
3981
18.99
------------------------------------Input Records:
20962
100.00

August 2010

22506
11
---------22495
Centroid Geo
Count
%
------------------------------------(1) Street Seg Match
349
6.33
(4) 7 Digit Centroid
974
17.66
(5) 5 Digit Centroid
4093
74.23
(9) Non Match
98
1.78
-----------------------------------Input Records:
5514
100.00

LAUS Program Manual 3-52

Error Report: The Error Report is a detailed report that shows you exactly which
records were assigned error codes during processing. The system uses these error
codes to indicate why it was unable to match the address in the USPS directories.
By reading these codes, you might be able to correct the data in specific records
or find a pattern of incorrect data entry.
Whereas the Job Summary Report tallies how often error codes occur, the Error
Report goes one step further and lists the actual addresses that were assigned
those codes. The error report can be customized to display any fields that are
included in the input file. Generally, the record number, the error code and the
addresses fields are written to the error report.
Not all error types result in the failure to assign geocodes. Many records that are
assigned errors may contain errors that pertain to only USPS address
standardization requirements. In addition, many errors can be corrected by the
system. The next page shows an example of an error report.

August 2010

LAUS Program Manual 3-53

Error Report
RUN ON: 21-Dec-2006 - 12:40:11PM
LAUS: STATE = TEST
----------------------------------------------------------------------REPORTING ON: D:\LAUS\ST\STOUT.TXT
RECORD
4
5
6
7
13
15

ERROR
E101
E212
E213
E214
E420
E423

ADDRESS

CITY

ST

ZIP

2 MASSACHUSETTS AVE NE
2 MASSACHUSETTS AVE NE
2 MASSACHUSETTS AVE NE
MASSACHUSETTS AVE NE
2 MASSACHUSETTS AVE

WILMINGTON
WILMINGTON
WASHINGTON
WASHINGTON

DC 99999
DC
DC 99999
DC
DC

10:40:08am
Local Area Unemployment Statistics
-----------------------------------------------------------

21-Dec-2006

Error Code List
Start Report at Record #
Max # of Records to Print
Nth Select
Report Type
Error
----E101
E212
E213
E214
E216
E302
E412
E413
E420
E421
E422
E423
E425
E427
E428
E429
E430
E431
E439
E500
E501
E502
E503
E504
E600

:
:
:
:
:

ALL
1
22
1.00
PW

Description
--------------------------------------------------Last line is bad or missing
No city and bad ZIP
Bad city and no ZIP
Bad city and bad ZIP
Bad ZIP, can't determine which city match to select
No primary address line parsed
Street name not found in directory
Possible street name matches too close to choose
Primary range is missing
Primary range is invalid for street/route/building
Predirectional needed, input is wrong or missing
Suffix needed, input is wrong or missing
Suffix & directional needed, input wrong or missing
Postdirectional needed, input is wrong or missing
Bad ZIP, can't select an address match
Bad city, can't select an address match
Possible addr. line matches too close to choose one
Urbanization needed, input is wrong or missing
Address found in Early Warning Sysem directory
Other Error
Foreign Address
Input record entirely blank
ZIP not in area covered by partial ZIP+4 directory
Overlapping ranges in ZIP+4 directory
Marked by USPS as unsuitable for delivery of mail

August 2010

LAUS Program Manual 3-54

How the Residency Assignment System works
The following sequence occurs when the system processes an
address:
1. Parse: The address is broken down into components (ZIP,
city, state, house number, street name, etc.).
2. Pre-standardize: Parsed components are pre-standardized to match to the
patterns of the directories by converting the data to full capitals, correcting
nonstandard abbreviations, and stripping out punctuation and extra spaces.
3. Match last line: The system reads the city, state, and ZIP and searches for
matching data in the City and ZIP to City file directories. First, by looking up the
city and state to find all ZIPS for the city, then, it looks up the input ZIP to find all
possible cities for the ZIP. By comparing the results of these two look-ups, the
program can verify that the last-line components agree with each other. At this
step, it may also correct the spelling of the city and state abbreviation. If the lastline components do not agree, the program expands the search to encompass a
larger metro area.
4. Match address line: The system searches the databases looking for records that
might match the input address line and secondary address. It evaluates all
potential matches and assigns a confidence score to each one. The program then
selects the record with the highest confidence score. (For a record to be chosen, it
must have a high confidence score, and score higher than the other possible
matches.) Once the program has chosen a matching record, it can finalize the ZIP
and assign geocodes selected by the State such as county FIPS and county name.
Assignment of FIPS Place Codes
The system is comprised of two separate databases: the Centroid level database
and the more precise data base, the Address-Level database. The FIPS place
codes are assigned from the Address-Level database. This is a more
comprehensive database that requires more precise input data requirements to
assign a geocode. The system will assign 5-digit place codes to all incorporated
areas and Census designated places (CDP) and minor civil divisions (MCD).
However, in order for the system to assign a place code, the input record must be
able to be corrected and matched to an entry in the Address-Level database. This
is called an exact Address-Level match or a roof top assignment. If the input
record contains errors such as missing street name, the city may be able to be
assigned through the Centroid level database; however, there would not be
enough detailed data in the record to make an exact address level match.
Address-level assignment:
A discrete range, a street name, directionals, a suffix, and a ZIP Code are required
when assigning Address-Level information. Also, Address-Level requires an
exact match on ZIP Code as well as an exact match on each component of the
address line in order to make a match.

August 2010

LAUS Program Manual 3-55

The Address-Level database is made up of 10 individual databases that represent
different geographic regions of the nation. These 10 databases are update
quarterly.
Centroid level matches:
A Centroid assignment does not require precise information for each address
component. Its precision is at the block-face level or 9-digit ZIP Code and reverts
to the 7-digit or 5-digit ZIP Code level in areas where data are less extensive. The
3 levels of Centroid matching are described below. The Centroid database is
updated bimonthly.
9-digit (ZIP+4 match or address level): The most specific Centroid geomatch. It is accurate to the block face and is the valid address
closest to the center of the block. Usually defines one side of a city
block.
7-digit (ZIP+2 match or ZIP sector): Designates a subdivision within a
ZIP Code. Zip+2 codes contain smaller groups of streets and
blocks than the 5-digit ZIP Codes. The 7-digit is calculated and in
general is the balance point of a smaller area within the 5-digit
area. The 7-digit area typically encompasses an 8 X 6 block area.
The Centroid can be polled off of the balance point by the density
of deliverable addresses within this area. Typically, the 7-digit
Centroid would not be polled more than a couple of miles off of
the calculated balance point, given the size of an 8 X 6 region of
city blocks.
5-digit (ZIP Code match or Centroid level): The least precise GeoCensus
match. It covers the greatest area of the 5-digit, 7-digit, and 9-digit
levels. A 5-digit ZIP Code area can be any size (i.e. many city
blocks or entire towns and villages).
Assignment of Codes not in the RAS Databases
As mentioned previously, the RAS databases do not contain codes for place
created or changed since the decennial census. However, the databases are
continually updated with the latest geographic information and do contain the
elements to identify any location.
In order to access the geographic data in the RAS databases and convert them to a
specific code, a search-and-replace table is used. A search-and-replace table lists
each search value and its replacement value. It can be either an internal table
created within a job file or an external table that resides in a separate file.
Place Codes: The most effective method to assign codes to places that are not
identified in the Address-Level database is to use census tracts and blocks in
conjunction with a search-and-replace table. This method is called the Census
method. Both the Address-Level database and the Centroid database are capable
of assigning a 10-digit census tract/block field.
The Census method uses geographic definitions of places by census tracts and
blocks. These definitions are obtained from the US Census Bureau. They are
August 2010

LAUS Program Manual 3-56

updated twice a year so they contain the latest information on new places and
places that have changed since the last decennial census.
The new Census method compares geographic codes from RAS to a crosswalk of
Census data. A fifteen digit geographic identifier is created from the two digit
State FIPS, three digit county FIPS, and the Census tract and block. The Census
block is the smallest area of geography shared by RAS and available Census files.
Even at this level of geographic precision, certain blocks will not be unique to a
single FIPS place or MCD. Adding county FIPS decreases, but does not
eliminate, the number of shared Census blocks.
To accommodate this limitation, the Census method uses two separate tables. The
first, much larger table consists of all the unique Census block / FIPS place pairs.
These census blocks are not shared between multiple FIPS places. Any record
with an address located in such an area will receive the associated FIPS place
code.
While this table accounts for a large portion of the State geography, it does not
exhaust it; as mentioned above, some Census blocks are shared by two or more
places within one county. Because we can not code more precisely, an additional
table is used. This table takes the fifteen digit code for the block and appends to
the end the official FIPS place names for all places common to that block. RAS
creates a similar code by appending the official USPS city name to its fifteen digit
code for each record and compares the result to the search and replace table. If a
match is found, the associated FIPS code is returned.
If RAS cannot find a match in either table, the system will attempt to provide a
code from its internal geocoding database. This is only available for perfectly
matched addresses. Typically, this provides codes for only out-of-State areas,
which are not part of the Census tables. Occasionally, in-State areas will also
receive a code via this method. These records reside in Census blocks with
multiple FIPS places. If the address has an acceptable USPS name that is not a
FIPS place name, then it will not match in the second Census table. It will,
however, receive a code from the RAS internal database if the record is perfectly
matched.
The following input is requested but not required from the State analyst:
Level of precision is required for use of the Census table coding:
The National Office highly recommends the use of the Census
method for codes “0” and “1”, which are perfect and near-perfect
address matches. As of January 2007, the geographic codes and
boundaries in the Census tables are six years more recent than the
internal RAS database and thus provide a better reflection of
current geography. Because match codes “4” and “5” are less
precise, the decision is left to the analyst as to whether to use the
method for these records. Use at these codes will extend the level
of geocoding returned by the system at the cost of more

August 2010

LAUS Program Manual 3-57

imprecision. No coding is available for records that receive a
match code of “9”.
By default, the National Office will use the Census method for
codes 0 -5.
Use of the second (“overlap”) table:
The second Census table is used for those Census blocks shared by
multiple FIPS places. This table provides little extra coding; early
testing show less than 1/10 of 1% of all records receive a FIPS
code via this table. Disabling this table will provide a small
decrease in overall processing time. Upon request, the National
Office can provide the analyst with approximate coding gains and
performance time losses that result from the use of this table.
By default, the National Office uses this table.
Use of RAS internal database for extra geocoding
As mentioned above, the RAS internal geocoding database is used
to supplement the output of the Census table method. In instances
where Census blocks were in a FIPS place in 2000 but are no
longer so, this will incorrectly return a FIPS code.
The analyst may direct the National Office to one of three options:
1.) Use the RAS internal database for all uncoded records;
2.) Use the RAS internal database for out-of-State records
only; or
3.) Do not supplement the output with the RAS internal
database.
By default, the National Office uses the RAS internal database for
all uncoded records.
State Specific Codes: This category consists of any codes that are not universally
recognized but are internal to a specific State. These codes may identify local UI
offices, UI call centers, New England city and town codes, or any other internal
code that may be used by States.
To utilize this function, a search-and-replace table must be created. A State must
provide the specific codes and their definition criteria for the search-and-replace
table. The table is created and added to the State’s RAS job files by LAUS
program office staff.

August 2010

LAUS Program Manual 3-58

Assessing geocoding precision
There are three different tools available for States to assess the level of geocoding
precision for their claims files. These are 1) the GeoCensus Summary, 2) the
Match Code, and 3) the Quarterly Geocoding Precision Report.
1) GeoCensus Summary: The GeoCensus Summary is found in the job summary
report. This report is generated each time a file is processed and contains
summary statistics on the number of records processed, the number and percent of
records by address type (including PO boxes), and the number and percent of
errors along with a description of the errors. The GeoCensus Summary is located
on the last page of the report and identifies the number and percent of records
assigned to a county or place by the Address-Level and Centroid databases.
Below is an example of the GeoCensus summary.
GEOCENSUS SUMMARY--------------------------------------------------------------TOTAL INPUT RECORDS:
1646
LESS 5 DIGIT FAILURE RECORDS:
0
---------NET INPUT RECORDS FOR GEO PROCESSING:
1646
ADDRESS LEVEL GEO
CENTROID GEO
COUNT
%
COUNT
%
------------------------------------------------------------------------(0) MATCHED
1178
72.14
(1) STREET SEG MATCH
24
5.13
(8) NON MATCHED
455
27.86
(4) 7 DIGIT CENTROID
142
30.34
(5) 5 DIGIT CENTROID
302
64.53
(9) NON MATCH
0
0.00
------------------------------------------------------------------------INPUT RECORDS:
1633
100.00
INPUT RECORDS:
468
100.00

Any records not containing all the address components will automatically bypass
the Address-Level database and therefore are not counted in the non matched
summary, item (8). As a result, and in the example above, the (8) non matched
count (455) may be less than the total number of records processed by the
Centroid database (468).
2) Match Code: The match code is a 1-digit field that identifies the level of coding
precision of the latitude and longitude assignment. A lower the number indicates
a more precise the assignment. The match code values correspond with the levels
of accuracy found in the GeoCensus summary and are listed below. (Note: These
codes can be appended to each record. States that wish to have the match code
field added to their output file should contact their regional office.)
The following codes are assigned:
(0) - Matched to Address-level
(1) - 9-digit match in Centroid (ZIP+4, center of ZIP+4 address range)
(4) - 7-digit match in Centroid (ZIP+2, postal sector)
(5) - 5-digit match in Centroid (close to center of ZIP - highest delivery
area is weighted and latitude/longitude adjusted toward that area)
(8) - Not matched in Address-level
(9) - Both options tried, but no match in either
3) Quarterly Geocoding Precision Report: The new Quarterly Geocoding
Precision Report is a snapshot of the geocoding precision for all States for
August 2010

LAUS Program Manual 3-59

particular quarter. It contains the same elements as the job summary report but
shows the cumulative percentage of records at the various levels of geocoding.
Address
Level
State

Pct of Exact
File Processing Total
PO
Match
Type
Date
Records
Boxes
(0)
Pct.

Centroid
Street
Segment
(1)

7-Digit
Centroid
(4)

5-Digit
Centroid
(5)

Non-Match
(9)

Cumulative Cumulative Cumulative Cumulative
Pct.
Pct.
Pct.
Pct.

ST PROMIS 11/13/06

32,034

0.92

92.2

93.4

96.9

100

0.0

ST

61,535

13.7

85.2

87.7

96.3

100

0.0

LADT

11/22/06

This table enables States to determine if they need to assign place codes for
records not coded by the Address-Level database. BLS recommends that States
achieve at least a 70 percent match rate for place codes. Looking at the above
example, the Address-Level assignment is at 85.2 percent for the LADT file. By
adding the Centroid street segment match (ZIP+4 match or address level) it brings
the total to 87.7 percent. Therefore the State may want to have the system assign
place codes for the records with match codes equal to “1” (the Centroid street
segment match). To code 96.3 percent of all records, the State would have to
assign place codes for records with a match code of “4” and for 100 percent, a
match code of “5”. Assignment of place codes for match code levels other than
“0” is discussed further in the next section.
This table will be produced at the beginning of each calendar quarter (January,
April, July, and October) and provided to States via EUSWeb. If a State sends
multiple files to the RAS system, statistics will be developed for each type of file,
and then a summary of all State files will also be produced.

August 2010

LAUS Program Manual 3-60

Adjustments to city claims for disaggregation
States should keep in mind that the use of RAS can affect their
disaggregation of experienced unemployment to the city level and
county level. RAS adds or corrects county and city codes to
claims with missing or incorrect codes. Any changes in the county
or city codes of claimant records must be reflected in both the numerator and the
denominator in the disaggregation ratios in the following ways.
o

City codes added or corrected can change the numerator in the
disaggregation ratios.

o

City codes removed from claims with addresses that are not within city
limits will affect the numerator in the balance-of-county claims
disaggregation.

o

Claims records without city codes should be included in the balance-ofcounty and added to the county claims total.

All addresses, including those not within city limits, will have the county code
added to the record. County records with addresses that do not reside within cities
are summed together to find the number of balance-of-county claims. This
number is the numerator for the balance-of-county disaggregation. The
denominator is formed by summing all claims for the county, including those
within and without LAUS cities. This must equal the sum of claims for all cities
within a county and the balance-of-county number.
States should review their current disaggregation ratios and adjust them based on
the geocoding assignment from RAS. The following examples are scenarios in
which States would have to adjust their current disaggregation ratios due to RAS
geocoding to ensure that the numerator and denominator in the ratios are
consistent.
1.) The Village of Elmwood Park is an official LAUS city. However,
Chicago, IL, is an acceptable USPS city name for addresses located in
Elmwood Park. Prior to using RAS, records may have been identified
according to the postal address city and coded to Chicago. RAS codes
these records according to physical location, so that they are instead
coded to Elmwood Park. This increases the disaggregation ratio for
Elmwood Park and lowers the ratio for Chicago, increasing Elmwood
Park’s share of the unemployed for Cook County, IL.
2.) Large portions of Sarasota County, FL, are unincorporated. Many
addresses have the postal address of Sarasota and Venice, even though
they reside in the Sarasota-Bradenton Unincorporated Area and not within
city limits. For unincorporated places, RAS would not assign a place
code. If records were previously coded based on the postal city, the use of
RAS would decreases the numerators of the disaggregation ratios for
Sarasota and Venice, decreasing those cities’ shares of Sarasota County,
FL, unemployed.

LAUS Program Manual 3-61

Geographic Resources
The follow Web sites offer services that are helpful in checking
and mapping addresses, and FIPS codes.

Geographic Names Information System (GNIS)
http://geonames.usgs.gov/pls/gnispublic
The Geographic Names Information System (GNIS), developed by the USGS in
cooperation with the U.S. Board on Geographic Names (BGN), contains
information for almost 2 million physical and cultural geographic features in the
United States and its territories. The Federally recognized name of each feature
described in the data base is identified, and references are made to a feature's
location by State, county, and geographic coordinates. The GNIS is our Nation's
official repository of domestic geographic names information.
This site allows you to query the Federal codes (formerly known as FIPS55)
database based on the feature name, county, class code, MSA code or place code.
It can be used to obtain information on the FIPS place codes assigned by the
Residency Assignment System, such as the physical coordinates (latitude and
longitude).
TIGER/Line® Files
http://www.census.gov/geo/www/tiger/
The TIGER/Line files contain the geographic definition of places by census tracts
and blocks. These files are used with the RAS search and replace tables to
identify places and assign codes for place codes not currently in the RAS
databases.
There is one TIGER/Line file (in a compressed format) for each county or county
equivalent. The file names consist of TGR + the 2-digit state FIPS (Federal
Information Processing Standards) code + the 3-digit county FIPS code (i.e.
TGR01031.ZIP for Coffee County, Alabama.) Each state folder contains
individual county files as well as a Counts file. The county files are stored in
compressed format and are compatible with PK Ware's PK Zip software. The
COUNTSnn.TXT files (where "nn" is the state FIPS code) show the counts for
the number of records for each record type by county for a state. If the count for a
particular record type is 0, then that record type does not exist for that county.
The Census Bureau produces the TIGER/Line files in ASCII text format only;
therefore, the data are NOT in the form of map images. To create maps with the
TIGER/Line files, one would typically use a Geographic Information System
(GIS) package or other mapping software.
Users are responsible for converting or translating the files into a format used by
their specific software package. For information on how to use the TIGER/Line
data with a specific software package one should contact the company that
produced that software.
LAUS Program Manual 3-62

American Fact Finder
http://factfinder.census.gov/servlet/AGSGeoAddressServlet?_lang=en&_program
Year=50&_treeId=420
The Census Bureau’s American FactFinder provides access to detailed tables and
maps for population, housing, and businesses. The address lookup feature can be
used to identify the validity of an address and map its location.
Census TIGER Maps
http://tiger.census.gov/cgi-bin/mapbrowse-tbl
Use this site to map the LAUS city and FIPS place coordinates obtained from
GNIS. The results will display the physical location of the place and the LAUS
city.
US Postal Service
http://zip4.usps.com/zip4/welcome.jsp
Use this site to check is an address is valid.
FIPS 55-DC3 Index
http://www.itl.nist.gov/fipspubs/55new/nav-top-fr.htm
This site contains flat files of all place codes for all States.
FIPS 55-3 Class Code Definitions
http://www.itl.nist.gov/fipspubs/55new/class76.htm
The Residency Assignment System assigns FIPS place code for incorporated
cities and census designated places (CDP). This site provides definitions for class
codes that can be obtained from The Geographic Names Information System
(GNIS) query or the FIPS 55-DC3 Index site.

LAUS Program Manual 3-63

4

Inputs to LAUS Estimation:
Establishment Data Sources

Introduction

T

here are two establishment-based data sources for employment estimates:
the Current Employment Statistics (CES) program and the Quarterly
Census of Employment and Wages Program (QCEW), commonly
previously referred to as the ES-202 program. The next two sections provide an
overview of the se two programs.

The Current Employment Statistics Program
The Current Employment Statistics (CES) program is responsible for a FederalState cooperative monthly survey of 140,000 business establishments nationwide
that operates in all States, the District of Columbian, Puerto Rico, and the Virgin
Islands. These 140,000 businesses and government agencies represent
approximately 410,000 individual worksites. Each month, the survey is conducted
in order to provide detailed industry data on employment, hours, and earnings of
workers on nonfarm payrolls for the nation, each State, and all 372 metropolitan
statistical areas and 8 areas in Puerto Rico defined by the U.S. Office of
Management and Budget.
The CES is an establishment survey that measures payroll jobs, unlike the Current
Population Survey (CPS) which is a household survey that measures employed
persons.
CES Concepts
Establishment. An establishment is defined in the CES as an economic unit, such
as a factory, mine, or store, which produces goods or provides services. It is
generally at a single physical location and engaged in one, or predominantly one
type of economic activity. Where a single location encompasses two or more
August 2010

LAUS Program Manual 4-1

distinct activities, these are treated as separate establishments, provided that
separate payroll records are available.
Employment. Employment is the total number of persons employed full- or parttime in nonfarm establishments during a specified payroll period. Temporary
employees are included. In general, data refer to persons who worked during, or
received pay for, any part of the pay period that includes the 12th of the month.
For Federal government establishments, employment statistics relate to civilian
employees only and are reported for the number of persons who occupied
positions on the last day of the calendar month. Persons are considered employed
if they receive pay for any part of the specified pay period, but they are not
considered employed if they receive no pay at all for the pay period. Therefore,
persons who are on paid sick leave (when pay is received directly from the firm),
on paid holiday, on paid vacation, or who work during a part of the pay period
even though they are unemployed or on strike the rest of the period are counted as
employed. Not counted as employed are persons who are on layoff, on leave
without pay, on strike for the entire period, or who were hired but have not yet
reported to work during the pay period.
Since proprietors, the self-employed, and unpaid family workers do not have the
status of paid employees, they are not included. Also excluded from the
employed are farm workers and domestic workers in households. Salaried
officers of corporations are included.
CES Estimation
The estimation methodology for the CES combines annual benchmarks from the
Quarterly Census of Employment and Wages program with monthly data from a
sample survey to produce estimates of emplo yment, hours, and earnings. All
firms with 1,000 employees or more are asked to participate in the survey, as is a
sample of firms across all employment sizes. In 2010, the CES sample consisted
of about 140,000 businesses and government agencies that represented
approximately 410,000 individual worksites drawn from a sampling frame of UI
tax accounts. The sample rotation plan allows most firms to report for 4 years and
then be rotated out of the sample for a similar period. A sample of smaller firms,
with probability of selection proportionate to size, is also selected. The sample
frame is the master list of establishments reporting to the Unemployment
Insurance system and maintained as the Universe Maintenance System by BLS.
Sample distribution is obtained by stratifying the universe of establishments for
each industry into employment-size classes. A total sample size sufficient to
produce adequate employment estimates is then determined and distributed
among the size classes in each industry based on the average employment per
establishment and the relative importance of each size class to its industry. This
amounts to distributing the total number of establishments needed in the sample
among the cells according to the ratio of the employment in each cell to the total
employment in the industry.
August 2010

LAUS Program Manual 4-2

Data are collected from the establishments surveyed on the report form BLS 790
or electronic equivalent. (The CES survey is often referred to as the 790
program.) Employment estimates are made at what is termed the basic estimating
cell level, and are aggregated upward to broader levels of industry detail by
simple addition. Basic cells are defined by industry (usually at the five- or sixdigit NAICS level). Within the construction industry, stratification by geographic
region also is used.
To obtain all-employee estimates for a basic estimating cell, the following five
steps are necessary:
1. A total employment figure (benchmark) is obtained for the basic estimating
cell as of a specified month (March).
2. For each report, employment is multiplied by the sample selection weight to
obtain weighted employment for the months for which estimates are being
made and for the previous month.
3. For each cell, the ratio of the weighted all employees sample total in one
month to that in the preceding month (termed the weighted link-relative) is
computed for sample establishments that reported for both months.
4. Beginning with the benchmark month, the all-employee estimate for each
month is obtained by multiplying the all-employee estimate for the previous
month by the weighted link-relative for the current month.
5. Add a net birth/death estimate from the model described below.
This method, the “link-relative technique”, produces month-to- month changes for
a matched sample of industry establishments. Aggregate monthly estimates are
produced by industry and geographic area.
Business birth and death modeling. A net birth/death factor is added to national
and State employment estimates to produce the monthly published estimates.
Regular updating of the CES sample frame with information from the UI universe
files helps to keep the CES survey current with respect to employment change due
to business births and deaths. The timeliest UI universe files available however,
always will be a minimum of 9 months out of date. Thus, the CES survey cannot
rely on regular frame maintenance alone to provide estimates of the employment
effects of business births and deaths. BLS utilizes a model-based approach for
this component.
While both the business birth and business death portions of total employment are
generally significant, the net contribution is relatively small and stable. To
account for this net birth/death portion of total employment, BLS has an
estimation procedure with two components. The first component uses business
deaths to impute employment for business births. The second component is an
August 2010

LAUS Program Manual 4-3

ARIMA time-series model designed to estimate the residual net birth/death
employment not accounted for by the imputation
Benchmarks
The establishment survey constructs annual benchmarks in order to realign the
sample-based employed totals for a specific month each year with the UI-based
universe counts for that month. These population counts provide an annual pointin-time census for employment. Until 2007, benchmark levels replaced the
original sample -based estimates from April of the previous year to March of the
benchmark year for each month. Improvements in the receipt of UI data and in the
standardization of State operations have enabled all States to replace estimates with
UI data beyond March of the benchmark year. In the March 2009 benchmark, 42
States and the District of Columbia used third-quarter 2009 UI data (that is, through
September 2009) in their benchmarking, and eight (8) States used second-quarter
2009 UI data (through June 2009).

Universe counts are derived from the administrative file of employees covered by
UI. Approximately 99 percent of in-scope private employment is covered by UI.
A benchmark for the remaining 1 percent is constructed from alternate sources,
primarily records from the Interstate Commerce Commission and the Social
Security Administration. The full benchmark developed for March replaces the
March sample-based estimate for each basic cell. The monthly sample-based
estimates for the year preceding and the year following the benchmark are also
then subject to revision.
Monthly estimates for the year preceding the benchmark month are readjusted
using a “wedge back” procedure. The difference between the final benchmark
level and the previously published sample estimate is calculated and spread back
across the previous 11 months. The wedge is linear. For example, when the
benchmark month is March, eleven-twelfths of the March difference is added to
the February estimates, ten-twelfths to the January estimates, and so on, back to
the previous April estimates which receive one-twelfth of the March difference.
This assumes that the total estimation error since the last benchmark accumulated
at a steady rate throughout the current benchmark year.
Estimates for the 11 months following the benchmark month are also
recalculated each year. These post-benchmark estimates reflect the application
of sample-based monthly changes to new benchmark levels, and the recomputation of bias adjustment factors for each month. Bias factors are updated
to take into account the most recent experience of the estimates generated by the
monthly sample versus the full universe counts derived from the UI.
Reliability of Estimates
Although the relatively large size of the CES sample assures a high degree of
accuracy, the estimates derived from it may differ from the figures that would be
August 2010

LAUS Program Manual 4-4

obtained if it were possible to take a complete census using the same procedures.
Although the estimates are adjusted annually to new benchmarks, estimates
subsequent to the benchmark month have several potential sources of error. The
amount added each month for new establishments, for example, may be too high
or too low. Changes in the industrial classification of establishments that result
from changes in their product or activity between benchmark months are not
reflected. In addition, small sampling and response errors may accumulate over
several months as a result of the link relative technique of estimation used
between benchmarks.

August 2010

LAUS Program Manual 4-5

The Quarterly Census of Employment and Wages Program
Background
The Quarterly Census of Employment and Wages (QCEW) program, also referred
to as the ES-202 program, is a cooperative endeavor of BLS and the workforce
agencies of the 50 States, the District of Columbia, Puerto Rico, and the Virgin
Islands. Using quarterly data submitted quarterly by the agencies, BLS
summarizes employment and wage data for workers covered by State
unemployment insurance (UI) laws and for civilian workers covered by the
program of Unemployment Compensation for Federal Employees (UCFE).
The QCEW program is a comprehensive source of employment and wage data by
industry at the national, State, and county levels. Unlike the CPS and CES
programs which are monthly sample surveys, the QCEW program is collected
quarterly and provides a virtual census of nonagricultural employees and their
wages. In addition, about 44 percent of all workers in agricultural industries are
covered.

Sources of Data
There are five sources of data for the QCEW program. They are initial status
reports, quarterly contribution reports, multiple worksite reports, Federal
Government reports, and annual refiling survey forms.
Initial Status Reports
Initial status reports are filed by new employers with the State UI tax unit to
initially register their business. These reports provide basic business identification
and classification information to establish a UI account. The employer’s liability
for UI taxes is determined from information provided in this report.
Quarterly Contributions Report (QCR)
QCRs are filed quarterly by all UI-liable employers to the State UI tax unit. These
reports provide the name and social security number of covered workers who
worked or received pay for the pay period which included the 12th of each month,
the total wages paid to covered workers, the portion of total wages subject to
unemployment insurance tax, and the employer contribution amount.
Multiple Worksite Report (MWR)
The MWRs are filed quarterly with the State QCEW unit by most employers with
more than one bus iness establishment. The MWR provides establishment- level
employment and wages data not otherwise available on the QCR.
Federal Government Reports
These reports are filed quarterly by most federal government agencies to report
August 2010

LAUS Program Manual 4-6

employment and wages data to the State QCEW unit, in accordance with the
Unemployment Compensation for Federal Employees (UCFE) program. Data for
non-defense federal agencies are provided to the State QCEW unit; information
for civilian employees of the Department of Defense is reported directly to BLSWashington.
Annual Refiling Survey Forms
UI- liable employers are surveyed by the State QCEW units periodically to verify
their location(s) and industry activity (ies). Employers are asked to verify
physical location, mailing address, and industry and ownership information and to
provide corrections if necessary

Data Compilation
State agencies code and summarize the raw data, check for missing information
and errors, and prepare imputations of data for delinquent reports. Each
establishment is classified by its industrial activity and then independently by one
of five ownership categories. (The five ownership categories into which
establishments are classified are private industry, Federal government, State
government, local government, and foreign or international government.) An
establishment is an economic unit, such as a factory, mine, or store, which
produces goods or provides services. It usually is at a single physical location and
engaged in one, or predominantly one, type of economic activity, for which a
single industrial classification may be applied. The North American Industry
Classification System is used to classify the industry of each establishment.
Employment data represent the number of workers on the payroll of covered
employers during the pay period including the 12th day of the month. Persons on
the payroll of more than one firm are counted in each firm. Workers are reported
in the State and county of the physical location of their job. Persons on paid sick
leave, paid holiday, paid vacation, and so on, are included, but those on leave
without pay for the entire payroll period, are excluded. The employment count
also excludes employees who earned no wages during the entire applicable period
because of work stoppages, temporary layoffs, illness, or unpaid vacations, and
employees who earned wages during the month but not during the applicable pay
period.
UI Coverage
Each State has determined its own laws regarding UI coverage, but they have
been greatly influenced by the federal government. The Federal Unemployment
Tax Act (FUTA) provides tax incentives that have ensured States’ conformity
with the minimum coverage standards set down in FUTA.
In general, a covered employer is defined under the FUTA as one who has a
quarterly payroll of $1500 in the calendar year or preceding calendar year, or one
August 2010

LAUS Program Manual 4-7

worker in 20 weeks. While many States have chosen to expand coverage beyond
the FUTA standards, the notable exceptions and limitations are noted below.
Agriculture
For the majority of States, only employers with ten or more workers in twenty
weeks, or who paid $20,000 or more in wages in any quarter, are subject to
unemplo yment insurance laws. Farm owners/operators are excluded from
coverage in all states.
Domestic Service
Private households, social clubs, and college fraternities and sororities which
employ domestic help and pay wages of $1,000 or more in a quarter are subject to
unemployment insurance laws.
Nonprofit Organizations
Coverage is required for nonprofit organizations with four or more employees in
20 weeks. Almost half of the States, however, have elected more expansive
coverage, typically covering any organization with even one employee in twenty
weeks. Ministers employed by religious organizations to perform ministerial
duties are excluded from nonprofit coverage.
Self-employed Individuals and Unpaid Family Members
As defined by the unemployment insurance laws, employment is the hiring of
workers by others for wages. Self- employed individuals are therefore excluded,
except in California, where they may elect to pay contributions for self-coverage.
Relatives are not covered unless they receive pay from the official business
payroll. However, the employment of minors by their parents, or parents by their
children, is excluded.
Railroads
Interstate railroad workers are covered by the Railroad Unemployment Insurance
Act administered by the Railroad Retirement Board and thus are not included in
the QCEW data. Workers on intrastate and scenic railroads may be covered for UI
and included in the QCEW data.
State and Local Government Elected Officials and Others
All State and local government employees are covered under State UI laws
with the exception of elected officials, members of the judiciary, State
national and air national guardsmen, temporary emergency employees, and
policy and advisory positions.
Student Workers at Universities, Interns and Student Nurses
College and university students employed by the school at which they are
August 2010

LAUS Program Manual 4-8

enrolled, such as work-study students, are excluded from coverage. Many States
also exclude the spouses of students who work at the university if the employment
is part of a program to provide financial assistance to the student. Student nurses
employed by hospitals as part of a training program are not covered. Similarly,
medical school graduates working as interns in hospitals are excluded from
coverage.
Armed Forces
Military personnel are excluded from State unemployment insurance coverage.
They are covered under a separate program, Unemployment Compensation for
Ex-Servicemen, but are not included in QCEW data. Civilian defense workers,
however, and all other federal employees covered under the Unemployment
Compensation for Federal Employees (UCFE) program are part of the data
reported to the QCEW program.
Agents on Commission
Insurance and real estate agents who are paid only by commission are
excluded from coverage in almost all of the States.
Earnings Data
Total wages, for purposes of the UI quarterly reports submitted by employers,
include gross wages and salaries, bonuses, tips and other gratuities, and the value
of meals and lodging, where supplied. In a majority of the States, employer
contributions to certain deferred compensation plans, such as 401(k) plans, are
included in total wages. Total wages, however, do not include employer
contributions to Old-Age, Survivors’, and Disability (OASDI); health insurance;
unemployment insurance; workers’ compensation; and private pension and
welfare funds.
Uses
The QCEW data serve as the basic source of benchmark information for
employment by industry and by size of establishment in the Current Employment
Statistics program.
The Unemployment Insurance Name and Address File, developed in conjunction
with the ES-202 report, also serves as a national sampling frame for establishment
surveys by the Producer Price Index, Occupational Safety and Health Statistics,
Employment Cost Index, and other compensation programs.

August 2010

LAUS Program Manual 4-9

Differences: Establishment Data Sources versus the CPS
The household and establishme nt data complement one another, each providing
information that the other cannot supply. Population characteristics, for
example, are obtained only from the CPS, whereas detailed industrial
classifications are much more reliably derived from establishment reports.
Certain differences can be accounted for, others cannot. It is useful to be aware
of the CPS/CES/QCEW differences for estimation and analysis purposes. Some
of the important differences are discussed below.
Place of Work vs. Place of Residence. CES and QCEW data are produced
according to the location of the establishment; CPS data provide residency-based
employment estimates.
Jobs versus Employed People: CES and QCEW develop estimates of jobs. The
CPS estimates employed individuals. Workers holding more than one job will be
included more than once in the CES and QCEW employment counts since they
will appear on more than one payroll record or contribution report. Persons
counted by the CPS are counted only once even if they hold multiple jobs.
Reference Period Differences. The reference period for the CPS is the calendar
week including the 12th of the month, except in December when it is often the
5th. The reference period for CES and QCEW is the payroll period including the
12th of each month, which could be weekly, biweekly, semi- monthly, for
example.
Employment Coverage Differences. The CPS definition of employment is total
employment, comprising wage and salary workers (including domestics and other
private household workers), self-employed persons, and unpaid workers who
worked 15 hours or more during the reference week in family-operated
enterprises. Employment in both agricultural and nonagricultural industries is
included. The CES and QCEW definitions reflect nonfarm wage and salary
employment and do not include self-employed and unpaid family workers. They
include some, but not all, domestics in private households and agricultural
workers.
CES and QCEW estimates include 14- and 15-year olds while the universe for the
CPS is limited to 16 years of age and older. CES does not cover workers who are
on unpaid absence for the duration of the pay period. These workers may be
considered employed in the CPS depending on their job attachment as determined
in the course of the interview. Workers who are on strike for the entire pay period
of the establishment are not included in the CES and QCEW estimates, but are
considered employed in the CPS.

August 2010

LAUS Program Manual 4-10

Uses of CES Data in the LAUS Program
The nonagricultural wage-and-salary estimates from the CES survey are used as
basic employment inputs for several LAUS estimating procedures. CES estimates
are used as variables for the State employment models. CES data are used in
adjusting place of work employment to place of residence and as current inputs to
labor market area employment, where available.
Use of CES in the State Employment Models
The statewide CES employment is a variable in the State employment model and
in the models for New York City, Los Angeles, and the respective balances of
New York and California. The CES data are adjusted to include individuals
involved in labor- management disputes. Those differences between the CPS and
CES discussed in the previous section do not require separate adjustments in the
model because coefficients are computed separately for each State from their own
data and the data relationships. The coefficients represent historical data
relationships and describe the degree to which the trends in the CES and CPS
employment series are related in each State. The CES variable is always included
in the employment model because nonagricultural wage and salary workers
represent such a large proportion of the employed. See Chapter 6, Statewide
Estimation, for more information on modeling techniques.
Use of CES for Estimating Current LMA Employment
The 372 metropolitan areas plus the 8 areas in Puerto Rico defined by OMB use
the nonagricultural wage and salary CES estimates in developing monthly LAUS
total employment estimates. If a Labor Market Area (LMA) is not covered by the
CES program—for example, a micropolitan or small LMA—but does have a
sample-based employment series developed under State auspices, these estimates
are used in the Handbook methodology. For LMAs without sample-based
employment estimates, nonsample (synthetic) estimation methods, using QCEW
estimates, yield place-of-work nonfarm employment. See Chapter 7 for details
on producing estimates for areas outside the CES program.
Use of CES in the Dynamic Residency-Adjustment Ratios
Current monthly area nonfarm employment estimates, which are establishmentbased, are converted to a residency basis by the application of the Dynamic
Residency-Adjustment Ratios. Commutation ratios were developed using Census
2000 employment data that displayed place of residence cross-tabulated by place
of work. The data provided information on county-to-county commuting flows
for each State in the United States and Puerto Rico. To determine significant
commutation patterns for a given labor market area, ratios between the Census
employment by place of residence by place of work and the CES March-April
2000 average employment by place of work were developed for each labor market
area and for all other areas to which at least 100 of the estimating area’s residents
commuted.
August 2010

LAUS Program Manual 4-11

When multiplied by the current month’s sum of CES total nonagricultural wage
and salary employment and the number of labor disputants, these ratio s yield the
residency-based LAUS total nonagricultural wage and salary employment for
each labor market area.
See Chapter 7 for a more detailed description of this adjustment.

August 2010

LAUS Program Manual 4-12

5

Inputs to LAUS Estimation:
Census Data

Introduction
he Census of Population and Housing, conducted every ten years by the
Bureau of the Census, is primarily intended to provide the population
counts necessary for apportionment of seats to the U.S. House of
Representatives and for determining legislative district boundaries. The
decennial census also has increasingly become a source of data for other uses and
provides socioeconomic and demographic data in addition to population
estimates.

T

The LAUS program methodology uses decennial census data
for adjusting establishment-based employment estimates to
residency-based employment estimates, for estimating certain
employment and unemployment components in the
Handbook methodology, and disaggregating or apportioning
labor market area estimates to smaller areas.
Post-censal population estimates are used in the State CPS estimation
methodology and the LAUS employment/population index share disaggregation
of the estimates for counties and, in some States, cities over 25,000. Ongoing
population estimation is conducted by the Bureau of the Census through a
Federal/State cooperative program. Statewide population estimates are produced
annually for the United States and counties; sub-county estimates are produced
biennially. Data are additive to the next level of geography, i.e., the State is the
sum of its counties. Except for the decennial census year, population estimates
pertain to July 1 of the reference year.

August 2010

LAUS Program Manual 5-1

Prior to the 2010 census, the questionnaire came in two forms: the short form,
which includes questions found on every form (100-percent questions) and the
long form, which also includes sample questions. The majority of individuals
received a short form where questions regarding household relationships, sex,
race, age, marital status, Hispanic origin and housing are asked. Approximately
one-sixth of the population received the long form where sample questions
include the following topics: (1) social characteristics such as education, place of
birth, ancestry, disability, and veteran status; (2) economic cha racteristics such as
labor force status, occupation, industry, class of worker, place of work, work
experience, and income; and (3) more detailed housing questions.
The 2010 census was limited to the short- form questionnaire which was sent to all
househo lds in the United States and Puerto Rico. Information previously gathered
by the long- form questionnaire is now provided by the Census Bureau’s
American Community Survey, which is collected in a sample of approximately 3
million households annually. The LAUS program is considering ways to utilize
information from the American Community Survey.

The Decennial Census: Enumerated and Sample-Based
Data
Tabulations based on the 100-percent enumerated questions are prepared down to
the block level. Tabulations for the sample questions are also prepared down to
the block level, but because they are based on a sample, the data are reliable only
for larger areas. Areas for which statistics are derived include Census regions and
divisions, Metropolitan Statistical Areas, Micropolitan Areas, Urbanized Areas,
Urban/Rural areas, Census county divisions, Census Designated Places, Census
tracts, Block Numbering Areas, Block groups, Blocks, Alaska Native village
statistical areas, Tribal designated statistical areas, and Tribal jurisdictional
statistical areas.

Sampling for Designation of Long Form Recipients
The basic sampling unit for the long form census is the housing unit, including all
occupants. For persons living in group quarters, the sampling unit is the person.
Two sampling rates are employed. In counties, incorporated places, and minor
civil divisions estimated to have fewer than 2,500 persons, one- half of all housing
units and persons in group quarters are included in the sample. In all other places,
one-sixth of the housing units and persons in group quarters are sampled. The
purpose of this sample design is to provide reliable estimates for small places.
When both sampling rates are taken into account, approximately 19 percent of the
housing units in the nation are included in the sample.

August 2010

LAUS Program Manual 5-2

Estimation of Census Sample Data
The estimation procedure used for the sample survey involves an iterative ratio
estimation technique, called proportional fitting, similar in concept to the
estimation procedures used for the Current Population Survey. In each tabulation
area, a characteristic total is estimated by summing the weights assigned to the
persons or housing units in the area. Initial weights for both households and
persons are assigned as the approximate inverse of the probability of selection for
the Census sample. Weighting areas are then created with a minimum sample of
400 persons.
Within a weighting area, the ratio estimation procedure is conducted in four
stages for both persons and occupied housing units. The first stage identifies 17
household-type groups which include classification by the number of persons in a
house and type of housing unit (e.g., persons in a housing unit with or without
children, group quarters, etc.). The second stage determines the sampling rate of
the weighting area. The third stage classifies persons as household/nonhouseholder and housing units as single- or multiple-units in a structure. The
fourth stage applies 180 aggregate age/sex/race/Hispanic origin categories.
Groups within these four categories are combined, if needed, to increase
reliability. In the final step, the initial weights undergo four stages of ratio
adjustment by which each group within each stage is multiplied in two iterations
by the ratio of the complete census count to the sum of the initial and subsequent
stage weights for each sample person.
Sample data are considered less reliable than enumeration, or 100-percent
questionnaire, data. However, estimated standard errors can be used to construct
confidence intervals around the sample estimates. These reliability estimates do
not account for nonsampling errors that are inevitable in a survey as extensive as
the decennial census, and which occur in enumerated as well as sample-based
data.

Nonsampling Errors
Nonsampling error can occur in the enumerated and the sample-based data, and
can introduce bias into the data as well as increase the total error associated with
the estimates. The Census Bureau tries to control for such error during collection
and processing procedures. Types of nonsampling error include undercoverage,
respondent and enumerator error, processing error, and nonresponse.
Every census results in an undercount, i.e., some people are missed. These
undercounts can occur by age, sex, and race categories. The Census Bureau
compares its data to other aggregate data sources to analyze the demographic
count differences. It also conducts a post-enumeration survey by taking a sample
of areas within the US and doing a very accurate count of the persons in those
areas. This allows the Census Bureau to estimate the extent of undercount. For the
1990 census, the total undercount of the population was less than 2 percent. The
2000 Census was negatively impacted by a data collection problem pertaining to
group quarters, particularly in towns with high percentages of college dormitory
August 2010

LAUS Program Manual 5-3

residents. In particular, the form used to collect labor force status information in
group quarters (the Individual Census Report, or ICR) was processed such that a
very large number of incomplete forms were systematically and erroneously
allocated to unemployment, resulting in implausibly high unemployment rates
being reported for these areas. BLS obtained special tabulations of householdonly data from the Census Bureau to utilize in certain estimating procedures. In
these tables, the household-only employment and unemployment measures
replace the corresponding total civilian measures.

August 2010

LAUS Program Manual 5-4

Differences: Census versus CPS/LAUS Estimates
Historically, there have always been differences between the Census and
the CPS programs in their respective estimates of unemployment. Prior to
1990, however, the census-based unemployment estimates generally
tended to be very close to the CPS estimates. For the 1990 census, this
relationship changed, with the 1990 census unemployment estimates
considerably higher than those from the CPS. The 2000 Census
unemployment estimates continued this pattern, with the Census
unemployment rate considerably higher than the national CPS.
There are several important differences between methodology used in
CPS/LAUS and the methodology used in the decennial census. These
produce differing unemployment estimates. It is important to know how
to interpret these differences and explain why the CPS/LAUS estimates
are regarded as the more reliable and accurate estimates.
1.) The census is a self-enumeration survey, while the CPS

survey is conducted in an interviewer-controlled
environment. This provides the CPS with more accurate
and detailed response information because interviewers are
present to clarify questions.
2.) The CPS questionnaire asks seven specific employment-

related questions to arrive at the labor force classification.
The census questionnaire asks only four. Misclassification
can occur as a result of fewer employment-specific
questions.
3.) The CPS has rigorous quality control procedures.
Interviewers are trained extensively, proficiency checks
are conducted regularly, and a portion of each month’s
households are reinterviewed as a quality control
measure.
4.) The CPS has a definite reference period, i.e., generally the
week including the 12th of the month. The census reference
period is officially April 1, but the questionnaire instructs
the respondent to provide information as of the week before
the questionnaire is completed.
5.) There is a known first month- in-sample reporting bias
whereby unemployment rates tend to be higher the first
time a household reports information. In the CPS,
households are interviewed for 4 months, not interviewed
for 8 months, and then interviewed again for 4 months, so
that 25 percent of the sample could be reflecting this bias.
The census is a one-time survey. Consequently, the entire
census could be affected by first-time reporting bias.

August 2010

LAUS Program Manual 5-5

Because of these reasons, both BLS and the Census Bureau have agreed that the
superior estimator of the labor force is the CPS.
The LAUS program uses decennial census data only where no other source of
data is available. The direct use of decennial census data is generally avoided
because of the superiority of the CPS and because relationships in the data are
unlikely to remain fixed over an entire decade.

August 2010

LAUS Program Manual 5-6

Uses of Decennial Census Data in LAUS
Uses of Labor Force Estimates
Census employment estimates are used in the employment/population
index share disaggregation method, which is used in conjunction with the
claims-based unemployment disaggregation method for counties and
cities. Because more current estimates of employed residents are not
available, the decennial Census estimate of this group is moved over time
by changes in annually prepared population estimates. In other words, the
Census employment to population ratio is maintained over the decade.
Decennial labor force estimates of total employed and unemployed for
sub-county areas are the basis of the Census-share disaggregation method.
The use of Census data for disaggregating labor force estimates is required
when UI claims data by county or city of residence are not available. The
method uses ratios of employment and unemployment in subareas to the
respective larger area totals. In this method, the relative distribution of
employed and unemployed is fixed for the decade.
In order to develop place-of-residence employment estimates for LMA’s,
census nonfarm employment levels and residence to work commutation
pattners, in combination with CES payroll employment levels, are used as
residency adjustment factors for monthly establishment-based
employment estimates. (Net commutation patterns from the census are
fixed for the decade.)
The census employment levels of agricultural employment and all-other
employment (self-employed, unpaid family workers, and domestics in
private households) are the benchmark levels for current LMA estimates
of these components.
The census estimates of all-other employment are also used in the
stratification of States into three groups for the purpose of developing
monthly change factors. The monthly change factors, referred to as Step-3
Ratios, are then used to estimate monthly all-other employment levels.
Census journey to work commutation data, which identify place of
residence and place of work estimates, are used in the designation of
LMA’s, including metropolitan areas, micropolitan areas, and small
labor market areas.

August 2010

LAUS Program Manual 5-7

Uses of Population Data
Census total population data have been
used in the population-share
disaggregation method for determining
sub-county estimates. This method is
used only when subarea UI claims data
or census labor force data are not
available. The disaggregation is based on
the ratio of total population in a subarea
to total population in the larger area,
applied to current employment and unemployment estimates for the larger
area. This method is applied only after the claims-based/population-based
disaggregation and census-share methods have been used to establish
estimates for the larger areas. Approval from the appropriate regional
office must be obtained before employing this disaggregation method.
Decennial census population estimates for States, and the subsequent
intercensal estimates, serve as the population controls in CPS estimation.
In a ratio estimation procedure, known population totals are applied to
sample ratios to improve the accuracy of the sample-based estimates of
levels.
Age-specific population counts are used in the distribution of new entrant
unemployed and reentrant unemployed estimates to States. They are also
used in the claims-based unemployment disaggregation of LMA entrants
and reentrants.

LAUS Program Manual 5-8

Employment/Unemployment Data
Data
Total Employment
Total Unemployment
Employment:
Total employment
Agriculture
All-other

Commutation Data

Use
Disaggregation of employment estimates
Disaggregation of unemployment estimates
Determination of appropriate weighting for
dynamic residency adjustment factors
Agricultural employment benchmark
Stratification of areas based on 1990/2000 relative
change; domestics and self-employed/unpaid family
benchmark
Definition of metropolitan, micropolitan, and small
labor market areas; also used in dynamic residency
adjustment factors

Population Data
16+ civilian, noninstitutional population
for States
Total population to
sub-county levels
Total population 16-19,
20+ to sub-county
levels (cities over
25,000)

CPS population controls
Disaggregation of employment from LMAs to counties
and counties to places
Claims disaggregation of LMA unemployment
estimates

LAUS Program Manual 5-9

Post-Censal Population Estimation
Post-censal population estimates are used in the State CPS estimation
methodology and the LAUS employment/population index share disaggregation
of the estimates for counties and, in some States, cities over 25,000. Ongoing
population estimation is conducted by the Bureau of the Census through a
Federal/State cooperative program. Statewide population estimates are produced
annually for the United States and counties; sub-county estimates are produced
biennially. Data are additive to the next level of geography, i.e., the State is the
sum of its counties. Except for the decennial census year, population estimates
pertain to July 1 of the reference year.

Statewide Estimation
National population estimates from the Bureau of the Census, which account
directly for births, deaths, and legal immigration, are done on a relatively
straightforward basis. The population of the States must be estimated using less
direct methods, because interstate migration, a large component of the change in
State populations, cannot be accounted for directly in the way that births, death,
and legal immigration can. There are two methods that are used, and the official
State estimates are an average of the two. The sum of the States is made to equal
the national total in a process known as proportional “raking”.
The first method for statewide estimation uses available administrative records.
This method measures interstate migration for persons under age 65 by using
address changes on tax returns and the number of exemptions claimed. For the
population of age 65 and over, migration is measured using the change in the
number of Medicare enrollees.
The second method uses a composite of a number of different factors. They are:
o
o

o

migration of the school-age population based on school enrollments;
a regression on changes in various population indicators measurable from
various data sources, for the population of adults under 65 years old; and
Medicare enrollments, for the change in the population 65 and older.

The age pattern of migration used in producing estimates of State populations by
age and sex (which affects the total 16 and over) depends primarily on the age
distribution of the base population, changes in school enrollments, and the
relationship between school-age migration and adult migration as of the last
census.

Substate Estimation
Substate population estimates are issued by the Bureau of the Census in two
series. In the P-25 series, Census publishes provisional estimates for States,
counties, MSA’s, and incorporated places. The annual P-26 series includes
LAUS Program Manual 5-10

revised estimates for all levels except incorporated places.
The estimation methodology is unique for each State, but generally includes an
average of three methodologies, with an appropriate “rake” to insure that areas
sum to the Statewide total.
o

o

o

Regression (ratio-correlation) Method. A multiple regression equation is
used to relate changes in a number of different data series to change in
population distribution. Independent variables may include automobile
registrations, elementary school enrollments, resident births for various
periods, and Federal income tax returns, among others.
Component Method II. This method employs vital statistics to measure
natural increase and school enrollment to measure net migration. These
estimates are specific to the civilian population under age 65. To this are
added an estimate of the population 65 and older based on Medicare
statistics and an estimate of the resident military population based on
Department of Defense data.
Administrative Records Method. This is an alternative component method
which uses individual Federal income tax returns to measure civilian
inter-county migration, and reported birth and death statistics to estimate
natural increase.

LAUS Program Manual 5-11

6

Development of
Statewide Estimates

Background

H

istorically, CPS samples have not been sufficiently large to produce
reliable monthly estimates directly from the survey for all States. As
a result, indirect methods have been used to estimate employment and
unemployment. As far back as 1960, Statewide estimates of employment and
unemployment were developed under uniform Federal procedures, using the
Handbook method. With the introduction of CPS State estimates in the
1970’s, a six- month moving average ratio adjustment to CPS levels
augmented the Handbook estimate. In the
late 1970’s, the Levitan Commission was
established to review the measurement of
the labor force in the United States.
Among the recommendations made by the
Commission in its report of 1978 was that
BLS explore replacing the Handbook with
an econometric approach to subnational
estimation.
Building on work done by Mathematica Policy Research under contract to
BLS, preliminary models were developed in the mid 1980’s. In order to
involve States directly in the research, the State Research Group, made up of
State Research Directors and BLS staff, was established in 1986 with the
support of the Interstate Conference of Employment Security Agencies
(presently called the National Association of State Work Force Agencies).
Regression and time series techniques were employed, with the models
extensively evaluated using empirical methods as well as recognized
statistical theory.

August 2010

LAUS Manual 6-1

Modeling can address the small samples in each State which result in
unacceptably high variation in the monthly CPS estimates of State
employment and unemployment. To produce less variable labor force
estimates as well as produce more stable seasonally adjusted estimates, BLS
developed time series models which ‘‘borrow strength’’ over time by using
historical series of sample observations for a given State to increase its
effective sample size. On average, the variance of month-to-month change in
the model estimates is about one third of the size of the CPS variance.
A type of regression model known as the Variable Coefficient Model (VCM)
best met the criteria. The VCM is so named because the coefficients in its
equations are allowed to vary over time to reflect structural changes in the
State’s data. The cha nging coefficients are estimated by the forward filter, a
widely used statistical technique that evaluates structural change against
sampling variability. The forward filter, also referred to as the Kalman Filter,
enables the VCM to handle the different relative accuracies that result when
an estimate draws upon data from several sources.
In 1988, a year of dual estimation of BLS and the States helped the states
make the transition from the Handbook to the VCM. In 1989, this new
method was implemented in 40 States. The remaining States were using
monthly CPS estimates of employment and unemployment directly.
During the early 1990’s, ongoing research at BLS brought about another
improved model that better dealt with error estimation and incorporated new
time series variables. Known as the Signal-Plus-Noise model, it also uses
variable coefficients and the forward filter. The Signal-Plus-Noise model
was implemented in January 1994. In 1996, time series modeling was
extended to the 11 more populous direct-use States because of reductions in
the size of the CPS sample.
The 2005 LAUS Redesign introduced a new generation of LAUS models.
The objectives of the new generation models were to implement direct
model-based seasonal adjustment with reliability measures and to improve
the benchmarking procedure by incorporating real-time monthly
benchmarking. At the same time, 6 area models were introduced along with
corresponding Balance of State models.
Real-time benchmarking addressed a number of concerns with the prior
generation of LAUS models. It reduces annual revisions by incorporating the
CPS benchmark on a current basis. It eliminates prior model biases and
benchmarking issues. It ensures that national events and shocks to the
economy will be reflected in State estimates as they occur. It also eliminates
the discrepancy between the sum-of-States estimates and the national notseasonally-adjusted totals. The following table illustrates differences
between the LAUS sum-of-States unemployment rates and the national CPS

August 2010

LAUS Manual 6-2

rates for both seasonally-adjusted and not-seasonally-adjusted estimates in
the years prior to the introduction of the third generation LAUS models.
Difference Between LAUS sum-of-States and CPS national
unemployment rates, 1996-2004 (LAUS minus CPS)
Month

1996

1997

January
February
March
April
May
June
July
August
September
October
November
December

-0.2
-0.1
-0.2
-0.2
-0.2
0.0
-0.1
0.0
0.0
-0.1
-0.1
-0.1

-0.2
-0.2
-0.3
0.0
-0.1
-0.1
0.0
-0.1
0.0
0.1
0.1
-0.1

January
February
March
April
May
June
July
August
September
October
November
December

-0.1
-0.1
-0.1
-0.2
-0.2
0.0
-0.2
0.1
0.0
-0.1
-0.3
-0.2

-0.2
-0.3
-0.2
-0.2
0.0
-0.1
-0.1
0.0
0.0
0.0
0.1
-0.1

1998
1999
2000
2001
Not Seasonally Adjusted
-0.1
0.0
-0.1
-0.2
-0.1
-0.1
-0.1
-0.2
-0.3
0.0
-0.3
-0.3
0.1
-0.1
-0.1
-0.2
0.0
0.0
-0.2
0.0
-0.1
-0.1
-0.1
-0.1
-0.1
-0.2
-0.1
-0.1
-0.2
-0.2
-0.2
-0.4
-0.1
-0.1
0.0
-0.2
-0.1
0.0
0.0
-0.3
0.0
0.0
-0.1
-0.3
0.0
0.0
-0.1
-0.4
Seasonally Adjusted

2002

2003

2004

-0.4
-0.3
-0.4
-0.3
-0.2
-0.3
-0.3
-0.3
-0.1
-0.1
-0.4
-0.4

-0.3
-0.3
-0.2
0.2
-0.3
-0.4
-0.2
-0.3
-0.2
-0.2
-0.2
-0.1

-0.2
-0.1
-0.3
-0.2
-0.2
-0.2
-0.2
-0.2
-0.1
-0.2
-0.2
-0.2

-0.2
-0.1
-0.1
0.0
0.0
-0.1
-0.1
-0.1
0.0
-0.1
0.0
-0.1

-0.2
-0.1
-0.1
-0.4
-0.3
-0.4
-0.3
-0.2
-0.2
-0.2
-0.5
-0.4

-0.1
-0.1
0.0
0.2
-0.3
-0.5
-0.3
-0.2
-0.3
-0.3
-0.3
-0.1

-0.1
-0.1
-0.2
-0.3
-0.3
-0.2
-0.2
-0.1
-0.2
-0.3
-0.2
-0.2

-0.1
-0.2
0.0
-0.2
0.0
-0.1
-0.2
-0.1
-0.1
0.0
-0.1
-0.1

-0.1
-0.2
-0.1
-0.1
-0.2
-0.1
-0.1
-0.1
0.0
0.0
-0.1
-0.1

-0.2
-0.2
-0.2
-0.2
-0.1
-0.1
-0.1
-0.3
-0.3
-0.4
-0.4
-0.5

The new models also address consistency issues ensuring that the sum of the
State estimates equal that of the nation every month. As part of the real-time
benchmarking procedure, each month the State’s estimates are controlled to a
Census Division. There are 9 Census Divisions which are in turn controlled
to the national CPS.

August 2010

LAUS Manual 6-3

In January 2010, BLS implemented smoothed-seasonally-adjusted (SSA)
estimates as the official seasonally-adjusted series for States’ labor force data.
SSA estimates incorporate a long-run trend smoothing procedure, resulting in
estimates that are less volatile than those currently produced by the LAUS
estimation methodology. The use of the SSA methodology is effective in
reducing the number of spurious turning points in current estimates. More
importantly, SSA estimation can reduce revisions in historical estimates and
remove the potential disconnection between historically benchmarked and
current estimates.

August 2010

LAUS Manual 6-4

Model Structure
The model structure introduced in 2005 utilizes both univariate and bivariate
modeling approaches. Univariate modeling is based only on the past values
of the CPS unemployment or CPS employment, and is utilized for Division
and Area models. This approach combines the two models; the time series
model of the Signal and the Noise model of the CPS survey. The sur vey
estimates used in the models are strengthened by the application of almost 3
decades of CPS sample data.
For State estimates, the bivariate modeling procedure is used. Bivariate
modeling of the series depends on the past values of the CPS and the past
values of a related series (payroll employment and UI claims) and the
relationship between the CPS and the input series
.Model Structure
Geographic Level
Type
Divisions
Univariate
Additive
States
Bivariate
Additive
Area

Univariate

Form

Unemployment-Multiplicative
Employment-Additive

The State models incorporate five features that are tailored to the properties
of each State’s data series. These features include the smoothness of the
trend, the stability of seasonal patterns, survey error, the relationship of the
CPS trend to State input trends, and the presence and types of outliers.

Signal-Plus-Noise Approach
The State CPS estimates are broken out into
the Signal, which represents the true value
of the employed or the unemployed and the
Noise, which represents survey error
inherent in the CPS sampling procedures.
The observed CPS estimate consists of a true, but unobserved labor force
value plus noise, which occurs because the estimates are derived from a
probability sample and not the entire population.
CPSt = Signalt + Noiset
Signalt = true value
Noiset = survey error

August 2010

LAUS Manual 6-5

The signal/noise estimation models are based on a modeling approach that
accounts for and extracts the noise from the CPS time series data, thus
providing a better estimate of the signal. An important component of the
noise in the CPS data is sampling error; its characteristics are known, or at
least can be estimated from survey design information. Two other factors
that account for the noise are irregular movements in the data and occasional
outliers. When there is a change in the CPS level of employment or
unemployment, that change is a combination of the change in the true labor
force signal and the change in the noise. The goal of the models is to isolate
the signal from the noise to avoid distortions in the CPS estimates and obtain
the best possible estimates of the true labor force values.
The models for the employment and unemployment estimates are a
combination of two processes: a signal estimation and a noise estimation.
The signal is a time series model that is based on historical data relationships
that are used in estimating current true labor force values, so a long historical
CPS time series is required. While the time series model of the signal
depends on past relationships, it does not require that these relationships be
fixed over time. A very important feature of this model is a built- in selftuning mechanism, known as the forward filter, which automatically adjusts
the regression coefficients and trend and seasonal components to adapt to
gradual structural changes as they occur.
Sudden, unpredictable changes in the time series relationships are handled by
incorporating outlier effects into the model.
The noise estimation clears up the distortion caused by the CPS sampling
error.

August 2010

LAUS Manual 6-6

Signal
In addition to survey error, there are other sources of variation in the CPS
time series. These sources are identified as the seasonal, trend and irregular
components and are taken into account in the modeling procedure.
Based on the decomposition of a time series into the trend, seasonal, irregular
and survey error components, the model form may be either additive or
multiplicative. In a multiplicative model, the seasonal variation is
proportional to the level of the series. As the trend rises, the magnitude of the
seasonal variation around the trend rises. The magnitude of the survey error
is inversely proportional to the level of the series. The standard errors are
therefore relative measures of error.
Time Series Model of the CPS
CPSt =

Tt + St + It +et
Tt x St x It x et

Sigt =

CPSt -et
CPSt / et*

Where:

Survey Error = et
Seasonal = St
Trend = Tt
Irregular = It
Model Estimation
The first step of the estimation process is to correct the CPS estimate for
survey error. This step also occurred in the generation of models that were in
use up to 2005.
CPSt – et* = Sigt*
The second step is to seasonally adjust the error corrected CPS. The seasonal
adjustment procedure is model based. In the previous generation of models,
seasonal adjustment was performed independent of the model by the use of
ARIMA X-11 software. Once the seasonal factor is removed from the error
corrected CPS, the remainder consists of the trend and the irregular
components.
Signalt* - St* = Tt* + It*
The following example illustrates an additive time series model of the CPS.
As mentioned above, the CPS is decomposed into trend, seasonal and
irregular components.
CPSt = Tt + St + et
August 2010

LAUS Manual 6-7

Where:

Tt = global linear trend
St = fixed seasonal pattern
et = purely random (irregular)
The global linear trend model represents a linear relationship between
dependent variable (CPSt ) and t, where t is a time indicator. The magnitude
and direction of growth are fixed by the slope. The growth per period is
determined by the β units. The fixed intercept affects the level of the series.
The initial level is determined by α, but has no effect on growth. The
smoothest possible trend would require that that growth lies on a straight line
centered through the series.
Tt = α + βt
t = 1, ...N
The fixed seasonal model consists of 12 coefficients, one for each month of
the year. Each coefficient, or factor, measures the seasonal effect on the
series for a given month. Additive seasonal factors have positive and
negative values that indicate deviation above and below the trend level due to
seasonality. Summing all 12 months of seasonal factors will equal 0. The
fixed seasonal pattern repeats each year.
Sm = cm
m = month index 1, 2, ... , 12
S1 + S2 + ...+ S 12 = 0
In this example the trend for time period t has an intercept of 10 and a slope
of 0.26.
Tt * = 10 + 0.26t
S1 * = 2.0, S2 * = 3.5, S 3 * = 4.0, ... , S12 = 0
First, the CPS is corrected for survey error. The survey error model uses an
autoregression approach with the current standard error as a weighted sum of
its previous values plus the current random error, v t. The coefficients are
provided from sample survey information.
et = 0.34et-1 + 0.19et-2 +0.10et-3 +0.02et-4 +0.02et-5 +0.02et-6, ... + v t.
A regression equation is used to model survey error to account for the overlap
in CPS (autocorrelated error) and changes in reliability. The autocorrelated
component, et, is adjusted by a variance inflation factor (VIF). The VIF is
based on standard errors computed for the CPS.

August 2010

LAUS Manual 6-8

etadj = et *VIF t
Removing the estimated error from the CPS yields a value for the signal
which equals the value for the signal computed from the trend and seasonal
components.
CPSt – et* = Signalt*
Signalt* = (10 + 0.26t) +Sm
et* = CPSt - Signalt*
Next the error corrected CPS is seasonally adjusted by removing the seasonal
component.
Signalt* - Sm * = Tt*

Variable Regression Coefficients
The simple model may be generalized to handle real series. Most series have
changing trends and evolving seasonality. The trend component cannot
respond to a change in the direction of the series. The seasonal component
cannot respond to changes in the seasonal pattern.
Using variable regression coefficients (VC) allows the coefficients to vary
over time so the model can adapt to changing patterns. In estimating the
coefficients, more recent observations receive more weight than earlier
observations. Past data that are less relevant to current conditions are
discounted.
When the trend coefficients vary over time, the trend is able to adapt to
changing patterns. A poorly fitting fixed linear trend may miss important
turning points in the series. A variable slope trend resolves this problem.
Tt = α t +βtt
Similarly the seasonal component adapts to changing patterns and the
seasonal factor for month m will change over time. Seasonal factors that are
fixed may no longer reflect the current situation while seasonal factors are
made adaptive with the variable coefficient.
Sm,t = cm,t

August 2010

LAUS Manual 6-9

In addition, the VC is a self-tuning mechanism where the model adapts itself
without requiring special intervention. Each component has a “hyper
parameter” associated with it that determines how much it changes over time.
The hyper parameter is identified as σi. If σi = 0, then the component is
fixed. If σi > 0, then the component changes continuously over time. The
hyper parameters are estimated from State data.
Component
Trend level (intercept)
Trend slope
Seasonal
Irregular

Fixed
σlevel = 0
σslope = 0
σseasonal = 0
σirregular = 0

Varying
σlevei > 0
σslope > 0
σseasonal > 0
σirregular > 0

Noise
Accounting for CPS Sampling Error
There are two properties of the CPS, all controlled through the models, which
affect the time series data: changing reliability and the correlated sampling
error.
Changing Reliability

Changing reliability is due to one or a combination of several factors. These
factors include survey redesigns after decennial censuses, sample size
changes due to budget cuts or special supplementation, and variations in
labor force levels. Because of these factors, the CPS sampling error variance
is not fixed over time.
As the reliability of the CPS estimates changes, so do the weights used to
estimate the signal. The estimated signal is a combination of an estimate
based on the time series model of the signal historical data and current CPS
estimates corrected by a model-based estimate of sampling error. The
reliability of the CPS can change over the years. As it improves, less weight
is given to the time series model and more weight to current CPS estimates.
The reverse is true for periods when the reliability weakens. Thus, the
estimated signal is a weighted average of a predicted signal based on
historical data and the current CPS estimate. This is represented by the
equation below:
Signalt = (1-w t) Signalt (prediction) + wt (CPSt - Nt )

where:
Signalt = the model estimate of the signal.
wt = the weight, between 0 and 1, given to the current CPS.
Signalt (prediction) = the model-based prediction of the signal
based on historical relationships.
Nt = the noise.

August 2010

LAUS Manual 6-10

The lower the reliability of the CPS, the less weight is placed on the current
CPS; the higher the reliability, the higher the weight.
Correlated Sampling Error
Because of the CPS 4-8-4 sample rotation method, there are significant
overlaps in the samples used by the CPS. (See Chapter 2.) Each month
three- fourths of the sample from the previous month is interviewed, oneeighth of the sample is interviewed for the first time, and one-eighth is
resuming interviews after being out of the sample for 8 months. Each month
one-half of the households from 12 months earlier are interviewed. The chart
below shows the proportion of the households in the current sample that were
also in the sample k months ago. For example, 75 percent of the households
in the sample this month were in sample last month, 50 percent were in two
months ago, etc. Note that samples from 4 to 8 months and over 15 months
apart have no households in common.

The use of a rotation system requires the periodic replacement of the sample.
To cover a decade under the 4-8-4 scheme, 15 samples are needed. A key
feature of the replacement scheme is that successive samples are generated in
a dependent way. Once an initial sample of households is selected,
replacements are obtained from nearby addresses. For each original sample,
the 14 succeeding ones needed to cover the decade are usually taken from the
same neighborhood.
The overlap in the CPS sample is important because it introduces strong
autocorrelation in the sampling error. That is, the current value of the
sampling error (either an overestimate or an underestimate of the true value)
will depend on its own past values. For example, suppose the unemployment
rate for the sampled households in the current month is higher than the rate
for the entire population. Since 75 percent of these households will remain in
the sample next month, the unemployment is likely to be overestimated
again.
The extent of this autocorrelation depends not only on the overlap in the
sample but also on the stability of the labor force characteristic being
estimated. The overwhelming majority of workers spend most of their time
in the labor force as employed rather than unemployed. Accordingly, errors
in the employment estimates will be more strongly autocorrelated than in

August 2010

LAUS Manual 6-11

unemployment estimates since employment is a more stable characteristic of
the households being sampled.
While CPS standard error estimates have been routinely produced for State
CPS data, estimates of the error autocorrelations have not. Obtaining this
information is potentially very costly, involving complex calculations on
huge micro data files. However, as part of BLS model research activities, a
method has been developed to estimate the autocorrelations that requires only
State CPS data for each rotation group.
The following graph presents the time profile of a CPS error series for a
typical State for both employment and unemployment. The vertical axis
gives the weights that show how the effect of a CPS error occurring in a
given month is distributed over future months, and the horizontal axis
indicates the number of months following the occurrence of the error. For
the current month (zero months ahead), the weight equals one since the full
impact of the error is felt in the month that it occurs. A value of 0.6 for 1month ahead, for example, indicates that 60 percent of the error in the current
month carries over into the next month’s estimate.

The strong autocorrelation in the CPS sampling error has important
consequences. First, sampling error will account for long-run movements in
the CPS. Ordinarily, we think of sampling error as having a transitory effect
on a series. If completely random, its effects would quickly average out.
This means the weights wo uld drop to zero for all months following the
occurrence of the error.
Another important consequence of correlated sampling error is that these
errors are unlikely to average out over a 12-month period. If the error were
August 2010

LAUS Manual 6-12

completely random, the number of overestimates and underestimates should
be about the same. However, because of the strong autocorrelation in the
error, if the CPS underestimates one year, it is likely to do so the next year.
Outliers
CPS data are occasionally affected by outliers. These outliers are CPS values
that are inconsistent with the expected behavior of either the signal or noise
component. There are two possible causes—a nonrepresentative sample
resulting in a noise outlier, or a real non-repeatable event, such as bad
weather, strikes, etc., that cause an outlier in the signal. Because these
outliers represent sudden changes, they may cause special problems for a
model. In fact, we define an outlier as an observation that breaks the pattern
of behavior predicted by the model. It is not necessarily an extreme value in
the observed series. For example, a series may not change much from one
month to the next, but an outlier may have occurred if the series normally has
a large seasonal increase.
Even though there may be extreme observations in the CPS accompanied by
a few large prediction errors, it is not necessarily good practice to make
special adjustments to the model to fit those observations more closely. The
purpose of the model is to capture the normal time series behavior of the
signal. Thus, the model must be flexible enough to adapt to structural
changes in the signal, but if too flexible, it will fail to filter out the noise. If
there is prior information about the occurrence of an outlier, then an
adjustment may be justified. Otherwise, adjusting the model for outliers is
important only to the extent that they distort diagnostic testing, cause bias in
parameter estimates, or lead to a deterioration in current performance.
Determining the type of outlier is crucial to deciding how to adjust the model
for its effects. Even though there are many complex patterns of outliers, the
three below tend to be the most common types of outliers that occur in time
series data:
1.) An additive outlier (AO) affects the series for only one month, such as a
sudden increase followed by a decrease.
2.) A temporary change (TC) in the level of the series causes an abrupt change
in the series followed by a gradual return to its former level
3.) A permanent level shift (LS) refers to an abrupt shift that persists
indefinitely into the future, or until an offsetting shift in the opposite
direction occurs.

The outlier may be due to a real change in the labor force or result from the
measurement process, which includes sampling and other types of
measurement errors. The origin of the outlier determines whether it should
be included in the signal or the noise component. Ideally, this should be
resolved by seeking external information about the potential causes. In
practice, such information is rarely available. Since highly transitory outliers
in the CPS are more likely to be due to the measurement process than a real

August 2010

LAUS Manual 6-13

economic event, the usual procedure is to assign these types of outliers (AO
and TC) to the noise component. On the other hand, a permanent shift in
level is considered a real effect and assigned to the trend component of the
signal. However, such identification requires that significant number of
months of data be available following the occurrence of the outlier in order to
identify the type of outlier that occurred. Therefore, outlier identification
cannot be made in current estimates. Models are monitored on a current
basis to detect the occurrence of outliers in the current year. Once enough
data become available to identify the nature of the outlier, its effects are
incorporated into the model specification and implemented during the annual
re-estimation of the models.

Univariate Trend Models
The univariate models are used to correct the CPS for survey error and to
seasonally adjust the model estimates. The local linear trend (Tt) of this
model is comprised of a variable coefficient trend where the intercept and the
slope change over time.
Tt = α t + βt-1
where:
α t = Tt-1 + ∇α, t
β t-1 = β t-2 + ∇β, t-1
∇ = per period change
A change in the local trend can result from a change in the intercept or a
change in the slope. When there is a change in the intercept, there will be an
up-down shift in the level. The slope, on the other hand, changes the trend
more gradually.
The smoothness of trend is based on whether the intercept or the slope
accounts for most of the change. It the intercept is dominant, the trend
appears rough. If the slope is more important, then the trend looks smooth.
Based on this, the local linear trend can appear in three forms: smooth, rough
and general. The type of trend is determined from the data by estimating the
empirical variability in the intercept and slope (hyper parameters).
A smooth trend can result from the trend having a fixed intercept and a fixed
slope (global trend), or a fixed intercept and a changing slope. The trend line
shows continuous change but not abrupt shifts. Turning points in the series
are well defined with local peaks and troughs.
Tt = α t + βt-1
Hyper parameters: σlevei = 0 and σslope > 0
A rough trend is caused by a changing intercept with no slope. All change is
due to shifts in the level. This give s the trend line a jagged look with many

August 2010

LAUS Manual 6-14

small changes in direction. Occasional large shifts tend to be associated with
major business fluctuations.
Tt = α t
Hyper parameters: σlevei > 0
The general form has shifting intercepts and slopes. It may approach
behavior of either the rough or smooth trend, depending on the relative size
of the change in the intercept and slope components.
Tt = α t + βt-1
Hyper parameters: σlevei > 0 and σslope > 0

Bivariate Trend Models
The bivariate trend models are used to seasonally adjust the State inputs from
the univariate model and improve the seasonal adjustment of the CPS. This
model explicitly estimates the relationships between the trends of the
univariate models and determines how strongly correlated the CPS trend is
with the UI claims and CES employment series.
The univariate model approach takes steps to correct the CPS trend for
survey error. It is represented in the following equation.
Tcps,t = αcps,t + β cps,t
The bivariate model, which addresses the seasonally adjusted State inputs, is
represented in the following equation where x is the UI or the CES.
Tx,t = α x,t + β x,t
Next the bivariate model relates the CPS to the State inputs, or Tcps,t to Tx,t.
There is a variety of possible relationships. There can be relationships
between the trend levels (intercepts), between growth rates (slopes), or some
combination of both.
The strength of these relationships is measured by trend correlation
coefficient s. A value of 1 indicates a perfect relationship and a value of 0
signifies no relationship.

August 2010

Corlevel =

1
0

perfect relationship
no relationship

Corslope =

1
0

perfect relationship
no relationship
LAUS Manual 6-15

The relationship between the trend levels is identified as the following
equation where the h coefficient relates a level shift in Tcps,t to Tx,t. and hα x,t is
the common trend. The R coefficient is the residual unique to the CPS. If h
= 0, then the Corlevel = 0 and there is no relationship between the CPS and the
State input. If Corlevel = 1 then Rα,t = 0 since the two are perfectly related
there can be no CPS residual.
α cps,t = hα x,t + Rα,t
Similarly the relationship between the growth rates is identified as the
following equation where the g coefficient relates growth rates in Tcps,t to Tx,t.
and gβ x,t is the common trend. The R coefficient again is the residual unique
to the CPS. If g = 0, then the Corslope = 0 and there is no relationship between
the CPS and the State input. If Corslope = 1 then Rα,t = 0 since the two are
perfectly related there is no CPS residual.
β cps,t = gβ x,t + Rβ,t
For strong relationships to exist it is necessary that the two serie s have trends
of the same type, i.e., rough, smooth, or general; however this is not
guarantee of a strong relationship.
Even if Cor = 0, it does not necessarily mean that there is no relationship
between the CPS and the State inputs. Keep in mind that both series must be
related given our understanding of what they measure. Only in a restricted
sense does it mean that there is no net linear relationship. By saying that
there is no net linear relationship we mean that the State inputs (CES or UI
series) provide no useful information about the trend in the CPS beyond what
we already know from the historical CPS series.
To determine if a net linear relationship exists, a trend model is fitted to the
noise corrected CPS. This essentially involves correlating the CPS with its
own past values. Next, establish if the State inputs adds any additional
information. If not, there is no net linear relationship. Since non-linear
relations hips may exist, the CES and UI data are used in all State models.

Area Models
In 2005, area models were introduced for the Chicago-Naperville-Joliet, IL
metropolitan division, Cleveland-Elyria-Mentor, OH metropolitan area,
Detroit-Warren-Livonia, MI metropolitan area, Miami-Miami Beach-Kendall
metropolitan division, New Orleans-Metairie-Kenner, LA metropolitan area,
Seattle-Bellevue-Everett metropolitan division. (Model-based estimation of
the New Orleans-Metairie-Kenner, LA, metropolitan area was suspended
following Hurricane Katrina.) Each of these area models is paired with a

August 2010

LAUS Manual 6-16

Balance of State (BOS) model. The BOS is modeled directly rather than
computing it as a residual of a State model less the area model. This is done
because modeling the BOS provides error measures. If the residual approach
is used, all the error would be aggregated in the BOS. In addition, the BOS
CPS data are often more reliable than the area data.
At the area level, univariate models are used. The unemployment model is
multiplicative in form and the employment model is additive. The area
model and the BOS model for a State are controlled to the State model
estimates.

August 2010

LAUS Manual 6-17

Description of the Employment Model
Overview
The basic form for the signal/noise employment model is a regression
equation that uses the monthly CPS employment level as the endogenous
variable and the CES employment level as the explanatory variable. Each
State employment model can be thought of as having a regression equation
form, with a variable coefficient component (CES employment level for the
employment model) and two time series components, which reflect the State's
CPS seasonal and trend movements not accounted for by the CES.
Chapter 4 discusses the conceptual and coverage differences between the
CPS and CES series. While the time series model adjusts for these
differences, knowledge of the CPS/CES differences is important to
understanding the nature of the model's CES variable. The model accounts
for these differences automatically because the regression coefficients,
residual trend and seasonal components are computed separately for each
State using State-specific data. Knowledge of the survey differences is also
useful in understanding why the trends in these two series diverge at times.
Illustrated below is the basic structure of the employment model. The model
consists of three components.
Emp = Emp Signal + Emp Noise
Signal = βCESEM + Trend Residual + Seasonal Residual
Noise = CPS Error, Irregular, Transitory Outliers

Description of Signal Component
CES Base Variable
From the CES survey, a monthly estimate is developed of the total number of
persons on establishment payrolls who received pay for any part of the
employer's pay period that includes the 12th of the month. In the model, the
CES employment is used as the major data source for the model's target: the
employed portion of the labor force. Data for major strikes are added to the
CES estimate from which the employment is calculated.

Time Series Components
The part of the signa l that is unaccounted for by the CES variable is
represented by the residual seasonal and trend components. The trend
component is adjusting for long-run systematic differences between the CES
and the CPS series. Time trend equations with variable intercepts and slopes
are used to estimate the trend. By allowing the parameters of the trend
component to change, the estimated trend component can adapt to change in
the data. If there are frequent changes in the level and/or slope of the trend,
more weight is given to recent observations in estimating the trend.

August 2010

LAUS Manual 6-18

The seasonal component is decomposed from the time series model of the
CPS. As is the case for the trend, the component is allowed to vary to permit
adaptation to changing relationships between the seasonal patterns of the CPS
and CES. Because of definitional differences between the CPS and CES, this
component is necessary. Differences in seasonality between the two series
occur principally because there are large seasonal variations in employed
persons on unpaid absences who are counted as employed in the CPS but not
in the CES, and to seasonal variation in agricultural employment. For a
complete discussion of these differences, see Chapter 4.

August 2010

LAUS Manual 6-19

Description of the Unemployment Model
Overview
The basic form for the signal/noise unemployment model is a regression
equation that uses the monthly unemployment level as the endogenous
variable and the unemployment insurance claims (UI) as the explanatory
variable. Each State model can be thought of as having a regression equation
form, with a variable coefficient component and two time series components
which provide flexibility for the State's CPS seasonal and trend movements
not accounted for by the explanatory variable.
Below is the mathematical representation of the basic structure of the
unemployment model.
The model consists of three variables.
Unemp = Unemp Signal + Unemp Noise
Unemployment Signal = βUI + Trend Residual + Seasonal Residual
Unemployment Noise = CPS Error, Irregular, Transitory Outlier

Description of Signal Component
The Base Variable—UI Claims
The most important variable in the unemployment model is the UI claims
count. This is a measure of the number of workers who are currently
unemployed and receiving UI benefits. Since the CES data are adjusted to
include strikers (CESADJ), the continued claimant count should exclude any
known strikers. (In some States, strikers may be eligible for unemployment
compensation). The statewide estimate of continued claimants without
earnings follows the standards outlined in Chapter 3.
The main weakness of the claims data is that these data are the by-product of
the UI tax system and therefore are subject to changes in the State’s laws,
making them a biased cyclical indicator. Since 1980 there has been a marked
deterioration in the cyclical sensitivity of the claims data. Several factors may
account for this. For example, in the latter stages of a severe recession, many
workers exhaust their UI benefits and are dropped from the count. As a
result, the UI claimant count tends to diverge from the total rate when
unemployment is high and converge with the total rate when unemployment
is low. Nationally, there have been changes in the long-term relationships
between the total unemployment rate and the claims rate. From the 1970’s
through the 1980’s, the proportion of CPS total unemployed collecting UI
benefits nationally dropped from nearly 60 percent to less than 40 percent.
Since the 1980’s, the proportion has generally been between 30 and 40
percent.

August 2010

LAUS Manual 6-20

The seasonal pattern of the claims data also differs in important ways from
the CPS. Most notably, in the summer months, the entry of students into the
labor force is not reflected in the UI data. The seasonal component in the
model controls for this as well as for other seasonal differences. The model
controls for the cyclical bias in the UI data by changing the magnitude of the
regression and trend components. (See Chapter 3 for a discussion of the
differences between the CPS and UI.)

Time Series Components
The part of the signal that is unaccounted for by the claims count is
accounted for by the seasonal and trend variables. Time trend equations with
variable intercepts and slopes are used to estimate the trend. By allowing the
parameters of the trend component to change, the estimated trend component
can adapt to change in the data. If there are frequent changes in the level
and/or slope of the trend, more weight is given to recent observations in
estimating the trend.
The seasonal component is decomposed from the time series model of the
CPS. As is the case for the trend, the coefficients are allowed to vary to
permit adaptation to changing seasonal patterns.

August 2010

LAUS Manual 6-21

Detailed Description of the Estimation Process
Overview
The two approaches to estimation are real-time and historical time. Realtime is a sequential process and makes an estimate for one month at a time
immediately after each new CPS estimate becomes available. Historical
time, on the other hand, is a batch process. Data are accumulated over time
and processed together at once.
The advantage of real-time estimation is that up-to-date estimates are
produced without delay. However, there are some disadvantages. The errors
in the estimates are largest at the end of the series. The trend and seasonal
components appear less smooth. The process requires revisions.
The historical estimate process addresses the disadvantages that occur with
the real-time process. It produces smaller errors because it makes use of all
available information. The revisions are smaller when new data are added.
The trend properties are clearly displayed. The only drawback is that the
estimates are not timely.

Forward Filter
The current estimation procedure uses the forward filter algorithm. To
produce the current month’s estimate the forward filter requires only two
inputs. One is the prior month’s estimates of the signal and noise. The other
input is the current month’s CPS, CES, UI claims and population data. No
other historical data are needed.
The forward filter produces estimates at time t, taking into account
information available at this time. As new CPS data become available after t,
the estimate at t is not revised with new CPS data. States have the
opportunity to update their UI claims and CES data each month, but prior
month estimates are not updated with the latest CPS observation. Thus this
procedure only goes forward and does not look backward.
The forward filter provides estimates without delay. It is computationally
efficient as it requires no more work to process the last observation than it
does the first.

Model Re-Estimation
The model re-estimation algorithm revises the forward filter estimates to
incorporate all data that become available after the estimate for reference
month. It accounts for all information available after time t, i.e., over the
whole sample. Re-estimation is done once a year and requires processing all
of the current year and historical data.

August 2010

LAUS Manual 6-22

The re-estimation process uses all available data and thus is more accurate
than the forward filter procedure. It is less sensitive to the erratic movements
in the CPS. As a result, it provides much smoother model components.

Benchmarking
The purpose of benchmarking is to control for potential bias. The previous
unemployment models were slow to adapt to national shocks, while the
previous employment models tended to overestimate. To address these
limitations, the new generation model estimates are controlled monthly so
they sum to the national CPS. This constraint ensures that model
employment and unemployment estimates will adapt to national shocks
without delay. The national CPS is a relatively reliable benchmark.
Real-time benchmarking adjusts the forward filter estimates to a monthly
control one month at a time. Yearly data are not complete until the last
month. Historical benchmarking adjusts all the values of the smoothed series
at once. This procedure us es complete years of data and is more stable for
the trend.
The procedure to benchmark estimates to the national CPS is comprised of 3
stages. In stage 1 the model estimates are produced for the 9 Census
Divisions and the aggregated series are constrained to sum to the national
CPS. In stage 2 the State model estimates in each Division are summed to
the benchmarked Division model estimate from stage 1. In stage 3 the area
model estimates are summed to the benchmarked State estimate from stage 2.
As with the State models, the Division model structure is based on the
decomposition of a time series into the trend, seasonal, and irregular
components. A model of the CPS survey error component is also added.

August 2010

LAUS Manual 6-23

Benchmark Control Structure
National CPS (Control)

Division Model of the Signal (Stage 1)

State Model of the Signal (Stage 2)

Area Models (Stage 3)

Real Time Benchmarking
The benefits of real time benchmarking to the national CPS are numerous.
The estimates are consistent with reliable monthly national estimates. It
provides protection from national shocks to the economy such as recessions
or catastrophic events like the September 11 terrorist attacks. There is more
consistency between the current year estimates and historical estimates. The
year end revisions are smaller. Error measures are provided. However, this
procedure may introduce additional variability into the current year estimates
due to random fluctuations in the benchmark adjustment made each month.
The two approaches to benchmarking are external adjustments and internal
adjustments. External adjustments are made after the estimation and are
referred to as pro-rata or ratio adjustment. This type is used in the real time
benchmarking application. The internal adjustment occurs during the
estimation. Theoretically it can produce better reliability measures, though it
is difficult to implement and is still being researched.
In the external adjustment, Division model estimates are benchmarked to the
national CPS. The national CPS (USCPS) is divided by the sum of the 9
Division model estimates. This ratio is then applied to each Division model
estimates to arrive at the pro-rated Division estimate (AdjustedD,t).
Step 1
AdjustedD,t = ModelD,t

USCPS

ΣModelD,t
D

The next step is to benchmark the State model estimates to the Division
controls developed in the previous step. The adjusted Division model
estimate (AdjustedD,t) is divided by the sum of the State model estimates for

August 2010

LAUS Manual 6-24

that Division. This ratio is then applied to each State model estimates in the
Division to obtain the pro-rated State estimate (AdjustedS,t).
Step 2
AdjustedD,t
AdjustedS,t = ModelS,t

ΣModelS,t
S

The final step is to pro-rate the area model estimate by controlling it to the
adjusted State model estimate.
Step 3
AdjustedS,t
Adjusteda,t = Modela,t

ΣModela,t
a

In the real time benchmarking procedure, the model estimate that is directly
adjusted is the forward filter not-seasonally adjusted estimate, that is the
trend plus the seasonal components. Implicitly, all the components are also
ratio adjusted by the same factor.
Adjusted trend t = trend t (prfactort )
Adjusted seasonalt = seasonalt (prfactort )
AdjustedD,t
Prfactort =

ΣOriginalS,t
S

Pro-rating the State level maintains relative State model shares in the
Division total. It does not prevent State trends from moving in different
directions. Additionally, larger States get larger absolute adjustment, while
smaller States are not dominated.
AdjustedD,t
AdjustedS,t = ModelS,t

ΣModelS,t
S

ModelS,t
=

ΣModelS,t

AdjustedD,t

S

Benchmarked model estimates may be somewhat more variable then the
original model estimates. The proportionate change in the original model

August 2010

LAUS Manual 6-25

estimates are not exactly preserved due to fluctuation in the benchmark
adjustment factor from month-to-month. Only if the adjustment factors are
equal to a constant (k) over time, will the proportionate change in the original
model estimates be exactly preserved.
AdjustedD,t

=k

ΣOriginalS,t
S

AdjustedS,t
AdjustedS,,t-1

k originalS,t
=

k originalS,t-1

originalS,t
=

originalS,t-1

For seasonally-adjusted estimates, an additional step is employed at this time
to reduce the volatility produced by the application of the pro-rata factors.
This step is called Smoothing and is discussed in an upcoming section.

August 2010

LAUS Manual 6-26

Official Estimates
This benchmarking procedure does not eliminate the need for end-of-year
revisions; however, it does reduce the size of the revision compared to the
previous method. The smaller annual revisions to the real time model leads
to less-over-the year distortion and facilitates analysis of the estimates
between historical and current year estimates.

Historical Benchmarking
The end-of-year processing involves entering the revised UI claims, CES
employment, and population values and re-estimating the forward filter. The
resulting smoothed forward filter estimates become the benchmarked
smoothed estimates and replace the real time benchmarked estimates.

End-of-Year Processing Steps
Revise Pop, CPS, UI, & CES
?
Make historical estimates
?
Benchmark historical estimates
?
Replace concurrent estimates
In the ‘benchmark historical estimates’ step, the first stage is to benchmark
the Division models. The control total is the not-seasonally-adjusted national
CPS estimate of employment or unemployment. Then, develop the historical
not-seasonally-adjusted model estimate. Next, re-compute the pro-rata
factors for not-seasonally-adjusted estimates. Benchmark both the notseasonally-adjusted and seasonally- adjusted estimates using the new pro-rata
factors.
The second stage is to benchmark the State model estimates of employment
and unemployment. The control total is the not-seasonally-adjusted Division
estimates of employment and unemployment developed in stage 1.Then, recompute the pro-rata factors for not-seasonally-adjusted historical model
estimates. Then, benchmark both the not-seasonally- adjusted and the
seasonally-adjusted estimates using the new pro-rata factors.

August 2010

LAUS Manual 6-27

Pro-rata factor for Historical State NSA & SA Series
Bmk NSA Historical Div Mdldt
prfactor dt

=

___________________________________

Sum of NSA Historical State Model dt
Monthly benchmarking of the seasonally-adjusted estimates does eliminate
large revisions due to end-of-year trending but the resulting volatility must be
addressed.

August 2010

LAUS Manual 6-28

Smoothed-Seasonally-Adjusted Estimates
There are a number of sources of volatility in the LAUS estimates. These include
sampling error in CPS and outliers, real time monthly benchmarking, seasonality,
uncertainty at the end points of the series, frequent level shifts in the trend, and real
outliers in the current year (for example, Hurricane Katrina). Volatility is a problem for
model estimation when month-to- month change is unexplained, not related to
predictable survey error or seasonal patterns, lacks persistence, and is difficult to explain
in terms of long-run movements. Major sources of volatility in the models can be
controlled for (at least partially). Normal survey error behavior is controlled with the
survey error model. One-time outliers have potential solutions—either through end-ofyear smoothing or the outlier regression model. Seasonality, a significant source of
volatility, is removed with seasonal adjustment.
The principal volatility that we want to try to control for is that which arises from realtime monthly benchmarking. One approach to smoothing seasonally-adjusted
benchmarked estimates is through the use of moving averages, or filters. Symmetric
averages “move” through a time series from period to period by shifting the time periods
to be included forward by 1 period. The center of a moving average is at time t, the
point being smoothed. The weights of the other time periods add to one. A simple
moving average applies the same weight to all time periods. A weighted moving
average gives more weight to central observations. Asymmetric moving averages are
utilized for time periods with no future observations (for example, current time). The
center of an asymmetric moving average is at time t-1, the point being smoothed is at
time t. A simple average lags the values by 1 period. A weighted average gives more
weight to the value in the current period, making it more responsive to change.
Smoothing methods don’t work as well at the end of a series as they do in the middle.
Time-series models involve averaging over time. At the end of series we must put more
weight on a relatively few number of observations When we move toward the center of
the series, the weights are spread more evenly over a larger number of observations since
we have data following the time point for which we make an estimate. It is always more
difficult to tell what is happening at the end of a series. We have only past data and are
missing information from future data. We are less certain whether movements are due to
trend, seasonal, or irregular components. It is only well after the event that we can form
a clear view as to what has happened to the trend.
For the LAUS time series models, we need a set of filters to smooth all of the points in
the series. We begin by designing a symmetric filter with “good” properties for
historical series. Then, we derive asymmetric filters that converge to the symmetric as
more data become available and minimize filter revisions. This results in a “family of
related filters” consisting of a given symmetric filter and all of the necessary asymmetric
filters needed when there are not enough data points for the symmetric filter.
The Henderson 13 term Trend Filter family is the filter utilized in LAUS smoothing. Its
symmetric filter is relatively short. The half length is 6, which means it needs only 6

August 2010

LAUS Manual 6-29

months of future data. It is a fairly effective smoother, and generally, results in little
distortion to the trend cycle. The Henderson asymmetric filter requires 6 terms. It does
not over smooth the trend cycle or lag turning points. It reduces volatility in the
benchmarked seasonally-adjusted estimates.
Six different filters are used to smooth the historical series:
Estimates for last
year
Up to June
July
August
September
October
November
December

Length
13
12
11
10
9
8
7

Number of past
observations
6
6
6
6
6
6
6

Number of future
observations
6
5
4
3
2
1
0

For real-time (concurrent) estimation, we use a single asymmetric filter (one-sided)
moving average. The filter length is 7 – 1 current observation and 6 previous values
looking backward. The Henderson asymmetric filter does not preserve turning points
due to irregular variation in the series. Much of this variation is noise. If one-time blips
are real economic events, we will treat them as outliers.
Use of the SSA methodology provides more continuity between current and historical
estimates. For example, the SSA weighted average estimates for January of the current
(production) year overlap the SSA weighted average for December of the prior year
(which has been annually revised). There are 5 data points in common for the January
and December estimates.

August 2010

LAUS Manual 6-30

Error Measures
In general, point estimates are never 100 percent
accurate. To convey the limitations of the data to our
data users it has been the Bureau of Labor Statistics’
policy to publish error measure if the methodology
permits. With the introduction of the new generation
of models in 2005, we are now able to publish error
measures.
There are two uses of error measures. One is reliability which gives us an
idea of how far are the estimates from the truth. The second is for ana lysis by
giving us an idea of what we can say about the truth.
Standard Error
The standard deviation of errors in the estimates gives a measure of the
dispersion of the error around a mean of zero. The larger the standard
deviation (Stder), the more likely an individual estimate is far from the true
values.
Example: A point estimate of 238,000 persons may have a Stder of 25,600
persons.
Coefficient of Variation
From the Stder other error measures that facilitate analysis can be computed.
The coefficient of variation (CV) is a common reliability measure that is
useful for comparing the estimates of different size or scale. The CV is
computed by dividing the standard error by the estimate.
CV =

Stder

=

Estimate

25,600
238,000

= 0.11

Confidence Intervals
The Stder is also used to construct confidence intervals. A confidence
interval for an estimate give us upper and lower limits around the estimate
where the true value likely to be located with a given level of certainty or
confidence.
Estimate ± k Stder
If k = 1.96, the significance is at a 95 percent confidence level. If k = 1.645, it
is at a 90 percent level.

August 2010

LAUS Manual 6-31

Significance
These error measures can be used to determine the difference between a State
estimate and the estimate for the nation or another State is statistically
significant. It is also used to reveal if the over-the- month change is
significant.
Estimate - Mean
z=
Stder
If z = 1.96, then the difference is significant at the 95 percent level. If z =
1.645, then it is significant at the 90 percent level.
Error measures assist the analysis of month-to- month change, the differences
from the US and other State estimates and reliability.
Availability of Measures
Error measures for both smoothed-seasonally-adjusted and not-seasonallyadjusted estimates are available monthly in the STARS tables. Table 2
provides standard errors (Stder) for the point estimates, while Table 3
provides standard errors for over-the-year change, and Table 7 provides
standard errors for over-the-month change.

August 2010

LAUS Manual 6-32

LAUS Estimation:
Labor Market Area Estimates

7
Introduction

I

n the late 1940's, when sub-national labor force estimation was first attempted,
employment and unemployment estimates were developed for large labor
market areas as well as States, underscoring the importance of substate labor
market information. Today, the LAUS program creates estimates for 2,352 Labor
Market Areas (LMAs) that exhaust the geography of all States, the District of
Columbia, and Puerto Rico.

UI
Claims

Exhau stees

Non-Ag
Wage &
Salary

Entrants

All Other
Emp

Ag
Emp

Estimates for most LMAs are produced
independently by a building block approach,
which uses current unemployment insurance (UI)
data and current nonfarm employment estimates
as basic inputs. In addition, components of the
labor force not covered by the basic source data
are developed using larger-area and decennial
Census relationships. This methodology is
referred to as the Handbook procedure.

When the Handbook methodology was first introduced as a standard procedure
for sub-national labor force estimation in 1960, it was viewed as an attempt to
approximate the results of a CPS-type household survey, but without the
prohibitive cost of conducting such a survey. The Handbook method utilized a
system of estimates that was reflective of the labor market structure of the 1960’s
in terms of UI coverage. Over the years, refinements were made to the
components and basic input data to improve comparability and consistency within
the States and with the standard definitions of the labor force as embodied in the

November 2010

LAUS Program Manual 7-1

CPS. In addition, the Handbook procedure has been streamlined to reflect
expanded UI coverage and economic and behavioral changes in the labor market.
Today, the Handbook methodology consists of 16 line items that can be broken
out into employment and unemployment estimation procedures. This chapter will
provide details for each Handbook line and the associated inputs entered into the
LAUS State System Plus (LSS Plus) software. (The LSS Plus Users’ Guide is
available on EUS Web at \\Eusdr1\lnsoutbox\LSS PLUS Manual v2010.1.0.pdf.)

Handbook Line Items
Employment
Line

Description

1

Non-agricultural Wage & Salary Employment

2
3

All-other Employment
Agricultural Employment

4

Total Handbook Employment (lines 1 + 2 + 3)
Unemployment

Line

Description

5
6

UI Claims
UCFE Claims

7
8

Rail Road Claims
Total Claims (lines 5 + 6 + 7)

9
10
11

Unemployed Exhaustees
Non-covered Agricultural Unemployment
Unemployed excluding Entrants (lines 8 + 9 + 10)

12
13

Re-entrants Ratio
Re-entrants

14
15
16

New Entrants Ratio
New Entrants
Total Unemployment (lines 11 + 13 + 15)

November 2010

LAUS Program Manual 7-2

Additivity
Prior to 1977, the Handbook estimates were the final LAUS estimates for LMAs.
Beginning in that year, additivity of the substate Handbook estimates to statewide
modeled estimates was introduced to address methodology issues and Federal program
allocation needs.
The sums of Handbook employment and unemployment for all LMAs in a State tend to
be lower than statewide estimates due to the greater difficulty in obtaining some of the
input data elements at the substate level. Forcing the LMA estimates to sum to the
statewide totals corrects for any methodological deficiencies in a proportional manner,
allowing for the complete, to-the-dollar distribution of federal funds to areas when LAUS
data are used in the allocation algorithm.
Additivity is considered a separate methodological step that follows Handbook
estimation. (See Chapter 9 for more details.)

November 2010

LAUS Program Manual 7-3

Labor Market Area Employment
Employment comprises all persons who did any work at all as paid employees, worked in
their own business, own profession, or on their own farm, or who worked 15 hours or
more as unpaid workers in an enterprise operated by a member of the family. It also
includes all those who were not working but who had jobs or businesses from which they
were temporarily absent because of vacation, illness, bad weather, labor- management
dispute, job training, child-care problems, maternity or paternity leave, or other family or
personal reasons, whether or not they were paid for the time off or were seeking other
jobs. (See Chapter 2 for more details.)
The Handbook method decomposes employment into three subcategories. The following
table provides a brief summary of the employment inputs entered into LSS Plus for each
Handbook line. The sections following the table discuss each item in greater detail.

Handbook Employment
Input description
(LSS Plus variable ID)

Input
Source

Line

Line Description

1

Non-agricultural
Wage and Salary
(NAWS)
Employment

DRRs (C01)

BLS

Establishment-based NAWS
(M01)
Labor- management disputants
(M02)

State

All-other
Employment

Census all-other Employment
(C02)
Step 3 Ratio Stratum (C03)

BLS

+ 2

Establishment-based NAWS
decennial base (C04)
Step 3 Ratios (S01 – S03)
+ 3

= 4

November 2010

Agricultural
Employment

Census Agricultural
Employment (C05)
Agricultural Employment
Change Factors (G01 – G21)
Total Handbook Employment

State

BLS
State
BLS
BLS
BLS

LAUS Program Manual 7-4

Non-agricultural Wage and Salary Employment
(Handbook line 1)
The Handbook calculation of residency-adjusted non-agricultural wage and salary
employment begins with input data that pertain to jobs by place of work (establishment
data) rather than employed people by place of residence (household data). The
conceptual differences between (1) the establishment-based inputs entered into LSS Plus
and (2) the household-based output desired require calculations to adjust the inputs to
conform to CPS concepts. The Handbook line 1 calculations apply Dynamic Residencyadjustment Ratios (DRRs) to the establishment-based input data to bridge the conceptual
gap.
The States provide two inputs that are entered into LSS Plus each month for the line 1
calculations:
•

Establishment-based Non-agricultural Wage and Salary Employment
(NAWS)
o Also referred to as “place-of-work NAWS” or “pre-adjusted NAWS”
o LSS Plus variable ID M01

•

Labor-Management Disputants
o Also referred to as workers involved in “work stoppages”
o LSS Plus variable ID M02

BLS provides one input that is entered into LSS Plus once a decade for the line 1
calculations:
•

Dynamic Residency-adjustment Ratios (DRRs)
o LSS Plus variable ID C11

The following sections will provide details regarding the sources and development of
these inputs.

Handbook line 1 input: Establishment-based Non-agricultural Wage and
Salary Employment (M01)
For most States, there is no single source of M01 data. Obtaining the M01 inputs for all
LMAs in a State usually requires the use of various data sources. The principal source is
the Current Employment Statistics (CES) survey. For those LMAs that are not within the
CES program’s scope, a sample-based employment series developed under State auspices
is the next best data source. If such a series is not available, a number of ways to produce
the input can be used. Details for all data sources and procedures are provided below.

November 2010

LAUS Program Manual 7-5

Current Employment Statistics (CES) data
The CES survey, which is also referred to as the “payroll survey” or “establishment
survey,” yields employment estimates for all metropolitan areas and most metropolitan
divisions. It is the principle source of the M01 input and sets the conceptual standard for
M01 values derived from other sources. The CES not-seasonally-adjusted estimate of
total nonfarm wage and salary employment should be used for LAUS estimation
wherever it is available. No adjustment of CES data is necessary to conform to the
definition of the LAUS M01 input.
The CES program conducts monthly revisions (data for the prior month are revised with
the release of the current month) and annual benchmarking revisions (data for the latest
18 months are revised at the end of each year). The LAUS program incorporates these
revisions during monthly and annual processing, respectively. Historic CES corrections
that extend beyond the standard two-year LAUS annual processing period are
incorporated if substantially large.
Quarterly Census of Employment and Wages (QCEW) data
The QCEW program publishes counts of establishments and jobs and aggregate wage
data that are derived from quarterly tax reports. The reports are submitted to State
employment security agencies by employers that are subject to either (1) State UI laws or
(2) the Unemployment Compensation for Federal Employees (UCFE) program. Each
quarter, QCEW releases monthly data. Data for any given year are not final until data for
the first quarter of the following year are released.
LAUS estimation generally occurs with only a one- month lag from the reference period.
In contrast, the volume of data processed by the QCEW program requires a longer lag of
6 to 9 months between the reference quarter and the release of data. To reconcile the
timely needs of the LAUS program with the longer lag of the QCEW program,
extrapolation is used for monthly processing. During annual processing, actual QCEW
data replace extrapolated data.
Adjustments made to both extrapolations of QCEW data and actual QCEW data are
discussed below, followed by details of the extrapolation process.

November 2010

LAUS Program Manual 7-6

Adjustments to QCEW data
Several steps are necessary to adjust QCEW job counts to conform to the definition of the
LAUS M01 input. Essentially, these adjustments bring QCEW data into alignment with
the CES data used for metropolitan areas.
1. Start with the QCEW total job count for the labor market area.
2. Subtract agricultural jobs (NAICS codes 111, 112, 114, and 115).
o These are accounted for in the Agricultural Employment estimate (line
3).
3. Subtract private household workers (NAICS 814).
o These are accounted for in the All-other Employment estimate (line 2).
4. Add Presumed Non-covered Employment (PNC).
o This includes various categories of jobs that are not covered by
unemployment insurance laws, but which are included in CES
estimates. These include:
•
•
•

College students working for their schools,
Commissioned insurance agents, and
Religious employees.

o Most States have a periodic survey that estimates PNC at the State and
metropolitan area level. The LAUS technician should consult with the
CES unit in their State to obtain PNC estimates for labor market areas.
5. Add estimates of railroad employment.
o These estimates are provided to States by the BLS national office.
Extrapolation of QCEW data
Due to the time lag associated with publication of QCEW data, extrapolation is necessary
for LAUS monthly processing. Two approaches exist:
1. Extrapolation using Area-based Change Factors
•

Historic over-the- month change s for each reference month are applied
to the prior month’s estimate.

•

The historic monthly changes can be those for the prior year or
averages of those for several prior years.

•

For example, to estimate March 2010, take the over-the-month QCEW
employment change from February 2009 to March 2009 and apply that
factor to the February 2010 estimate.

2. Extrapolation using Area Shares of the non-CES Balance of State
•

November 2010

Historic ratios of area QCEW employment to the QCEW employment
of the non-CES balance of State are applied to current CES balanceof-State estimates to yield area estimates.
LAUS Program Manual 7-7

•

The historic area shares can be those for the prior year or averages of
those for several prior years.

•

The CES balance of State estimate is found by subtracting all MSAs
within the State from the statewide estimate.

BLS ARIMA Forecasts
BLS produces nine-month forecasts of QCEW data for labor market areas for all States.
The forecast program utilizes historical monthly QCEW data for each labor market area
in a State. Once the most recent quarter of QCEW data are available to national office
staff, ARIMA software is run producing employment forecasts for each area. States need
to add PNC and railroad estimates to these forecasts before utilizing them as inputs to
LAUS employment estimation.
Small Domain Estimators
Illinois uses the National Opinion Research Center Small Domain Estimator model to
produce non-agricultural employment by industry for their non-CES areas. The model
utilizes CES sample, QCEW employment, PNC estimates, allocation of employment for
statewide reporters, and economic events not captured in the CES sample. At the end of
each year, the model estimates are benchmarked to QCEW employment.
Pennsylvania uses a Small Domain Estimator that is based on the inverse relationship
between unemployment insurance claims and employment—when workers get laid off,
claims go up and employment goes down; when workers return to work, claims go down
and employment goes up.
CES ACES System
Some States utilize features within the CES ACES estimation system to produce
employment estimates for labor market areas not included in the CES program.
Basically, this model calculates the change in employment for each sample cell and
applies that change to the prior month’s estimate.
Handbook line 1 input: Labor-Management Disputants (M02)
In addition to the non-agricultural wage and salary employment input (M01), data for
workers absent from their jobs due to labor- management disputes are needed. Workers
involved in these disputes can be either on strike (a work stoppage initiated by the
workers of an establishment) or lockout (a work stoppage initiated by the management of
an establishment). In both cases, the workers have jobs from which they are temporarily
absent and are therefore considered employed under the CPS definition. (See Chapter 2.)
Counts of workers involved in labor disputes during the reference week are entered into
LSS Plus for each Handbook area using the variable ID M02.

November 2010

LAUS Program Manual 7-8

Data on labor- management disputes can be obtained from the CES Strike Report and the
BLS Work Stoppages Program.
Handbook line 1 input: Dynamic Residency Ratios (C11)
The input data detailed above pertain to workers by their place of
employment. The CPS definition of employment pertains to employed
individuals by their place of residence. Because the goal of LAUS is to
parallel CPS concepts, the establishment data gathered and entered into
LSS Plus must undergo adjustment.
While there are several differences between establishment data and household data (see
Chapter 4), the largest source of difference at the LMA level is the discrepancy between
the location of establishment s, where jobs are counted, and the location of residencies,
where employed individuals live. The Dynamic Residency Ratios (DRRs) adjust the
establishment-based inputs to a place-of-residence, or household, basis.
Development of Residency-adjustment Methodology
Prior to 2005, residency adjustment of the establishment-based M01 and M02 inputs was
accomplished by a single ratio for each LMA. The ratio was calculated from 1990
Census data by dividing the number of residents employed in non-agricultural wage and
salary jobs within each LMA (obtained from Census data) into the total number of wage
and salary jobs in the LMA at the time of the Census.
Beginning in 2005, the single-ratio approach was replaced with the introduction of DRRs.
The general concept behind the DRR methodology is that an LMA’s resident
employment is a function not only of the jobs available within the LMA (the pre-2005
approach), but also of the jobs available in neighboring LMAs.
Commutation data from the 2000 Census were used to determine the appropriate
neighboring LMAs to include in the residency adjustment calculations of each area. The
largest commuter areas for each LMA were identified and, to reduce the complexity of
the calculations, the number of commuter areas was capped at four for each LMA. A
minimum of 100 commuters was set for a commutation area to be included (for New
England, the minimum was lowered to 50 commuters).
The DRR inputs and calculations are detailed below.
Inputs for DRR calculations:
1. Census 2000 commutation data
•

County-to-county (or, in New England, MCD-to-MCD) commuter flows from
the Census are aggregated to the LMA level.

•

The following are incorporated into the DRRs for each LMA:
§ Commuters residing and working within the same LMA, regardless of

November 2010

LAUS Program Manual 7-9

level of commuters, and
§ The four largest commuter flows to neighboring LMAs, where the
level of commuters is 100 or more (or, in New England, 50 or more).
2. Census 2000 Non-agricultural Wage and Salary Employment (Census NAWS)
•

Total Census employment is obtained (from Table P43 of Summary File [SF]
3) and the following are subtracted:
§ Agricultural workers (Table P51 of SF3),
§ Self-employed workers in own not incorporated business (Table P51),
§ Unpaid family workers (Table P51), and
§ Private household workers (Table PCT85 of SF4).
3. Establishment-based NAWS Employment decennial base (C04)
•

M01 and M02 data for March and April 2000 (the time of the Census) are
averaged.

The following equation displays the DRR calculations for LMA1 :
Ratio 1:
Employed residents LMA1 working in LMA1
Nonfarm employment Mar/Apr 2000 LMA1
Ratio 2:
Employed residents LMA1 working in LMA2
Nonfarm employment Mar/Apr 2000 LMA2
Ratio n:
Employed residents LMA1 working in LMAn
Nonfarm employment Mar/Apr 2000 LMAn

November 2010

X

X

Census NAWS1
Total Commuters1+2+…+ n
Census NAWS1
Total Commuters1+2+…+n

Census NAWS1
X
Total Commuters1+2+…+n

LAUS Program Manual 7-10

The following table displays an example DRR calculation based on the equations above:
(A)

(B)
Census 2000
Nonagricultural
Wage and
Salary Emp.

Area

Area of
Residence
Commuter
Area 1
Commuter
Area 2
Commuter
Area 3
Commuter
Area 4

(C)

(D)

Census 2000
Commuters
from Area of
Residence

Est.-based
NAWS Emp.
decennial
base (C04)

235,584

305,400

0.977555

0.754081

13,370

127,450

0.977555

0.102549

7,345

66,950

0.977555

0.107246

2,891

25,960

0.977555

0.108864

2,830

32,745

0.977555

0.084486

256,139

(E)
Control ratio
= [(B) for
Area of
Residence] / S
(C)

(F)
DRR (C11)
= (C) / (D) *
(E)

The following table displays an example Handbook line 1 calculation using the
DRRs from the table above:
(A)

Area

Area of
Residence
Commuter
+
Area 1
Commuter
+
Area 2
Commuter
+
Area 3
+

Commuter
Area 4

(B)

(C)

(D)

DRR
(C11)

Establishmentbased NAWS
Employment
(M01)

Labormanagement
disputants
(M02)

0.754081

311,508

0

234,902

0.102549

121,105

200

12,440

0.107246

73,645

0

7,898

0.108864

26,220

0

2,854

0.084486

33,072

0

2,794

= Line 1 for Area of Residence

November 2010

(E)
Residencyadjusted line 1
component
= (B) * [(C) +
(D)]

260,888

LAUS Program Manual 7-11

All-other Employment (Handbook line 2)
All-other employment includes the following types of workers that are not employed in
agriculture:
(1) The self-employed, who work in their own not- incorporated business,
(2) Unpaid family members, who work for a business owned by a family member, and
(3) Private household workers (or “domestics”).
These people are employed by the CPS definition and are the second largest category of
total employment behind non-agricultural wage and salary workers.
Two sources of all-other employment data exist--the decennial Census and the CPS. The
Census provides more geographic detail while the CPS is available on a monthly basis.
Total all-other employment is calculated using CPS estimates of all-other employment,
CES estimates of nonagricultural wage and salary data, and Census counts of area
nonagricultural wage and salary employment and all-other employment data. The Census
counts are used as a base period estimate.
During intercensal years, data for all-other employment are available from the CPS.
While these data are published on a monthly basis only for the nation as a whole,
unpublished data are available at the State level. Research has shown that the State CPS
data, together with State and area wage-and-salary employment, can be used to
extrapolate the Census all-other decennial benchmark for LMAs.

Development of All-other Employment Methodology
The original analysis which led to the first estimating methodology was based on an
examination of the relationship between all-other employment and wage-and-salary
employment in the Nation as a whole and in a randomly selected sample of areas using
the 1940 and 1950 Census data. It was found that, in both the areas and the U.S. total, the
relative change in wage-and-salary employment was accompanied by a proportional
relative change in all-other employment. In other words, slow wage and salary growth
was accompanied by slow all-other employment growth, and rapid wage and salary
growth was accompanied by rapid all-other growth.
It was also found that the proportional relative changes in all-other employment in the
areas and in the Nation were very close to each other. This meant that the relative change
in area all-other employment could be derived given the relative change in area wageand-salary employment and the ratio of the relative national change in all-other
employment to the relative national change in wage-and-salary employment.
Analyses utilizing data from subsequent Censuses corroborated the findings of the
original study. However, discrepancies between individua l areas, on the one hand, and
areas and the Nation on the other, proved quite common, and pointed out the need for
area adjustment. The CPS sample expansion of the 1970's provided additional
November 2010

LAUS Program Manual 7-12

geographic detail on all-other employment and allowed the opportunity for analysis and
testing of differences in the proportionality factor between States.
Following each Census, the relative change in wage and salary employment divided by
the relative change in all-other employment, is calculated and reviewed. Clusters of
States with similar proportionality constants are grouped into strata. Four strata were
defined following the 1980 Census and three were defined following the 1990 and 2000
Censuses. By grouping States into strata based on their ratio of relative change, it was
found that LMA all-other employment estimates could be improved. Specifically, using
the proportionality factor for State-based strata to estimate the all-other employment for
LMAs significantly reduced the range of error in estimating all-other employment.

Assigning Step 3 Ratio Strata (C03)
Following the 2000 Census, the proportionality constant, k, was used to assign States to
strata. Once the State-based strata were established, LMAs were assigned to the strata
using the same calcula tion of k.
The strata assigned remain in use for the entire intercensal period and determine the Step
3 ratio used in the Handbook calculations.
w t ÷ wt-1

k = at ÷ at-1
Where:
Variable
k
w
a
t
t-1

Description
Proportionality constant
March/April 2000 Average Non-agricultural
Wage and Salary Employment
Census All-other Employment
Current Census time period (March-April 2000)
Prior Census time period (March-April 1990)

Strata from the 2000 Census by k value
Stratum 1

Stratum 2

Stratum 3

k < 1.125

1.125 < k < 1.432

k > 1.432

The numeric stratum identifiers (1, 2, and 3) are used to identify each LMA’s
stratum in LSS Plus (variable ID C03). The system uses the stratum identifiers to
determine the appropriate Step 3 ratio to use in calculations.
Calculating Step 3 Ratios (S01 – S03)

November 2010

LAUS Program Manual 7-13

Each month, BLS calculates Step 3 Ratios using CPS and CES data for the States
that compose each stratum. The calculated ratios are provided to the States and
entered into LSS Plus. The following three steps are used to calculate the ratios.
Step 1.

Determine the all-other employment change ratio. The current
month CPS all-other employment estimates for each State in the
stratum are summed and divided by the sum of the March/April 2000
Census all-other employment estimates for the States in the stratum.

Step 2.

Determine the nonagricultural wage and salary employment
change ratio. The current month CES nonagricultural wage and
salary estimates for each State in the stratum are summed and divided
by the sum of the March/April 2000 (two-month average)
nonagricultural wage and salary estimates for the States in the
stratum.

Step 3.

Determine the ratio of relative change for each stratum. Divide
each stratum all-other employment change ratio by the corresponding
stratum wage and salary employment change ratio, i.e., Step 1 divided
by Step 2.

The resulting ratios are assigned the LSS Plus variable IDs of S01, S02, and S03
to correspond with strata (C03 values) of 1, 2, and 3, respectively.
Table 7-1 presents monthly Step 3 ratios by stratum.
Calculating Current All-other Employment
Once LMAs have been assigned to strata (once a decade) and Step 3 ratios have
been calculated (every month), Handbook calculations will use the March/April
2000 Average Non-agricultural Wage and Salary employment value (C04), the
establishment-based employment inputs for Handbook line 1 (M01 and M02), and
Census All-other Employment (C02) to calculate current all-other employment
(Handbook line 2). The following formula displays the line 2 calculation for each
LMA:

=

M01 + M02

x (C02) x (Step 3 Ratio)

C04

November 2010

LAUS Program Manual 7-14

Detailed explanations of the formula and an example of the calculation:
1) Calculate the non-agricultural wage and salary employment change ratio.
Divide the current month non-agricultural wage and salary employment,
including persons involved in labor- management disputes, by March/April
2000 nonagricultural wage and salary employment.
Non-ag. wage & salary employment (M01 + M02)

933,100

÷ March/April 2000 non-ag. employment (C04)

938,107

Area employment change ratio

0.994663

2) Calculate current all-other employment based on the unadjusted growth
rate of non-agricultural wage and salary employment. Multiply the ratio
calculated above (the non-agricultural wage and salary employment change
ratio) by Census All-other Employment.
Area employment change ratio
Census all-other employment (C02)

0.994663
53,589

Unadjusted All-other employment

53,303

×

3) Calculate all-other employment for the current month. Multiply the
estimate calculated above (unadjusted all-other employment) by the stratumspecific Step 3 ratio provided by BLS. This will adjust for the differing rates
of growth between non-agricultural wage and salary employment in the
stratum (based on CES data for the States in the stratum) and all-other
employment in the stratum (based on CPS data for the States in the stratum).
All-other employment change

×

Step 3 ratio (S01, S02, or S03 depending on C03)

All other employment (Handbook line 2)

November 2010

53,303
0.894000

47,653

LAUS Program Manual 7-15

Agricultural Employment (Handbook line 3)
Unlike the non-agricultural Handbook employment estimates, which split
employment by class of worker—wage and salary (line 1) and “all-other” (line
2)—the agricultural Handbook employment estimate encompasses all classes of
worker—wage and salary, self-employed, and unpaid family—in a single
estimate. This is accomplished by applying a monthly change factor to a
decennial base of agricultural employment obtained from 2000 Census data. The
following formula shows the calculation for each LMA.
L03 = ( C05 ) x ( Change factor )
Where:
Variable

Description (LSS Plus Variable ID)

L03

Handbook Agricultural Employment

C05

Census Agricultural Employment (C05)

Change
factor

Agricultural Employment Monthly
Change Factor (G01 – G21)

Development of the Agricultural Employment Methodology
Prior to the incorporation of 2000 Census data into the Handbook methodology,
the procedure for agricultural employment estimation utilized information from
the 1990 Census, the Current Population Survey (CPS), and the Department of
Agriculture’s Farm Labor Survey (FLS). As of 2002, the FLS ceased to provide
information for all farm workers and began limiting its quarterly publication to
information for hired workers only. Because hired workers account for only 35 to
50 percent of all agricultural workers, the FLS data became an inadequate
benchmark for Handbook agricultural employment estimation.
To be congruent with the CPS definition of employment, the self-employed and
unpaid family workers must be included in addition to hired workers. Because of
this, FLS data are no longer used. Currently, unpublished monthly estimates of
agricultural employment from the CPS are used in lieu of FLS data.
Agricultural Regions
The Agriculture Department, through the FLS, designated twenty-one estimating
regions. Fifteen of the regions were creating by grouping States that have similar
agricultural activities, while six others each comprise only one State. Though
LAUS no longer uses FLS data, the Handbook methodology continue s to utilize
the FLS agricultural regions. The regions are listed in the following table.

November 2010

LAUS Program Manual 7-16

Region
Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21

Agricultural
Region
Northeast I
Northeast II
Appalachian I
Appalachian II
Southeast
Florida
Lake
Corn Belt I
Corn Belt II
Delta
Northern Plains
Southern Plains
Mountain I
Mountain II
Mountain III
Pacific
California
Hawaii
Michigan
Minnesota
Wisconsin

State(s)*
CT, ME, MA, NH, NY, RI, and VT
DE, MD, NJ, and PA
NC and VA
TN and WV
AL, GA, and SC
FL
MI, MN, and WI
IL, IN, KY, and OH
IA and MO
AR, LA, and MS
KS, NE, ND, and SD
OK and TX
ID, MT, and WY
CO, NV, and UT
AZ and NM
OR and WA
CA
HI
MI
MN
WI

* Alaska, the District of Columbia, and Puerto Rico are not included in any agricultural
estimating region.

The States have the option to make use of data for agricultural regions other than
their own. This may be done if local knowledge of the agricultural economy
indicates that another region better reflects the State’s agricultural employment.
(In the case of interstate areas, the calculation for the whole area is determined by
the controlling State's selection.) Once the selection of an alternate regional
factor has been made, it must be continued until the next decennial Census. The
production of all current and benchmarked data must also reflect the selection.
States must request the use of alternative agricultural factors by means of the
atypical/exception request procedure.
Census Agricultural Employment (C05)
The 2000 Census incorporated the North American
Industry Classification System (NAICS), which replaced
the Standard Industrial Classification (SIC) system. Data
from the 2000 Census for agricultural employment include
logging as part of a subsector of agriculture (Forestry and
Logging, NAICS 113), whereas logging had been classified
as a manufacturing activity in the now-defunct SIC system. Numerous other
November 2010

LAUS Program Manual 7-17

differences between the two systems exist. For LAUS purposes, all employment
in the agriculture industry (NAICS 11) is included in the decennial base estimates
of Census agricultural employment. The five NAICS industry subsectors that
comprise the agricultural sector are listed in the table below. Agricultural
employment is obtained from the 2000 Census via Table P51 of Summary File
(SF) 3.
NAICS Subsector

Subsector Title

111
112
113
114
115

Crop Production
Animal Production
Forestry and Logging
Fishing, Hunting, and Trapping
Support Activities for Agriculture and Forestry

Agricultural Employment Monthly Change Factors (G01 – G21)
A change factor created from the CPS annual average of the current year over the
annual average of the prior year is used at benchmark time to rebase to the
decennial Census. This ratio is then multiplied by the annual change factor from
the previous year in order to link the change factor to the July to July benchmark
period and to move the decennial Census number forward. July is used as a base
month because it is the most agriculturally active month for the majority of States.
The annual change factor is the CPS annual average of the recently completed
year (y) over the CPS annual average of the prior year (y-1) multiplied by the
annual change factor from the previous year.

Annual change
factor

=

AACPS (y)
AACPS (y-1)

Annual change
factor (y-1)

These annual change factors form the base for all agricultural employment
estimation until the next decennial Census update. New monthly change factors
are provided following each year’s benchmarking activity to revise monthly
estimates from July of the previous year forward. The new July base is then used
until the next year's benchmark. The preliminary August-December estimates are
replaced with revised estimates, based on the new July adjusted base, providing a
consistent series through the next calendar year.
The current production month factor is created by applying the annual change
factor to a ratio of the reference month State CPS data over the July CPS of the
previous completed year agricultural number.

November 2010

LAUS Program Manual 7-18

Monthly factor = Annual change factor

Ref month CPS
July CPS

Each month the current change factor produced using the above formula is
applied to the LMA agricultural employment estimate from the 2000 Census to
arrive at the current month’s total agricultural employment estimate for the LMA.
Census LMA agricultural employment

× monthly factor
Total LMA agricultural employment

November 2010

1,820
1.037070

1,887

LAUS Program Manual 7-19

Labor Market Area Unemployment
Unemployment comprises all persons who do not have a job, have actively looked
for work in the prior four weeks, and are currently available for work. Persons
who were not working and were waiting to be recalled to a job from which they
had been temporarily laid off are also included as unemployed.
Receiving benefits from an Unemployment Insurance (UI) program has no
bearing on whether a person is classified as unemployed; however, statistics from
the UI system are the only current measure of unemployment available with a
high degree of geographic detail. Because of this, LAUS makes extensive use of
UI records in its Handbook estimation procedures for unemployment.
Differences between the CPS concept of unemployment and Handbook estimates
of unemployment are resolved via additivity controls to statewide estimates. (See
Chapter 9.)
The Handbook method breaks unemployment into two main components—
experienced and entrant unemployment—that are each subdivided into subcomponents. Total Handbook unemployment (line 16) is experienced
unemployment (line 11) plus new entrant and re-entrant unemployment (lines 15
and 13, respectively).
Handbook
Line

Description
1. Experienced Unemployed (lines 8 + 9 + 10)
•

11
8

Continued Claimants (lines 5 + 6 + 7)
o State UI
o Unemployment Compensation for
Federal Employees (UCFE)
o Railroad Retirement Board (RRB)

5
6
7

•

Exhaustees

9

•

Non-covered Agricultural Unemployment

10

2. Entrant Unemployment
•

New Entrants

15

•

Re-entrants

13

Total Handbook Unemployment (lines 11 + 13 + 15)

16

Experienced Unemployment (Handbook line 11)
November 2010

LAUS Program Manual 7-20

The largest component of unemployment comprises people that were employed in
the civilian labor force immediately before their current spell of unemployment.
These people are called the experienced unemployed.
The Handbook method estimates this component using three subcomponents:
(1) Continued Claimants (line 8)
•

Monthly counts of people receiving unemployment benefits
during the reference week

(2) Exhaustee Unemployment (line 9)
•

Estimates of people remaining unemployed after exhausting UI
benefits

(3) Non-covered Agricultural Unemployment (line 10)
•
•

Estimates of unemployed agricultural workers that are
ineligible for UI benefits
Only applicable for some States

Continued Claimants (Handbook line 8)
Actual counts of current UI claimants under State UI programs, the
Unemployment Compensation for Federal Employees (UCFE) program, and the
Railroad Retirement Board (RRB) program are included in the count of total
continued claimants. Continued claimants are defined as persons without
earnings certifying to a compensated or non-compensated week of unemployment
during the reference week.
Continued Claimants
Line

Line Description

Input description
(LSS Plus variable ID)

Input
Source

5

State UI
Continued Claims

Regular UI claims (M03)
Interstate UI claims (M04)
Commuter UI claims (M05)

State
State
State

+ 6

UCFE Continued
Claims

Regular UCFE claims (M06)

State

Interstate UCFE claims (M07)
Commuter UCFE claims (M08)

State
State

+ 7
= 8

RRB claims (M09)
RRB Continued
Claims
Total Continued Claims

BLS

Unemployed Exhaustees (Handbook line 9)

November 2010

LAUS Program Manual 7-21

Final payments to beneficiaries of State UI and UCFE programs are tracked by
week and form the main input for exhaustee estimation. The exhaustee
component represents a significant part of the overall unemployed estimates and
is a major contributor to inter-area variability in the estimates. Although States
know the number of individuals who receive final payments, they are unable to
track them after they leave the UI system. Each month, tabulations of weekly
counts (and monthly counts in some areas receiving intrastate commuter claims)
of persons who have received final payments from the UI system are used to
estimate the number of people who do not immediately find a job or discontinue
their job search after exhausting benefits.
Estimates are made to reflect unemployment in the same CPS reference week as
continued claims, that is, the week including the 12th of each month. (In some
years, the December reference week is the week including the 5th of the month.
LAUS technical memoranda advise States when this occurs.) In addition, persons
receiving final payments in previous weeks or months are carried forward in
decreasing numbers into successive periods, by applying a CPS-based “survival”
or continuation rate. This rate refers to individuals who are still actively seeking
and available for work. The estimate of current exhaustees for an area is therefore
“built up” over the period including the 19th of the previous month through the
week including the 12th of the current month, and includes an estimate of the
prior month’s unemployed exhaustees who remain unemployed in the current
month. The level of the pool of exhaustees can rise or fall depending on the
volume of final payments and the survival rate.
Inputs for Unemployed Exhaustees (line 9)
General Input Description

Specific Input description
(LSS Plus variable ID)

Input
Source

Weekly UI and UCFE
Final Payments

Regular (M10)
Interstate (M11)

State
State

Quarterly Survival Rates

Commuter (M12)
Rate group limits (S05 – S08)

State
BLS

Survival rates (S13 – S16)

BLS

November 2010

LAUS Program Manual 7-22

Development of Exhaustee Methodology
Prior to 1987, a national average long-term survival rate was used to estimate the
number of unemployed persons that had exhausted UI benefits. The survival rate
was based on a formula developed by Hyman Kaitz that used national annual
average CPS duration data as the prime input. The underlying premise which
establishes the efficacy of the Kaitz method can be stated as follows: There is a
close and parallel relationship between the rate of unemployment and the duration
of unemployment spells (i.e., the survival rate). However, the application of a
single national annual average survival rate to all States and areas, regardless of
recent local unemployment rate conditions, does not fully conform to Mr. Kaitz’s
theoretical model. Therefore, beginning in 1987, a more flexible, timely, and
effective application of the basic Kaitz long-term survival rate methodology was
made operational.
That method established an unemployment rate-based survival rate that can
change at the area level from month to month. Each quarter, the fifty States and
the District of Columbia are divided into four unemployment rate groups. Each
group represents a set of States within a given range of unemployment rates. In
addition, each group contains roughly twenty- five percent of the nationally
weighted unemployment. Using the Kaitz formula and quarterly average CPS
State duration data, a survival rate is developed for each of the four
unemployment rate ranges. Thus, on a monthly basis, States and areas can select
a survival rate that most closely relates to recent local unemployment rate
conditions. Thus, high unemployment rate areas select a higher survival rate and
have higher exhaustee levels and Handbook estimates than low unemployment
rate areas.
Research established a lagged correlation of two quarters between the
unemployment rate and the survival rate. Adding an operational lag of one quarter
results in the use of a given survival rate based on the area’s unemployment rate
nine months prior.
In implementing this procedure, the following occurs:
1. Every January, April, July, and October, four survival rates are issued
based on CPS data for the most recent quarter (4th, 1st, 2nd, and 3rd). (See
Table 7-3 for survival rates.)
2. Each month during a given quarter, areas select from among these survival
rates in developing LAUS estimates for the quarter of receipt (1st, 2nd,
3rd, and 4th).
3. The selection of the rate is based on the area’s total unemployment rate
nine months prior to the estimate month. This lag represents the two
quarter lagged relationship between the unemployment rate and the
survival rate and a one-quarter operational lag.

November 2010

LAUS Program Manual 7-23

Calculation of Exhaustees from Weekly Final Payments
The following two tables illustrate the steps involved in calculating exhaustees.
The example in the first table pertains to the two- month estimation period for
March (revised prior month) and April (preliminary current month) of 2009.
Each column of the worksheet is described in detail in the second table.

(A)

(B)

(C)

(D)

(E)

Reference
Month

Week
Ending
Date

Total
Final
Payments

Survival
Rate

Exhaustee
Estimate *

February

2/14/2009

31

2/21/2009

30

0.959

380 = ( 31 + 365 ) * 0.959

2/28/2009

29

0.959

393 = ( 30 + 380 ) * 0.959

3/7/2009

16

0.959

405 = ( 29 + 393 ) * 0.959

3/14/2009

17

0.959

403 = ( 16 + 405 ) * 0.959

3/21/2009

16

0.955

401 = ( 17 + 403 ) * 0.955

3/28/2009

17

0.955

399 = ( 16 + 401 ) * 0.955

4/4/2009

32

0.955

397 = ( 17 + 399 ) * 0.955

4/11/2009

27

0.955

410 = ( 32 + 397 ) * 0.955

0.955

417 = ( 27 + 410 ) * 0.955

March
(rev.)

April
(prelim.)

4/12/2009

(F)
Exhaustee Calculation
= [(C) + (E)]prior week x
(D)current week

365 = Starting Pool

* The exhaustee estimate for the reference week is rounded to the whole integer and becomes the
monthly Handbook Line 9 value. For the weeks between reference weeks, the exhaustee estimates
are unrounded. Rounded values are displayed here for ease of reference.

November 2010

LAUS Program Manual 7-24

Description

Column
(A)

(B)

(C)

(D)

Reference Month
•

The months for which estimates are being generated.

•

In the example, the estimation period pertains to March (revised) and April
(preliminary) of 2009. Estimates for these periods were created during the
month of May 2009. The last week of February is included to display the
prior month’s final payments and exhaustee pool.

Week Ending Date
•

The last day of each calendar week displayed by reference month. The last
week of each reference month is the CPS reference week (usually the week
including the 12th day of the month).

•

In the example, each reference month starts in the week following the prior
month’s reference week and extends to the current reference week.

Final Payments
•

The count of people receiving their last unemployment benefit payment
during the week indicated.

•

Final payments made from State UI programs and from the UCFE program
are included. RRB final payments are not tracked.

•

Note that final payments made during the reference week are also counted
as continued claims.

Survival Rate
•

(E)

(F)

The rate at which exhaustees and final payment recipients from the prior
week remain unemployed in the current week.

Exhaustee Estimate
•

A weekly estimate of the number of unemployed people who have
exhausted UI benefits.

•

The exhaustee estimate for the reference week is the monthly Handbook
line 9 estimate.

Exhaustee Calculation
•

The survival rate for the current week is applied to the sum of final
payments for the prior week and the exhaustee pool for the prior week to
yield the exhaustee pool for the current week.
Exhaustees n = survival rate n * (Exhaustees n-1 + Final Payments n-1 )
Where “n” is the current week and “n-1” is the prior week.

November 2010

LAUS Program Manual 7-25

Non-covered Agricultural Unemployment (Handbook line 10)
Generally, this component is a small part of unemployment,
but for some areas with large and highly-seasonal
agricultural sectors, it is very important. For the 19 States
that estimate non-covered agricultural unemployment, this
component accounted for less than one percent of total
Handbook unemployment; however, for individual areas
within those 19 States, the component accounted for up to a
quarter of the unemployed.
Direct estimation of agricultural unemployment may be used in States with at
least one LMA where agricultural employment is 25 percent or more of total
employment. States that qualify must obtain approval from BLS to estimate
agricultural unemployment directly. In such cases, this direct estimation must be
used in all labor market areas of the State. Other States may request approval for
atypical treatment of agricultural unemployment for a specific LMA if it can be
demonstrated that the lack of such an estimate has a deleterious effect on
estimates for that area.
In order to develop this estimate, the relationship between the unemployment rate
for agricultural workers and fo r those from nonagricultural activities is used. This
relationship varies monthly, reflecting differences in the seasonal patterns for the
two groups. Monthly fractional rates for estimating agricultural unemployment
are provided by the national office via the LSS Plus software. These rates, which
are known as “W factors,” were developed by combining separate rates for
agricultural wage and salary and agricultural self-employed and unpaid family
workers. Each group was appropriately weighted, based on CPS monthly
employment levels for the previous two years; the weighted rates are combined by
addition. The appropriate rate is then applied to the non-covered agricultural
employment estimate.
The following formula details the calculation of non-covered agricultural
unemployment. The table below the formula provides details.

If A01 > L03, then L10 = 0, otherwise L10 =
(L03 - A01) * (L08 + L09)
(L04 + L08 + L09)

November 2010

1

(L08 + L09)

w

(L04 + L08 + L09)

LAUS Program Manual 7-26

Where:
Identifier

Description

A01

Annual Average Covered Agricultural Employment
•
•

Obtained from QCEW data
Entered into LSS Plus at beginning of year

L03

Handbook Agricultural Employment

L04

Total Handbook Employment

L08

Total Continued Claims

L09

Exhaustee Unemployment

L10

Non-covered Agricultural Unemployment

w

Monthly fractional rate (or “w” factor)
•
•

Pre-loaded into LSS Plus
Preset across years, but variable by month

Month
January
February
March
April
May
June
July
August
September
October
November
December

November 2010

w Factor
0.934
0.934
0.972
0.791
0.606
0.586
0.623
0.638
0.662
0.676
0.900
1.014

LAUS Program Manual 7-27

Entrant Unemployment (Handbook lines 13 and 15)
For many unemployed individuals, the ir current spell of unemployment was not
immediately preceded by employment. These individuals entered the labor
market from outside the labor force after having completed military service,
family responsibilities, education, or other activities outside the civilian labor
force. These individuals are known as unemployed entrants.
Unemployed entrants are further divided into two groups:
(1) Unemployed New Entrants
•

Individuals who enter the labor market for the first time and do not
find jobs.

(2) Unemployed Re-entrants
•

Individuals who enter the labor market after a period of retirement
from the labor force and are unable to find jobs.

Estimates of new entrants and re-entrants are created for each State, and the
statewide estimates are then allocated to the LMAs within the State using annual
population data. The table below lists the inputs necessary for entrant estimation
in LSS Plus.
Inputs for Unemployed Entrants
Input/ Handbook Line description
(LSS Plus variable ID)

Line

12
13

x
=

Input
Source

Statewide Unemployed Re-Entrants (SRE)

BLS

Re-entrants allocation ratio (L12)
Unemployed Re-entrants ( = SRE * L12)

BLS

Statewide Unemployed New Entrants

BLS
BLS

14

x

New Entrants allocation ratio (L14)

15

=

Unemployed New Entrants ( = SNE * L14)

Statewide New Entrant and Re-entrant Unemp loyment (SNE and SRE)
Statewide new entrant and re-entrant estimates are available from the CPS;
however, the data are volatile and are not suitable for direct use. To reduce the
volatility of the CPS estimates and obtain inputs more suitable for Handbook
estimation, five years of CPS data for a given month are used to develop
weighted-average estimates. The following table shows the weights used, where
“y” is the current year.

November 2010

LAUS Program Manual 7-28

Year

Weight

y
y– 1
y– 2
y– 3
y– 4

0.40
0.25
0.20
0.10
0.05

Once the weighted estimates are calculated, the resulting data for the fifty States
and the District of Columbia are controlled to monthly national CPS estimates of
new entrant and re-entrant unemployment. This controlling step, which was
added to the methodology in 2010, ensures that use of data from earlier years does
not bias the overall level of Handbook entrant unemployment upwards or
downwards during times of generally falling or rising unemployment,
respectively.

Allocation of Statewide Entrants to Handbook Areas (L12 and L14)
Handbook calculations in LSS Plus distribute the statewide new entrant and reentrant estimates to the LMA level using each LMA’s share of statewide agegroup population data. BLS obtains the population data from the U.S. Census
Bureau and calculates the area-share allocation ratios each year. The ratios are
then provided to the States during annual processing. New entrants are
distributed using LMA shares of the population aged 16 to 19 years (L14 ratios),
while shares of the population aged 20 or more years are used to allocate reentrants (L12 ratios).

November 2010

LAUS Program Manual 7-29

Table 7-1 STRATA STEP 3 RATIOS, BY MONTH

2000
1
2
3

JAN
0.96300
1.03700
1.01800

FEB
0.97100
0.99200
1.02100

MAR
1.00100
0.98200
1.00600

APR
0.99900
1.01800
0.99500

MAY
0.91800
1.02300
1.02500

JUN
0.89400
1.03900
1.11800

JUL
0.87100
1.02200
1.25200

AUG
0.85500
1.00800
1.24900

SEP
0.86300
1.03500
1.22800

OCT
0.85700
0.99400
1.18100

NOV
0.85500
0.95400
1.12600

DEC
0.85500
0.96700
1.11900

AUG
0.83300
1.02400
1.18900

SEP
0.87700
1.02200
1.09400

OCT
0.88200
1.02100
1.11000

NOV
0.89300
0.98900
1.07100

DEC
0.89100
0.97300
1.11000

AUG
0.86500
1.04500
1.06600

SEP
0.88200
1.04700
1.06900

OCT
0.88500
1.03800
1.13900

NOV
0.91800
1.03800
1.13500

DEC
0.90900
0.99700
1.12400

AUG
0.94200
0 94200
1.11300
1.14600

SEP
0.91800
0 91800
1.09600
1.11700

OCT
0.91200
0 91200
1.08900
1.14000

NOV
0.92400
0 92400
1.08300
1.18100

DEC
0.88100
0 88100
1.06300
1.20700

AUG
0.97700
1.09700
1.15300

SEP
0.94900
1.06200
1.09500

OCT
0.95700
1.07700
1.10200

NOV
0.95000
1.04000
1.14000

DEC
0.91100
1.02600
1.13900

AUG
0.88800
1.06800
1.16200

SEP
0.91400
1.03000
1.16500

OCT
0.89700
1.05100
1.13800

NOV
0.85500
1.02700
1.17700

DEC
0.86800
1.01400
1.09400

2001
1
2
3

JAN
0.87400
0.99800
1.12500

FEB
0.87700
0.97400
1.14200

MAR
0.91600
1.00400
1.10000

APR
0.88600
1.00100
1.09000

MAY
0.84900
1.00700
1.07900

JUN
0.87500
1.03800
1.15200

1
2
3

JAN
0.84100
0.98300
1.10000

FEB
0.81100
0.97700
1.12100

MAR
0.81500
0.96900
1.11900

APR
0.85200
0.98600
1.12600

MAY
0.84200
0.99800
1.08600

JUN
0.87500
1.02300
1.02200

1
2
3

JAN
0.94700
0 94700
1.01500
1.15200

FEB
0.91800
0 91800
0.98300
1.12200

MAR
0.90300
0 90300
0.99100
1.16200

APR
0.90800
0 90800
1.01200
1.19900

MAY
0.85800
0 85800
1.02500
1.21000

JUN
0.87500
0 87500
1.09900
1.17400

JUL
0.88900
1.05500
1.17200

2002
JUL
0.88400
1.04900
1.06600

2003
JUL
0.90500
0 90500
1.13000
1.15900

2004
JAN
0.90300
1.09300
1.22600

FEB
0.92100
1.04200
1.17100

MAR
0.90200
0.99200
1.15800

APR
0.87900
1.02400
1.09800

MAY
0.87800
1.03700
1.13800

JUN
0.87400
1.06900
1.20000

1
2
3

JAN
0.93200
1.04800
1.23200

FEB
0.92500
1.04800
1.22700

MAR
0.93000
1.03900
1.17000

APR
0.94200
1.08400
1.12000

MAY
0.90300
1.05500
1.10500

JUN
0.89800
1.05700
1.09600

JUL
0.91700
1.11100
1.15000

2005
JUL
0.90200
1.11100
1.15600

Revised 02/01/11

Monthly Step 3 Ratios

7-31 LAUS Program Manual

1
2
3

2006
1
2
3

JAN
0.90200
1.06400
1.17500

FEB
0.92400
1.02900
1.22200

MAR
0.90600
1.01500
1.18900

APR
0.89000
1.03500
1.15900

MAY
0.87900
1.06700
1.08600

JUN
0.87500
1.07900
1.11100

JUL
0.85200
1.11800
1.17200

AUG
0.88900
1.08600
1.12900

SEP
0.90000
1.03900
1.18100

OCT
0.93000
1.04200
1.12100

NOV
0.90700
1.02500
1.09600

DEC
0.89200
1.03100
1.12300

AUG
0.88300
1.06000
1.05800

SEP
0.88900
1.03300
1.06900

OCT
0.89000
0.97300
1.07000

NOV
0.84100
0.96700
1.09700

DEC
0.80600
0.97200
1.04900

AUG
0.89400
1.00500
1.06400

SEP
0.90000
0.96800
1.03600

OCT
0.85500
0.92100
1.08500

NOV
0.80500
0.92400
1.06400

DEC
0.81800
0.91400
1.04200

AUG
0.93300
0 93300
1.02600
1.10800

SEP
0.90000
0 90000
1.03600
1.01200

OCT
0.87500
0 87500
0.98200
1.00300

NOV
0.89700
0 89700
0.99500
1.00400

DEC
0.88000
0 88000
0.98700
1.03700

AUG
0.87500
0.99000
1.02900

SEP
0.84500
0.99500
1.03000

OCT
0.85600
0.95200
1.10400

NOV
0.86400
0.94200
1.09400

DEC
0.85100
0.93100
1.07000

2007
1
2
3

JAN
0.86200
1.04200
1.12100

FEB
0.90200
1.03000
1.12100

MAR
0.89600
1.03600
1.17100

APR
0.89200
1.03500
1.10900

MAY
0.88600
1.03600
1.13900

JUN
0.91500
1.06600
1.15400

1
2
3

JAN
0.81500
0.99600
1.07300

FEB
0.83600
1.03600
1.00100

MAR
0.83100
0.98300
1.03800

APR
0.88400
0.99200
1.05800

MAY
0.88400
0.99100
1.10600

JUN
0.90600
1.00700
1.14200

1
2
3

JAN
0.83900
0 83900
0.91900
1.08800

FEB
0.86900
0 86900
0.96200
1.04000

MAR
0.89300
0 89300
0.98400
1.05900

APR
0.87600
0 87600
0.99000
1.07500

MAY
0.89500
0 89500
0.98600
1.09600

JUN
0.89700
0 89700
1.00500
1.19200

JUL
0.92500
1.05700
1.13800

2008
JUL
0.94100
1.02800
1.12100

2009
JUL
0.92100
0 92100
1.01500
1.07200

2010
1
2
3

JAN
0.88400
1.00100
1.02400

FEB
0.89800
1.01900
1.03000

MAR
0.86000
1.02300
1.05400

APR
0.88200
1.03400
1.02100

MAY
0.89000
1.02100
1.04300

JUN
0.88600
1.03400
1.01800

JUL
0.86100
1.03200
0.99400

Revised 02/01/11

Monthly Step 3 Ratios

LAUS Program Manual 7-31

Table 7-1 STRATA STEP 3 RATIOS, BY MONTH

HISTORICAL CPS AGRICULTURAL EMPLOYMENT MONTHLY ESTIMATION FACTORS
BY AGRICULTURAL ESTIMATING REGIONS
JANUARY 2000-JANUARY 2011
YEAR
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011

JAN
0.94296
0.65064
0.97302
0.76192
0.31621
0.42802
0.49678
0.36057
0.35036
0.37300
0.39055
0.43502

FEB
0.82231
0.57431
0.77667
0.75620
0.27084
0.45521
0.46529
0.39227
0.35298
0.37216
0.32767

MAR
0.98350
0.48666
0.72212
0.53165
0.40465
0.50089
0.57703
0.48160
0.38323
0.32489
0.27664

APR
1.00000
0.61047
0.72686
0.49161
0.45501
0.57252
0.59870
0.49171
0.56199
0.38254
0.38203

MAY
0.93335
0.46364
0.55443
0.47043
0.56755
0.54488
0.69079
0.58939
0.55909
0.47265
0.44918

JUN
1.00073
0.52264
0.71808
0.70798
0.70638
0.70657
0.60124
0.48171
0.69736
0.53740
0.43071

JUL
0.78851
0.70263
0.77487
0.77841
0.66571
0.72608
0.64586
0.60451
0.70684
0.60101
0.58632

AUG
0.77604
0.72600
0.71105
0.79800
0.69768
0.70725
0.68913
0.45726
0.50872
0.49552
0.45131

SEP
0.77415
0.76596
0.64971
0.73410
0.70128
0.68141
0.68850
0.44344
0.47528
0.36749
0.38267

OCT
0.70356
0.78392
0.66756
0.56654
0.47762
0.58445
0.66266
0.43478
0.48270
0.40821
0.37162

NOV
0.71093
0.77877
0.78556
0.44888
0.43720
0.60465
0.53702
0.37949
0.47095
0.45638
0.34955

DEC
0.64456
0.82099
0.76027
0.39125
0.44894
0.66228
0.55610
0.43513
0.47611
0.38049
0.45137

NORTHEAST II

2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011

1.32532
1.35157
1.57029
1.59441
1.52504
1.83486
1.46431
1.60261
1.26910
1.30951
1.27805
1.88062

1.35313
1.39957
1.61837
1.02250
1.38744
1.80895
1.28112
1.41584
1.60881
1.24674
1.56575

1.03846
1.09630
1.73527
0.94511
1.65764
2.14671
1.26833
1.67548
1.40344
1.16771
2.03111

1.00000
1.28606
1.67177
0.90532
1.56659
2.20933
1.27789
1.71369
1.37416
1.20404
2.41455

0.94307
1.02520
1.41139
0.88689
2.02152
2.26560
1.46802
1.99273
1.65422
1.55035
2.28300

1.17532
1.13958
1.60958
1.04091
2.03611
2.23061
1.44661
1.82367
2.10394
1.74565
2.14514

1.28740
1.33868
1.53985
1.36801
1.81118
1.81115
1.52270
1.54987
1.92050
1.58060
1.87173

1.01559
1.38572
1.53188
1.33800
2.32378
2.49557
1.56613
1.15734
1.81808
1.69599
1.65855

1.09981
1.46101
1.95876
1.01271
2.00031
2.06223
1.22126
1.32302
1.84194
1.52116
2.00914

1.06031
1.40312
2.32292
1.18006
1.97744
2.27454
1.48696
1.33701
1.53183
1.24639
2.33548

1.10260
1.16148
2.19023
1.59762
1.94720
1.83105
1.59291
1.36712
1.57012
1.25999
1.71076

1.30612
1.39524
2.09311
1.76959
1.45392
1.19474
1.66936
1.55630
1.51442
1.20550
1.80489

APPALACHIAN I

2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011

0.91908
0.82992
0.53765
0.77721
1.10000
1.09546
0.78623
1.01542
0.94554
0.77127
0.87825
0.64361

0.76992
0.77998
0.48993
0.63339
1.00538
0.95523
0.77970
0.87027
0.56275
0.73257
0.84356

1.05046
0.77904
0.40919
0.74754
1.08678
1.15638
0.72995
0.82828
0.75601
0.61366
0.82615

1.00000
0.73234
0.54749
0.70957
1.14374
1.16738
0.92560
0.93638
0.83371
0.62735
0.91657

1.06238
0.85833
0.72271
0.99891
1.15935
0.88922
0.77343
0.97337
0.83667
0.87749
0.90215

1.38645
1.07506
0.84392
1.02514
0.98921
0.99854
0.94166
0.95696
1.08561
0.87368
0.91414

0.92980
0.87703
0.85236
1.13165
1.03855
1.03840
0.95713
0.91842
0.88274
0.88434
0.87943

1.19878
1.32450
0.62017
1.12238
0.88543
0.97630
0.65416
0.73096
0.86249
0.66682
0.93941

1.24622
1.38372
0.87598
1.02766
0.86084
0.82396
0.71561
0.68177
0.90344
0.72506
0.87987

1.12359
1.06794
0.77353
1.39676
0.94853
0.78240
0.80004
0.94224
0.67755
0.73627
0.94789

0.92760
1.07333
0.65075
1.27904
1.02325
0.87279
0.93774
1.12731
0.77049
0.84109
0.83232

1.02312
0.84791
0.70350
1.15948
1.13145
0.81773
1.08425
1.03956
0.66292
0.77453
0.57432

Revised 02/14/2011

Historical Monthly Agricultural Factors

7-33 LAUS Program Manual

REGION
NORTHEAST I

REGION
APPALACHIAN II

YEAR
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011

JAN
0.70929
0.66860
1.22111
2.14319
1.36933
0.80702
1.16321
1.64101
1.03851
0.75287
0.90701
0.62952

FEB
0.77490
0.71399
1.58552
1.85521
1.36137
0.66866
1.25991
1.77017
1.32547
0.71200
0.82289

MAR
0.96237
0.96396
1.75495
2.20231
1.40238
0.82515
1.42169
1.76711
1.13440
0.72117
1.11224

APR
1.00000
0.98580
1.68057
1.57768
1.63149
1.08546
1.30711
1.46305
0.87616
0.80558
0.84540

MAY
1.03133
1.19889
1.46400
1.14203
1.05742
1.13965
1.19323
1.17091
0.96288
0.86537
0.83053

JUN
1.07813
1.01949
1.05297
0.96547
0.90127
1.32261
1.09229
1.16778
1.01363
0.83474
1.04717

JUL
1.10082
0.88502
0.95132
0.93817
0.90106
0.91894
0.93031
0.87851
0.90692
0.87226
1.01940

AUG
1.16976
0.90522
1.29513
1.37080
1.07770
0.98943
1.22341
0.97254
0.97778
0.92892
1.09308

SEP
1.05360
1.01084
1.34107
1.66507
1.08307
0.97615
1.39011
1.26147
1.09286
0.72949
0.91959

OCT
0.95684
1.09800
1.67274
1.52246
0.92918
0.79619
1.52218
0.98887
0.87707
0.76728
1.00600

NOV
0.73146
1.18020
1.46891
1.61922
1.24206
1.04668
1.42081
1.33179
1.00893
0.66113
0.91096

DEC
0.61322
1.32972
1.78210
1.44822
1.05475
0.91799
1.49501
1.14692
0.95562
0.62839
0.49166

SOUTHEAST

2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011

0.95483
0.79725
0.55653
0.65875
0.28610
0.58960
0.69919
0.43071
0.57006
0.67214
0.43354
0.50801

1.07659
0.61572
0.63790
0.63170
0.19397
0.60735
0.70067
0.53256
0.61064
0.61853
0.44733

1.19927
0.52041
0.69927
0.61162
0.50959
0.64717
0.79388
0.61254
0.60053
0.86667
0.45733

1.00000
0.49983
0.72625
0.61044
0.50129
0.61960
0.64735
0.83962
0.50753
0.80940
0.47177

0.78481
0.59930
0.61823
0.61162
0.71424
0.67933
0.72088
0.99593
0.56606
0.88611
0.58320

0.61993
0.26001
0.67584
0.64410
0.81209
0.69458
0.78668
0.67114
0.69093
1.34821
0.60807

0.74569
0.65430
0.74660
0.80711
0.74144
0.68239
0.85748
0.81240
0.62939
0.76770
0.72449

0.88022
0.53282
0.47786
0.81045
0.78347
0.52677
0.83706
0.67671
0.76700
0.75673
0.76873

0.82332
0.41047
0.44611
0.85421
0.59313
0.59918
0.90127
0.85007
0.77557
0.58655
0.78454

0.84968
0.49117
0.47339
0.88004
0.54276
0.57051
0.77298
0.86584
0.80941
0.66041
0.79692

0.83133
0.34548
0.58878
0.60653
0.40760
0.49083
0.73639
0.72818
0.88983
0.55810
0.55872

0.61922
0.42995
0.64634
0.38387
0.42512
0.52851
0.63165
0.78918
0.84467
0.59164
0.56312

FLORIDA

2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011

1.45180
1.06668
0.52422
0.38505
0.29844
0.50610
0.61478
0.54761
0.16598
0.29629
0.70151
0.65183

1.19803
0.93482
0.57981
0.48165
0.34624
0.61115
0.52184
0.41162
0.15105
0.24703
0.52993

1.09410
0.93927
0.77536
0.63888
0.35787
0.58879
0.53583
0.13186
0.13173
0.20711
0.49244

1.00000
1.11911
0.81411
0.59296
0.21699
0.32957
0.53737
0.09668
0.23701
0.32940
0.57981

0.81355
0.89446
0.68192
0.54685
0.35880
0.35177
0.29568
0.11491
0.37326
0.45951
0.52788

0.94980
0.49382
0.57887
0.51103
0.37216
0.32941
0.34917
0.26217
0.39509
0.43687
0.56076

0.68411
0.53844
0.46556
0.47770
0.43626
0.35449
0.38649
0.29828
0.29970
0.38095
0.50379

0.87047
0.60456
0.39454
0.35488
0.40388
0.38336
0.56852
0.29359
0.24562
0.43645
0.76446

0.59965
0.51576
0.41085
0.34466
0.42751
0.42018
0.62207
0.22039
0.22086
0.56604
1.06823

0.50527
0.54379
0.37313
0.32188
0.47945
0.69368
0.71330
0.17344
0.26448
0.70028
1.08458

0.48302
0.47263
0.24219
0.32254
0.37464
0.54544
0.79636
0.19922
0.30018
0.71562
0.89019

0.63042
0.24064
0.32224
0.42556
0.57526
0.43666
0.61279
0.15627
0.27488
0.70812
0.82451

Revised 02/14/2011

Historical Monthly Agricultural Factors

LAUS Program Manual 7-34

HISTORICAL CPS AGRICULTURAL EMPLOYMENT MONTHLY ESTIMATION FACTORS
BY AGRICULTURAL ESTIMATING REGIONS
JANUARY 2000-JANUARY 2011

HISTORICAL CPS AGRICULTURAL EMPLOYMENT MONTHLY ESTIMATION FACTORS
BY AGRICULTURAL ESTIMATING REGIONS
JANUARY 2000-JANUARY 2011
REGION

JAN
0.87287
0.56085
0.77122
0.89911
0.89910
0.55346
0.67457
0.60012
0.61819
0.71438
0.82839
1.03569

FEB
1.05878
0.53346
0.80322
0.82077
0.85794
0.57383
0.65528
0.75939
0.67300
0.74265
0.91779

MAR
1.00613
0.56403
0.90679
0.95839
0.74936
0.61798
0.72950
0.81388
0.72436
0.84198
0.97479

APR
1.00000
0.82099
0.93147
0.98440
0.84507
0.69041
0.79324
0.79498
0.80413
0.91867
1.10531

MAY
1.06250
0.87802
0.94818
0.99010
0.82433
0.77896
0.87934
0.83370
0.83419
0.81631
1.16355

JUN
0.84180
0.90959
0.97419
0.92817
0.82974
0.86206
0.85035
0.77266
0.87301
0.82661
1.05412

JUL
1.02932
1.17659
1.04934
0.92664
0.95718
0.84180
0.86019
0.78987
0.77356
0.90732
0.98903

AUG
1.02042
1.18112
0.99168
0.99611
0.83163
0.76804
0.96818
0.78633
0.64401
0.84807
0.85792

SEP
0.97537
1.10400
1.15509
0.87728
0.77159
0.78659
0.89660
0.77297
0.69434
0.75807
0.85334

OCT
1.00192
1.03292
1.27280
1.12772
0.69901
0.69140
0.92336
0.80904
0.63251
0.71918
0.97269

NOV
0.79302
0.82931
1.02936
1.01173
0.64018
0.60273
0.85217
0.69510
0.55650
0.71840
0.91137

DEC
0.63717
0.81969
0.97147
0.87827
0.58515
0.64742
0.72409
0.60407
0.50316
0.69514
1.00936

CORN BELT I

2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011

0.96734
0.59660
0.87312
1.00435
0.72901
1.09702
0.95683
1.18233
1.20002
1.06253
1.07880
1.53357

1.16700
0.80154
0.87544
0.75415
0.69795
1.29301
0.90805
1.02438
1.06470
1.00062
1.20203

1.00166
0.92330
1.07995
0.97789
0.72731
1.27132
0.99683
1.07346
0.82138
0.94793
1.17488

1.00000
0.76574
1.22703
1.12051
0.90532
1.48627
1.25881
0.99586
0.73498
0.98738
1.32935

1.10134
0.84375
1.21312
1.20848
0.95660
1.40450
1.14497
1.16764
0.99297
1.07824
1.35872

1.20600
0.73233
1.34821
1.33601
0.98163
1.32245
1.23479
1.19248
1.13699
1.16452
1.53005

1.26096
1.29686
1.41110
1.15694
1.18160
1.32645
1.35513
1.37135
1.41473
1.23063
1.38264

1.33933
1.19233
1.11592
1.02031
1.22037
1.27720
1.10222
1.32793
1.36779
0.93282
1.32261

1.35228
1.19943
1.19049
0.99366
1.46286
1.68614
1.04726
1.21000
1.12312
0.94185
1.08180

1.08908
1.01092
1.38625
1.09410
1.66532
1.64440
1.15874
1.25331
1.20573
1.00584
1.10602

0.92504
0.81878
1.14771
0.89095
1.59929
1.62039
0.94653
1.34769
1.23403
1.04901
1.16728

0.74289
0.85567
1.10564
0.69844
1.34254
1.16644
1.11589
1.12393
1.06799
1.12387
1.39411

CORN BELT II

2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011

0.79853
0.90225
1.02582
0.66738
0.60371
0.62374
0.84837
0.87367
0.95439
0.78290
0.82558
1.07872

0.86710
1.14254
0.90227
0.69785
0.69929
0.69133
0.70611
0.75867
0.85312
0.79515
0.98341

0.91480
1.02252
0.80408
0.68433
0.80418
0.71149
0.76357
0.84116
0.85208
1.03029
0.99384

1.00000
1.12872
0.71198
0.62742
0.99247
0.94415
0.91851
0.88404
0.95988
0.97458
0.93043

1.12080
1.18183
0.58575
0.63684
0.89546
0.92468
0.98177
0.95736
1.02399
0.86186
0.77302

1.00916
0.79893
0.69431
0.84656
1.00063
0.95985
0.94801
1.04303
1.03923
0.80944
0.73429

1.06962
1.14891
0.97787
0.87228
0.91468
1.00889
0.96697
0.88036
0.95993
0.96336
0.92841

0.97440
0.99096
0.92200
0.94576
0.94437
0.81736
0.93959
0.94427
0.91531
0.78221
0.87515

0.87511
0.92880
0.90332
0.87071
0.98254
0.73576
0.79599
0.92302
0.84486
0.79347
0.94301

0.81929
1.17009
0.84507
0.83442
0.87994
0.87034
0.91565
1.09700
1.05291
0.96886
1.11087

0.92834
1.01318
0.71651
0.68000
0.84662
0.83249
0.92093
0.97097
0.93258
0.87627
0.87216

0.74947
0.94712
0.67987
0.54347
0.72395
0.86878
0.92011
0.97324
0.80937
0.82410
0.93555

LAUS Program Manual 7-35

Revised 02/14/2011

Historical Monthly Agricultural Factors

YEAR
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011

LAKE

REGION
DELTA

YEAR
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011

JAN
0.96414
0.81551
1.09846
1.50921
1.14701
0.91811
1.08268
1.18139
1.39798
0.85297
0.73742
0.94006

FEB
0.79434
0.87702
0.98152
1.18626
1.05013
0.71478
1.13287
1.38383
1.41513
0.90250
0.78418

MAR
0.84434
1.10465
1.08648
1.32175
0.92631
0.66907
1.24734
1.26477
1.65529
0.95218
0.83235

APR
1.00000
0.86849
1.00326
1.28748
0.98219
0.97267
1.42857
1.20674
1.35407
1.05817
0.91040

MAY
0.91205
0.87585
1.21005
1.35191
1.00565
1.13808
1.20176
0.96958
1.18031
1.27740
0.96218

JUN
0.90121
0.80589
1.09030
1.40014
0.87222
1.21107
1.19026
0.84076
0.89645
1.13754
0.89204

JUL
0.96847
0.98611
1.14323
1.23190
1.11551
1.24262
1.20559
0.94157
0.96866
1.08196
1.11915

AUG
0.88569
0.87910
1.28280
1.51618
1.17256
1.06598
1.24633
1.22362
1.10377
0.97220
1.01596

SEP
1.05761
1.00218
1.20163
1.47237
1.09346
1.05964
1.52254
1.60231
0.94304
0.79877
0.78595

OCT
1.22778
1.08982
1.61390
1.70751
1.29218
1.24141
1.53031
1.52990
1.11783
0.86733
0.98878

NOV
1.17841
1.05119
1.51789
1.55346
1.16362
0.93102
1.44700
1.50058
1.27429
0.94088
0.93610

DEC
1.26016
0.99675
1.55545
1.18732
0.85021
0.83845
1.16303
1.32253
0.82784
0.93975
0.90413

NORTHERN PLAINS

2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011

1.01572
0.96740
1.11418
0.96490
1.02366
0.97332
0.97428
0.76018
0.73258
0.84511
0.87857
0.92396

1.09924
0.96187
1.03999
0.99879
1.00927
0.91199
0.93140
0.76249
0.70852
0.83039
0.98048

1.01789
1.02658
1.01678
1.04823
1.02891
0.98855
0.93555
0.79791
0.79467
0.85765
0.98573

1.00000
1.09071
1.09733
1.07199
1.07668
0.91227
0.93383
0.77671
0.73615
0.87442
0.97037

1.14141
1.10629
1.17697
1.18130
1.14470
0.91992
0.96789
0.85111
0.76967
0.86182
0.92722

1.23859
1.14954
1.16225
1.23303
1.20262
1.06323
1.11987
1.00412
0.94313
0.97144
1.06428

1.23216
1.09340
1.17052
1.22547
1.16420
1.08221
1.12451
1.02941
1.03752
1.00841
1.02061

1.18723
1.06567
1.10433
1.20903
1.06437
0.98555
1.02759
0.89268
1.02811
1.01846
1.04686

1.04829
1.04425
1.04316
1.18392
0.95699
1.04345
1.05510
0.92018
1.00724
1.01768
1.01603

1.08088
1.15585
1.11074
1.21025
0.96207
1.08832
1.02575
0.82305
0.87927
0.91657
1.01840

1.02472
1.06757
1.07140
1.14235
0.93576
1.01652
0.91272
0.73128
0.84778
0.99043
1.02623

0.99935
0.98783
0.96177
1.04755
0.99724
1.04020
0.86819
0.75166
0.86699
0.92303
0.91826

SOUTHERN PLAINS

2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011

1.13595
1.07412
1.15057
1.22097
1.17898
0.95102
0.81830
0.91485
1.05805
1.08191
0.95279
0.94622

1.10192
0.92116
0.96813
1.24139
1.08983
0.71566
0.70997
1.03144
0.86621
1.11308
0.92782

1.10352
1.08330
1.09474
1.47712
0.94489
0.80594
0.69114
1.00126
1.04353
1.12668
0.95802

1.00000
0.98726
1.11036
1.45585
1.20480
1.03346
0.78835
1.03424
1.03890
1.20610
1.02229

1.17774
1.12846
1.18305
1.43222
1.01865
0.92860
0.73438
1.04734
1.00518
1.26288
0.96188

1.47747
1.23240
1.13803
1.46564
1.38368
1.06217
1.03109
1.02590
1.14786
1.35373
1.10616

1.59380
1.42706
1.32134
1.38261
1.36290
1.24223
1.10422
1.00765
1.15314
1.15198
1.07643

1.44767
1.43321
1.23550
1.42465
1.49796
0.98414
0.82833
0.68755
1.17509
1.21951
1.01895

1.37980
1.33126
1.38681
1.56695
1.32658
0.95371
0.79219
0.87891
1.42018
1.06543
1.04096

1.20030
1.38911
1.33767
1.50711
1.15048
0.96579
0.76013
0.85863
1.22729
0.96695
1.04210

1.15649
1.27203
1.25394
1.45023
1.06192
0.85372
0.74870
0.91403
1.28576
1.06849
0.97392

1.26997
1.24962
1.12646
1.15730
0.91693
0.84190
0.93695
1.02896
1.28853
1.01007
1.02767

Revised 02/14/2011

Historical Monthly Agricultural Factors

7-36 LAUS Program Manual

HISTORICAL CPS AGRICULTURAL EMPLOYMENT MONTHLY ESTIMATION FACTORS
BY AGRICULTURAL ESTIMATING REGIONS
JANUARY 2000-JANUARY 2011

HISTORICAL CPS AGRICULTURAL EMPLOYMENT MONTHLY ESTIMATION FACTORS
BY AGRICULTURAL ESTIMATING REGIONS
JANUARY 2000-JANUARY 2011
YEAR
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011

JAN
1.09504
0.69411
0.61833
0.64275
0.64602
0.76141
0.79699
0.69931
1.01975
1.10162
0.79332
0.95014

FEB
0.98151
0.59975
0.56013
0.71583
0.68036
0.71552
0.76666
0.78246
1.08122
1.09558
0.89797

MAR
0.99354
0.61784
0.47366
0.73192
0.74941
0.81132
0.80769
0.87912
1.15372
1.15299
0.94723

APR
1.00000
0.72999
0.59630
0.84045
0.84774
0.79824
0.78519
0.76840
1.17286
1.12048
0.98813

MAY
1.00287
0.68617
0.76115
0.89053
1.01235
0.99986
0.90202
0.85603
1.01177
1.00483
0.97913

JUN
1.07058
0.88794
0.94933
0.86774
1.06516
1.15906
1.08129
1.03422
1.13145
1.02200
0.86957

JUL
1.03033
1.03125
0.94292
0.93496
1.01003
1.04010
1.06382
0.98455
1.11009
0.94344
0.88417

AUG
1.00381
0.98283
0.97218
0.96340
1.14682
1.06538
0.93335
1.06576
1.11278
0.92975
0.89990

SEP
1.07478
0.99900
0.87885
0.84476
1.01865
0.81894
0.87817
1.12040
1.12060
0.97492
0.89448

OCT
1.09697
0.97862
0.85010
0.74177
1.04281
0.91019
0.80466
0.99217
0.98816
0.78935
0.86638

NOV
0.81820
0.85272
0.75413
0.68452
0.89458
0.84297
0.77179
0.88481
1.02694
0.87142
0.92474

DEC
0.78179
0.78072
0.72134
0.63057
0.75637
0.84775
0.77147
0.92478
1.05036
0.82507
0.84744

MOUNTAIN II

2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011

0.67301
0.68265
1.47275
1.90693
1.49322
1.28799
1.16447
1.37839
1.34681
1.34876
1.18573
0.91294

0.83519
0.90223
1.45838
1.35479
1.42507
1.15537
1.21971
1.48550
0.84999
1.27435
1.07457

0.86818
0.86510
1.28685
1.06228
1.70417
1.24274
1.20348
1.27701
0.56817
1.26086
0.90528

1.00000
0.92792
1.31302
1.20208
1.69936
1.27732
1.39176
1.34013
0.28525
0.98864
0.87820

0.95883
1.03845
1.46833
0.99164
1.52642
1.06232
1.18111
1.38844
0.80729
0.93422
1.30569

0.95490
0.82000
1.70931
1.57921
1.90060
1.64340
1.90223
1.36295
1.03663
1.26541
1.52268

1.22684
1.22723
1.59991
1.73612
1.66173
1.54544
1.70850
1.56176
1.65979
1.88349
1.54998

1.12702
1.34352
1.65489
1.41025
1.42750
1.50243
1.79493
1.64425
1.64866
2.44936
1.85938

0.90973
1.64375
1.52324
1.46151
1.32942
1.59423
1.71858
1.57975
1.65815
2.58115
1.90014

0.91744
1.45568
1.73099
1.26166
0.87944
1.44971
1.28295
1.58460
1.65676
2.46647
1.69575

0.74080
1.49951
1.50960
0.95671
0.92910
1.37801
1.59357
1.67173
1.20414
1.69476
1.72074

0.67459
1.41704
1.94986
0.88576
0.80856
1.10060
1.23140
1.41002
1.31249
1.44542
1.19568

MOUNTAIN III

2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011

2.02626
1.83203
0.92873
0.79876
0.91285
0.29225
0.49382
0.74466
0.99770
0.44005
0.44343
0.60040

1.65235
1.20172
0.88422
0.73760
0.70100
0.21641
0.37618
0.62872
0.72246
0.43261
0.46570

1.29765
1.32306
0.98192
0.75969
0.59951
0.44231
0.42256
0.63971
0.59129
0.38871
0.42517

1.00000
0.99393
0.78689
0.60267
0.43188
0.54896
0.63583
0.62285
0.58687
0.68263
0.34618

1.36450
1.35137
0.87045
0.58044
0.61471
1.01867
0.67842
0.46983
0.71846
0.63076
0.34586

1.17068
1.30996
1.12595
0.93363
0.92920
1.15157
0.95222
0.72599
0.70905
0.71439
0.63463

0.96292
0.92513
0.93465
0.93548
0.67553
0.77501
0.91120
0.83508
0.79914
0.66751
0.54014

1.18060
0.77368
0.77397
0.87199
0.57347
0.75358
0.81859
0.84008
0.68446
0.76206
0.62648

0.71930
0.67231
0.92243
1.08739
0.62911
0.35551
0.74149
1.11290
0.64344
0.86497
0.72850

1.40956
0.73248
1.17841
1.29013
0.38173
0.37426
0.69784
0.91566
0.41056
0.66471
1.03742

1.75845
0.57285
0.92625
1.14698
0.38519
0.53768
0.67258
0.99148
0.39679
0.75215
1.10671

1.73058
0.70261
1.00125
1.14275
0.40249
0.50813
0.75684
0.78199
0.43724
0.84340
0.81292

Revised 02/14/2011

Historical Monthly Agricultural Factors

LAUS Program Manual 7-37

REGION
MOUNTAIN I

REGION
PACIFIC

YEAR
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011

JAN
0.70716
0.72669
0.46141
0.32530
0.49363
0.37362
0.44970
0.61328
0.54286
0.51833
0.55809
0.52883

FEB
0.75590
0.95332
0.48200
0.36605
0.65620
0.41902
0.58910
0.69573
0.60686
0.48310
0.63940

MAR
1.05111
1.03020
0.57958
0.32884
0.64096
0.41759
0.46557
0.57506
0.65070
0.44151
0.54179

APR
1.00000
0.83688
0.59868
0.46672
0.68689
0.46022
0.53693
0.50699
0.69721
0.47869
0.62328

MAY
1.04959
0.80365
0.69682
0.53447
0.68315
0.36534
0.49190
0.53893
0.52915
0.44478
0.38188

JUN
0.98536
0.81755
0.53382
0.59890
0.65938
0.43308
0.62890
0.62047
0.50030
0.52065
0.56602

JUL
0.78307
0.75606
0.62402
0.59011
0.74281
0.57297
0.72367
0.73609
0.69923
0.70134
0.71802

AUG
0.96614
0.77712
0.60446
0.46474
0.73760
0.48515
0.66174
0.64929
0.60245
0.59615
0.66563

SEP
0.90977
0.75296
0.50035
0.48778
0.55930
0.49182
0.52837
0.52578
0.54806
0.57239
0.63706

OCT
0.72071
0.61477
0.59345
0.47919
0.46192
0.38397
0.42784
0.48192
0.54351
0.59454
0.68835

NOV
0.98959
0.54157
0.45339
0.46356
0.35827
0.34593
0.45391
0.35636
0.42474
0.53560
0.60184

DEC
0.92465
0.54521
0.35090
0.48651
0.32663
0.37298
0.38378
0.42807
0.44652
0.48330
0.44192

CALIFORNIA

2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011

0.79513
0.93313
0.70435
0.61301
0.69957
0.44343
0.43837
0.62070
0.80237
0.77641
0.66176
0.64237

0.90336
0.82903
0.73619
0.59597
0.65108
0.44124
0.48851
0.64563
0.86819
0.78531
0.85096

0.72709
0.80997
0.55563
0.59434
0.61116
0.39096
0.49030
0.61552
0.89406
0.56580
0.66561

1.00000
0.98891
0.81795
0.65367
0.72067
0.47715
0.60440
0.66503
0.86611
0.70056
0.61640

0.98122
1.07683
0.70261
0.74572
0.83729
0.53340
0.71884
0.65616
0.90892
0.74528
0.94692

0.94770
0.89717
0.61900
0.77888
0.78598
0.68879
0.81569
0.71794
0.86066
0.72105
0.73947

0.99863
0.79054
0.88048
0.83406
0.72174
0.77538
0.78720
0.83323
0.84544
0.72006
0.87661

0.95784
0.90043
0.83004
1.19497
0.91715
0.76827
0.89470
0.67057
0.92731
0.73586
0.80159

0.91729
0.82444
0.84194
1.14101
0.78285
0.67608
0.73028
0.83833
0.88208
0.64433
0.63467

0.85287
0.74548
0.79248
1.06649
0.83316
0.60853
0.65139
0.84816
0.98748
0.62872
0.69411

0.71999
0.56492
0.51903
0.82108
0.76822
0.62671
0.58755
0.83156
0.84298
0.60904
0.60877

0.78695
0.55365
0.55720
0.73068
0.69368
0.48463
0.58882
0.89385
0.94241
0.60368
0.59872

HAWAII

2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011

1.22702
0.76809
1.55472
1.36924
2.65916
1.35401
0.86750
0.91431
0.74792
0.72079
0.57759
0.38465

0.90942
0.61241
1.45777
1.64882
2.73413
1.75604
1.11075
0.98205
0.57154
0.79950
0.83027

0.59427
0.77031
1.28740
1.23854
2.01171
1.97333
1.20695
1.09815
0.76505
0.87974
0.95634

1.00000
0.94562
1.24612
1.35146
2.30935
1.72819
0.82593
1.05473
1.06797
0.94436
0.63438

1.57722
0.75451
0.94624
1.38681
1.91861
1.32365
0.66906
1.00874
1.32134
1.46728
0.58335

1.84935
0.79362
1.21175
1.08068
0.32765
1.69256
0.84137
1.04388
1.25590
1.26719
0.59415

1.69527
1.46547
1.29918
1.10465
1.31989
1.40079
0.93923
0.96825
1.11452
1.08814
0.81955

1.74949
1.45036
1.68652
1.29515
0.80943
1.36987
0.88369
0.54617
1.00385
1.03120
0.93154

1.19395
1.18337
1.70451
1.40555
0.94748
1.20684
0.66823
0.71357
1.20246
0.93203
0.79020

1.22390
1.01200
1.87598
1.68657
1.09270
0.78273
0.63984
0.72020
1.02391
0.78151
0.61711

1.06279
0.78081
1.62086
2.45368
1.02005
1.01738
0.45787
0.91540
0.91713
0.74932
0.61329

1.07897
1.31135
1.47888
2.24686
1.36424
1.17983
0.60899
0.76154
0.67238
0.60674
0.40961

Revised 02/14/2011

Historical Monthly Agricultural Factors

7-38 LAUS Program Manual

HISTORICAL CPS AGRICULTURAL EMPLOYMENT MONTHLY ESTIMATION FACTORS
BY AGRICULTURAL ESTIMATING REGIONS
JANUARY 2000-JANUARY 2011

HISTORICAL CPS AGRICULTURAL EMPLOYMENT MONTHLY ESTIMATION FACTORS
BY AGRICULTURAL ESTIMATING REGIONS
JANUARY 2000-JANUARY 2011
REGION
MICHIGAN

JAN
0.87287
0.44322
0.71433
0.91274
1.00247
0.67602
0.95039
0.82405
0.81365
0.90244
1.09899
1.44275

FEB
1.05878
0.39623
0.75615
0.83997
0.97032
0.70400
0.94612
1.05072
0.88642
0.93264
1.22238

MAR
1.00613
0.40719
0.85523
0.98004
0.86784
0.76001
1.05733
1.11942
0.95468
1.05692
1.30597

APR
1.00000
0.64455
0.89120
1.01011
0.97846
0.84934
1.15560
1.07747
1.05998
1.15123
1.47987

MAY
1.06250
0.68197
0.92078
1.02017
0.96760
0.95766
1.28147
1.12239
1.10049
1.00848
1.56498

JUN
0.84180
0.69393
0.95781
0.96357
0.98402
1.05956
1.26522
1.01689
1.15242
1.01490
1.44413

JUL
1.02932
0.94133
1.03415
0.96651
1.12773
1.03967
1.29690
1.02938
1.02415
1.11453
1.37775

AUG
1.00082
0.96117
0.98177
1.04975
0.98380
0.96812
1.44627
1.02611
0.84541
1.05532
1.19511

SEP
0.93616
0.91569
1.14725
0.93659
0.91705
1.01057
1.32489
1.01003
0.90480
0.95834
1.18872

OCT
0.94311
0.87504
1.26770
1.20858
0.83553
0.91256
1.35178
1.05837
0.81573
0.92413
1.35498

NOV
0.71460
0.72836
1.03222
1.09839
0.77020
0.82259
1.23099
0.91121
0.70787
0.93675
1.26956

DEC
0.53915
0.73689
0.97961
0.96996
0.70937
0.89732
1.02442
0.79391
0.63002
0.92174
1.40607

MINNESOTA

2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011

0.87287
0.61726
0.78020
0.79835
0.74792
0.44892
0.50731
0.44693
0.44090
0.55170
0.73091
0.95711

1.05878
0.59927
0.80443
0.70960
0.71807
0.46250
0.48672
0.56442
0.47800
0.58113
0.81281

1.00613
0.63924
0.90711
0.83076
0.63370
0.49634
0.54077
0.60587
0.51249
0.65950
0.86812

1.00000
0.90561
0.92333
0.84344
0.71457
0.55428
0.58646
0.59404
0.56842
0.72228
0.98375

1.06250
0.97204
0.93081
0.83638
0.70124
0.62595
0.65000
0.62404
0.58684
0.66178
1.04008

0.84180
1.01301
0.94848
0.76358
0.70907
0.69298
0.62167
0.58163
0.61188
0.67885
0.95884

1.02932
1.28941
1.02001
0.74949
0.81560
0.67194
0.62432
0.59601
0.53258
0.74440
0.91399

1.02982
1.28354
0.95136
0.80913
0.70484
0.60787
0.70460
0.58908
0.45336
0.70433
0.79283

0.99417
1.18820
1.09759
0.71647
0.64990
0.61749
0.65454
0.57474
0.49798
0.63904
0.78859

1.03013
1.09948
1.19941
0.92248
0.58428
0.53632
0.67586
0.59769
0.46540
0.61567
0.89889

0.83062
0.86551
0.95017
0.83212
0.53037
0.46035
0.62608
0.50746
0.42304
0.62359
0.84222

0.68418
0.84414
0.88129
0.72762
0.47970
0.49083
0.53502
0.43451
0.39630
0.61304
0.93278

WISCONSIN

2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011

0.87287
0.57190
0.79897
1.01314
1.01374
0.59797
0.69149
0.63515
0.70209
0.78545
0.76345
0.85475

1.05878
0.54635
0.83378
0.94375
0.95260
0.62146
0.66358
0.79966
0.76635
0.81096
0.83890

1.00613
0.57876
0.94151
1.09983
0.81023
0.67015
0.73729
0.86045
0.82684
0.91897
0.87987

1.00000
0.83756
0.96887
1.13939
0.91398
0.74882
0.79962
0.84859
0.91841
1.00070
0.99908

1.06250
0.89643
0.98810
1.15775
0.87743
0.84457
0.88625
0.89376
0.95560
0.87459
1.04137

0.84180
0.92984
1.01681
1.10548
0.87237
0.93454
0.84779
0.84018
1.00238
0.87925
0.90521

1.02932
1.19868
1.09558
1.11629
1.01434
0.91498
0.85152
0.86406
0.89789
0.96565
0.81624

1.02226
1.20550
1.04778
1.18841
0.88320
0.82786
0.96528
0.86450
0.74023
0.88289
0.70803

0.97905
1.12915
1.23079
1.03371
0.82148
0.84106
0.90126
0.85419
0.79134
0.76741
0.70425

1.00745
1.05895
1.36609
1.32384
0.74648
0.73065
0.93460
0.89795
0.71230
0.70631
0.80275

0.80038
0.85373
1.12433
1.17255
0.68604
0.62732
0.87098
0.77761
0.61677
0.68579
0.75214

0.64638
0.84614
1.07629
1.00021
0.62964
0.66894
0.75103
0.68233
0.54756
0.64133
0.83302

Revised 02/14/2011

Historical Monthly Agricultural Factors

LAUS Program Manual 7-39

YEAR
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011

Quarterly Survival Rates
Table 7-3 Quarterly Average Exhaustee Survival Rates Based on Unemployment Rates
Survival Rate
Use Period

Months for Selecting
Survival Rates

January-March
2000

April-June
1999

April-June
2000

July-September
1999

July-September
2000

October-December
1999

October-December
2000

January-March
2000

January-March
2001

April-June
2000

April-June
2001

July-September
2000

July-September
2001

October-December
2000

October-December
2001

January-March
2001

January-March
2002

April-June
2001

April-June
2002

July-September
2001

July-September
2002

October-December
2001

October-December
2002

January-March
2002

Unemployment
Rate Range
0.0-4.0
4.1-4.5
4.6-4.8
4.9 and up
0.0-4.2
4.3-4.7
4.8-5.0
5.1 and up
0.0-3.8
3.9-4.0
4.1-4.6
4.7 and up
0.0-4.3
4.4-4.6
4.7-4.9
5.0 and up
0.0-3.8
3.9-4.1
4.2-4.8
4.9 and up
0.0-3.8
3.9-4.3
4.4-4.9
p
5.0 and up
0.0-3.5
3.6-4.0
4.1-4.5
4.6 and up
0.0-3.5
3.6-4.0
4.1-4.5
4.6 and up
00.-3.2
3.3-3.9
4.0-4.4
4.5 and up
0.0-4.2
4.3-5.0
5.1-5.7
5.8 and up
0.0-5.2
5.3-5.5
5.6-5.8
5.9 and up
0.0-5.5
5.6-6.3
6.4-6.8
6.9 and up

Weekly
Survival Rates
0.946
0.952
0.952
0.955
0.945
0.947
0.947
0.947
0.940
0.946
0.946
0.950
0.951
0.951
0.951
0.961
0.943
0.943
0.943
0.951
0.939
0.942
0.947
0.947
0.938
0.938
0.938
0.943
0.941
0.945
0.945
0.947
0.945
0.945
0.945
0.946
0.942
0.943
0.943
0.943
0.951
0.953
0.953
0.955
0.957
0.959
0.960
0.960
Revised 02/01/11

LAUS Prorgam Manual 7-41

Quarterly Survival Rates
Table 7-3 Quarterly Average Exhaustee Survival Rates Based on Unemployment Rates
Survival Rate
Use Period

Months for Selecting
Survival Rates

January-March
2003

April-June
2002

April-June
2003

July-September
2002

July-September
2003

October-December
2002

October-December
2003

January-March
2003

January-March
2004

April-June
2003

April-June
2004

July-September
2003

July-September
2004

October-December
2003

October-December
2004

January-March
2004

January-March
2005

April-June
2004

April-June
2005

July-September
2004

July-September
2005

October-December
2004

October-December
2005

January-March
2005

Unemployment
Rate Range
0.0-5.3
5.4-5.9
6.0-6.3
6.4 and up
0.0-4.9
5.0-5.5
5.6-6.2
6.3 and up
0.0-5.2
5.3-5.8
5.9-6.3
6.4 and up
0.0-5.5
5.6-6.5
6.6-7.6
7.7 and up
0.0-3.8
3.9-4.8
4.9-5.6
5.7 and up
0.0-4.7
4.8-5.8
5.9-7.0
p
7.1and up
0.0-4.0
4.1-5.0
5.1-7.1
7.2 and up
0.0-3.9
4.0-5.0
5.1-5.6
5.7 and up
0.0-3.0
3.1-4.0
4.1-4.7
4.8 and up
0.0-3.0
3.1-5.1
5.2-6.0
6.1 and up
0.0-2.8
2.9-4.7
4.8-5.7
5.8 and up
0.0-2.6
2.7-4.4
4.5-5.5
5.6 and up

Weekly
Survival Rates
0.956
0.961
0.962
0.965
0.977
0.979
0.981
0.981
0.950
0.951
0.953
0.954
0.957
0.960
0.965
0.965
0.960
0.960
0.960
0.961
0.953
0.953
0.957
0.963
0.958
0.958
0.960
0.961
0.962
0.962
0.962
0.964
0.959
0.959
0.959
0.961
0.955
0.955
0.956
0.959
0.957
0.957
0.959
0.960
0.958
0.958
0.958
0.959
Revised 02/01/11

LAUS Prorgam Manual 7-42

Quarterly Survival Rates
Table 7-3 Quarterly Average Exhaustee Survival Rates Based on Unemployment Rates
Survival Rate
Use Period

Months for Selecting
Survival Rates

January-March
2006

April-June
2005

April-June
2006

July-September
2005

July-September
2006

October-December
2005

October-December
2006

January-March
2005

January-March
2007

April-June
2006

April-June
2007

July-September
2006

July-September
2007

October-December
2006

October-December
2007

January-March
2007

January-March
2008

April-June
2007

April-June
2008

July-September
2007

July-September
2008

October-December
2007

October-December
2008

January-March
2008

Unemployment
Rate Range
0.0-2.7
2.8-4.4
4.5-5.3
5.5 and up
0.0-2.9
3.0-4.9
5.0-5.5
5.6 and up
0.0-2.7
2.6-4.2
4.3-4.9
4.9 and up
0.0-2.2
2.3-4.8
4.9-5.4
5.6 and up
0.0-2.8
2.9-4.5
4.6-5.0
5.1 and up
0.0-2.4
2.5-4.3
4.4-5.0
p
5.1 and up
0.0-1.8
1.9-3.8
3.9-5.0
5.1 and up
0.0-2.1
2.2-4.3
4.4-5.5
5.6 and up
00.-2.1
2.2-3.6
3.7-4.9
5.0 and up
0.0-2.1
2.2-3.5
3.6-5.1
5.2 and up
0.0-4.1
4.2-4.9
5.0-5.5
5.6 and up
0.0-2.9
3.0-4.4
4.5-5.6
5.7 and up

Weekly
Survival Rates
0.960
0.960
0.960
0.960
0.950
0.950
0.952
0.953
0.952
0.952
0.956
0.959
0.96
0.96
0.961
0.961
0.953
0.953
0.954
0.955
0.951
0.951
0.953
0.958
0.950
0.950
0.950
0.951
0.955
0.955
0.957
0.959
0.955
0.955
0.959
0.959
0.953
0.953
0.953
0.953
0.947
0.953
0.954
0.960
0.953
0.953
0.958
0.959
Revised 02/01/11

LAUS Prorgam Manual 7-43

Quarterly Survival Rates
Table 7-3 Quarterly Average Exhaustee Survival Rates Based on Unemployment Rates
Survival Rate
Use Period

Months for Selecting
Survival Rates

January-March
2009

April-June
2008

April-June
2009

July-September
2008

July-September
2009

October-December
2008

October-December
2009

January-March
2009

January-March
2010

April-June
2009

April-June
2010

July-September
2009

July-September
2010

October-December
2009

October-December
2010

January-March
2010

January-March
2011

April-June
2010

Unemployment
Rate Range
0.0-4.6
4.7-5.1
5.2-6.0
6.2 and up
0.0-5.7
5.8-6.3
6.4-7.3
7.4 and up
0.0-5.9
6.0-6.7
6.8-6.9
6.9 and up
0
4.9
8.1
10.0 and up
0.0-8.1
8.2-9.9
10.0-10.9
11.0 and up
0.0-7.8
7.9-9.7
9.8-10.9
p
11.0 and up
0.0-6.9
7.0-8.9
9.0-10.9
11.0 and up
0.0-7.7
7.8-9.9
10-11.9
12.0 and up
0.0-6.9
7.0-8.9
9.0-9.9
10.0 and up

Weekly
Survival Rates
0.953
0.958
0.959
0.965
0.955
0.956
0.962
0.965
0.955
0.958
0.959
0.965
0.964
0.964
0.969
0.973
0.967
0.971
0.972
0.975
0.967
0.970
0.975
0.976
0.972
0.974
0.977
0.978
0.975
0.977
0.980
0.982
0.966
0.974
0.978
0.979
Revised 02/01/11

LAUS Prorgam Manual 7-44

8 LAUS Estimation: Geography
Introduction
he Local Area Unemployment Statistics (LAUS) program is responsible for
estimation of unemployment rates for all areas below the national level.
This chapter will provide a comprehensive review of LAUS geography
beginning with the largest areas and proceeding through the area types in
descending order by size.

T

For each type of geography, the source of the delineation, the role of the
geography in LAUS estimation, and the estimation methodology will be noted.
Detailed information regarding each estimation methodology can be found in the
appropriate chapters.

November 2010

LAUS Program Manual 8-1

Census Regions and Divisions
The U.S. Census Bureau has designated supra-State regions and divisions for
which LAUS creates labor force estimates. The four Census Regions each
comprise two or more Census Divisions. The nine Census Divisions each
comprise three or more States. The table below summarizes the geographic
composition of these areas.
LAUS creates estimates for Census Divisions using statistical models that
incorporate data from the Current Population Survey (CPS). Estimates for Census
Regions are developed by summing model-based data for the Census Divisions.
Note that Puerto Rico, which is within the scope of the LAUS program, is not part
of any Census Region or Division.

Census Region
Northeast

Midwest

South

West

November 2010

Census Division

States

New England

Connecticut, Maine, Massachusetts, New
Hampshire, Rhode Island, and Vermont

Middle Atlantic

New York, New Jersey, and Pennsylvania

East North
Central

Illinois, Indiana, Michigan, Ohio, and
Wisconsin

West North
Central

Iowa, Minnesota, Nebraska, North Dakota,
South Dakota, Kansas, and Missouri

South Atlantic

Delaware, District of Columbia, Maryland,
Virginia, West Virginia, Florida, Georgia,
North Carolina, and South Carolina

East South
Central

Alabama, Kentucky, Mississippi, and
Tennessee

West South
Central

Arkansas, Louisiana, Oklahoma, and Texas

Mountain

Colorado, Montana, New Mexico, Utah,
Wyoming, Arizona, Idaho, and Nevada

Pacific

Alaska, California, Hawaii, Oregon, and
Washington

LAUS Program Manual 8-2

Bureau of Labor Statistics Regions
The Bureau of Labor Statistics (BLS) has subdivided the Nation into regions
similar to those designated by the U.S. Census Bureau. However, the BLS
regions comprise States directly, rather than via divisions that subdivide the
regions. As such, there are no “BLS Divisions.”
LAUS does not create estimates for BLS Regions.
administrative purposes only.

BLS
Region
Number

BLS Region

These areas are for

States

1

Boston / New
York

Connecticut, New York, Maine, Massachusetts, New
Hampshire, Rhode Island, Vermont, and Puerto Rico

3

Philadelphia

Delaware, New Jersey, Pennsylvania, District of
Columbia, Maryland, Virginia, and West Virginia

4

Atlanta

Alabama, Florida, Georgia, North Carolina, South
Carolina, Kentucky, Mississippi, and Tennessee

5

Chicago

Illinois, Iowa, Indiana, Michigan, Ohio, Wisconsin,
Minnesota, Nebraska, North Dakota, and South Dakota

6

Dallas / Kansas
City

Arkansas, Colorado, Kansas, Missouri, Louisiana,
Oklahoma, Texas, Montana, New Mexico, Utah, and
Wyoming

8

San Francisco

Alaska, Arizona, Idaho, Nevada, California, Hawaii,
Oregon, and Washington

States
LAUS publishes estimates for each of the 50 States in the Nation and for the
District of Columbia and the Commonwealth of Puerto Rico. (Though the
District and Puerto Rico are not States, the LAUS program generally treats them
as such for administrative purposes.)
Estimates for 48 of the States and for the District of Columbia are developed
using statistical models that incorporate CPS, CES, and UI data. For the two
remaining States—California and New York—estimates are created by summing
the respective Balance of State and substate model-based area within each.
Estimates for Puerto Rico are derived from a survey similar to the CPS.

November 2010

LAUS Program Manual 8-3

Balances of State
Eight Balances of State exist to facilitate model-based estimation for areas within
States. The LAUS program delineates these as the portion of a State that remains
after removing the substate modeled area within the State.
LAUS creates estimates for Balances of States using statistical models. For six of
these, the models are based solely on CPS, while for the remaining two Balances
the models are based on CPS, CES, and UI data. The two Balances that
incorporate CES and UI data in their models are those for California and New
York.
Model-based estimation of the New Orleans-Metairie-Kenner, LA Metropolitan
Statistical Area and the Balance of Louisiana ended in August 2005 following
Hurricane Katrina and its impact on the available CPS sample in the substate
modeled area. Since August 2005, data for the Balance of Louisiana are produced
by summing its constituent parts, rather than via model-based estimation.

State

Balance of State

Substate Modeled Area

California

Balance of California

Los Angeles-Long Beach-Glendale,
CA Metropolitan Division

Florida

Balance of Florida

Miami-Miami Beach-Kendall, FL
Metropolitan Division

Illinois

Balance of Illinois

Chicago-Joliet-Naperville, IL
Metropolitan Division

Louisiana *

Balance of Louisiana *

New Orleans-Metairie-Kenner, LA
Metropolitan Statistical Area *

Michigan

Balance of Michigan

Detroit-Warren-Livonia, MI
Metropolitan Statistical Area

New York

Balance of New York

New York city, NY

Ohio

Balance of Ohio

Cleveland-Elyria -Mentor, OH
Metropolitan Statistical Area

Washington

Balance of Washington

Seattle-Bellevue-Everett, WA
Metropolitan Division

* The Balance of Louisiana and New Orleans-Metairie-Kenner, LA Metropolitan
Statistical Area were dropped as model-based areas following Hurricane Katrina.

November 2010

LAUS Program Manual 8-4

Labor Market Areas
In the late 1940's when subnational labor force estimation was first attempted,
employment and unemployment estimates were developed for large labor market
areas as well as States, underscoring the importance of substate labor market
information. Labor market areas (LMAs) are identified in order to standardize and
promote comparability for the collection and use of labor force information in
administering various government programs. In the LAUS program, substate
estimates of employment and unemployment are prepared for all LMAs in the
Nation. Labor market areas are defined in terms of full counties in all areas except
New England, where Minor Civil Divisions (MCDs) are used to define LMAs. In
the criteria below, the term “county” includes county equivalents and, in New
England, refers to MCDs. (The Labor Market Area Directory for 2010 provides
titles and definitions for all labor market areas covered by the LAUS program. It
is on the BLS website, at http://www.bls.gov/lau/lmadir.pdf.)
Generally a labor market area is defined as an economically integrated geographic
area within which individuals can reside and find employment within a reasonable
distance or can readily change employment without changing their place of
residence. LMAs are either metropolitan areas (MAs), micropolitan areas (MCs)
or small labor market areas (SAs of CNs) and exhaust the geography of each of
the States, the District of Columbia, and Puerto Rico. The Office of Management
and Budget (OMB) is responsible for defining metropolitan and micropolitan
areas while the Local Area Unemployment Statistics Division (LAUS) of the BLS
performs this function for small labor market areas. Currently, there are 380
metropolitan areas, 590 micropolitan areas, and 1,382 small labor market areas.
Areas are designated on the basis of population, urbanization, and commutation
data. Since population and urban area data are inappropriate for defining the
generally less populous small labor market areas, commutation data are used to
determine which counties are deemed single-county labor- market areas and which
are combined into multi-county labor-market areas. Regardless of population
size, commuting flows are an indication of the degree of integration of labor
markets among counties.
Under the 2000 standards, “Metropolitan Statistical Area” and “Micropolitan
Statistical Area” are the terms used for the basic set of county-based areas defined
under this classification. In addition, the term “Metropolitan Division” is used to
refer to a county or group of counties within a Metropolitan Statistical Area that
has a population core of at least 2.5 million. A Metropolitan Division is most
generally comparable in concept to the now obsolete Primary Metropolitan
Statistical Area.
While a Metropolitan Division is a subdivision of a larger Metropolitan Statistical
Area, it often functions as a distinct social, economic, and cultural area with the
larger region. Metropolitan Divisions retain their separate statistical identities.
Federal agenc ies provide detailed data for each Metropolitan Division, just as they
did in the past for the Primary Metropolitan Statistical Areas.

November 2010

LAUS Program Manual 8-5

Standards for Defining Labor Market Areas
and Combined Statistical Areas
Metropolitan and Micropolitan Statistical Areas are
collectively called Core Based Statistical Areas (CBSAs). The
Metropolitan and Micropolitan Statistical Area Standards do
not equate to an urban-rural classification; all counties
included in Metropolitan and Micropolitan Statistical Areas
and many other counties contain both urban and rural territory
and populations.
Core
A Core is a densely settled concentration of population, comprising either an
urbanized area (of 50,000 or more population) or an urban cluster (of 10,000 to
49,999 population) defined by the Census Bureau, around which a Core Based
Statistical Area is defined.
Core Based Statistical Area
A Core Based Statistical Area (CBSA) is a statistical geographic entity consisting
of the county or counties associated with at least one core (urbanized area or
urban cluster) of at least 10,000 population, plus adjacent counties having a high
degree of social and economic integration with the core as measured through
commuting ties with the counties containing the core. Metropolitan and
Micropolitan Statistical Areas are the two categories of Core Based Statistical
Areas.
Metropolitan Statistical Area
A Metropolitan Statistical Area (MA) is a CBSA associated with at least one
urbanized area that has a population of at least 50,000. The Metropolitan
Statistical Area comprises the central county or counties containing the core, plus
adjacent outlying counties having a high degree of social and economic
integration with the central county as measured through commuting.
Micropolitan Statistical Area
A Micropolitan Statistical Area (MC) is a CBSA associated with at least one
urban cluster that has a population of at least 10,000, but less than 50,000. The
Micropolitan Statistical Area comprises the central county or counties containing
the core, plus adjacent outlying counties having a high degree of social and
economic integration with the central county as measured through commuting.
New England City and Town Area
A New England City and Town Area (NECTA) is a statistical geographic entity
that is defined using cities and towns as building blocks and that is conceptually
similar to the Core Based Statistical Areas in New England (which are defined
using counties as building blocks).

November 2010

LAUS Program Manual 8-6

Metropolitan Division
A Metropolitan Division (MD) is a county or group of counties within a CBSA
that contains a core with a population of at least 2.5 million. A Metropolitan
Division consists of one or more main/secondary counties that represent an
employment center or centers, plus adjacent counties associated with the main
county or counties through commuting ties.
New England City and Town Area Division
A New England City and Town Area (NECTA) Division is a city or town or
group of cities and towns within a NECTA that contains a core with a population
of at least 2.5 million. A NECTA Division consists of a main city or town that
represents an employment center, plus adjacent cities and towns associated with
the main city or town, or with other cities and towns that are in turn associated
with the main city or town, through commuting ties.
Combined Statistical Area
A Combined Statistical Area is a geographic entity consisting of two or more
adjacent CBSAs linked through commuting ties. Areas were combined based on
the employment interchange rate, defined as the sum of the percentage of
commuting from the CBSA with the smaller total population to the CBSA with
the larger total population and the percentage of employment in the CBSA with
the smaller total population accounted for by workers residing in the CBSA with
the larger total population. Pairs of CBSAs with employment interchange
measures of at least 25 were combined automatically. Pairs of CBSAs with
employment interchange measures of at least 15, but less than 25, were combined
if local opinion in both areas favored combination.

November 2010

LAUS Program Manual 8-7

Standards for Defining Small Labor Market Areas
While CBSAs are designated by the Office of Management and Budget (OMB),
the LAUS Division of the BLS designates small labor market areas. Similar to
the federal statistical areas developed by OMB, multi- entity small labor market
areas (SAs) were created based on 2000 Census-based commutation ties.
However, unlike the federal statistical areas, no population criteria were applied
for SAs.
(1)

Worker flows were examined, and counties combined into one small LMA
if either of the following conditions was met:
(a) At least 25.0 percent of the employed residents of one county
commute to work in another county
(b) At least 25.0 percent of the employment (persons working) in one
county is accounted for by workers commuting from another county.

(2)

Small LMAs, as is the case with metropolitan and micropolitan areas, are
required to be contiguous. Counties were first combined based on the
commutation criteria, and then potential multi-county small LMAs were
checked for contiguity. Noncontiguous portions of potential small LMAs
were considered separately. If the noncontiguous area contained more
than one county, it was be reevaluated using (1)(a) and (1)(b) above. If
the noncontiguous area consists of a single county, it was designated as a
separate small LMA.

(3)

Subsequent to the verification of contiguity described in (2) above,
commuting flows between adjacent small LMAs were evaluated. Those
areas for which the measures and thresholds specified in (1a) and (1b)
above are met were merged to form one small LMA.

(4)

For the New England Minor Civil Divisions (MCD) based small LMAs,
due to the very large number of small MCDs, residual MCDs were added
to contiguous small LMAs based on commuting flows and/or other
economic ties. That is, if, after applying the commutation criteria, an
MCD was identified as an individual small LMA, the MCD was added to
a contiguous small LMA, especially if the MCD was extremely small.
Also, there were a number of individual MCDs isolated between CBSAs.
For the purpose of Handbook estimation only, these MCDs were grouped
with adjacent micropolitan areas, with their LAUS estimates being
disaggregated from the larger area.

November 2010

LAUS Program Manual 8-8

Counties
LAUS creates estimates for all of the 3,219 counties in the Nation and Puerto
Rico, including county equivalents (Boroughs and Census Areas in Alaska,
Parishes in Louisiana, Independent Cities in Virginia, Municipios in Puerto Rico,
and various county/city areas in other States).
Methodologies vary for the counties. In most States, they are derived by
Handbook estimation or disaggregation, depending on the county’s status with
regard to Labor Market Area definitions. In the New England States, counties are
derived generally by addition of MCDs.
Cities / Subcounty areas
LAUS often uses the term “cities” to denote areas below the county level;
however, fewer than half of all subcounty areas included in the LAUS program
are legally incorporated as cities within their respective States.
For non-New England States, LAUS creates estimates for subcounty areas (cities
and other units of local government below the county level) with populations
greater than or equal to 25,000 people. Annual population estimates produced by
the Census Bureau are used to determine the set of areas to be included for LAUS
estimation.
For the New England States, LAUS creates estimates for all subcounty areas with
nonzero labor force levels in Summary File (SF) 3 of Census 2000.
The U.S. Census Bureau identifies two major categories of subcounty areas, each
of which are further broken down into two major subcategories:
(1) Places
a. Incorporated Places
These areas are legally incorporated under the laws of a State. The
majority of incorporated places for which LAUS creates estimates are
cities, though there are four other types of incorporated places for
which LAUS creates estimates as displayed in the following table.
Type of
Incorporated
Place *

States with LAUS areas
by Type of Incorporated Place

Boroughs

Connecticut, Pennsylvania, and New Jersey

Municipalities

Alaska and Pennsylvania

Towns

Arizona, California, Colorado, Florida, Illinois,
Indiana, Massachusetts, New Jersey, North Carolina,
South Carolina, Tennessee, Texas, and Virginia

Villages

Florida, Illinois, New Jersey, New York, and
Wisconsin

* Cities, the most common type of incorporated place, are excluded from
the table. Nearly all States have cities for which LAUS estimates are
created.

November 2010

LAUS Program Manual 8-9

b. Census Designated Places (CDPs)
CDPs are statistical entities and have no governmental
functions under State law. LAUS does not create estimates for
these areas.
(2) County Subdivisions
a. Minor Civil Divisions (MCDs)
MCDs have legal boundaries and names and governmental
functions or administrative purposes specified by State law.
LAUS creates estimates for MCDs in select States. For most,
the disaggregation method is used.
State

Number of MCDs recognized
by LAUS and Type

Connecticut

149 towns

Maine

498 *

Massachusetts

304 towns

Michigan

32 townships

New Hampshire

224 *

New Jersey

65 townships

New York

63 towns

Pennsylvania

37 townships

Rhode Island

31 towns

Vermont

241 *

Total

1,644

* Multiple MCD types exist in the three northernmost New England States,
including gores, locations, plantations, reservations, towns, townships, and
unorganized territories.

b.

Census County Divisions (CCDs)
CCDs are statistical entities and have no governmental
functions under State law. LAUS does not create estimates for
these areas.

November 2010

LAUS Program Manual 8-10

City Parts / Incorporated Place Parts
There are 116 incorporated places for which LAUS (1) creates estimates and (2)
recognizes territory in more than one county. For any incorporated place
estimated by LAUS, estimates for the county-specific parts are created if the
incorporated place has more than one county part with a nonzero labor force level
in Summary File (SF) 3 of Census 2000. The 116 incorporated places are mostly
cities, though there are towns in North Carolina and South Carolina and villages
in Illinois with LAUS-estimated county parts. Overall, LAUS creates estimates
for 252 parts of the 116 incorporated places.
Handbook Areas
Handbook areas are those areas fo r which the Handbook Method is used to
generate estimates. These areas exhaust the geography of the Nation and Puerto
Rico with the exception of the District of Columbia. For the District of Columbia,
no Handbook estimates are calculated.
The phrase “Labor Market Areas (LMAs)” is often used as a synonym for those
areas for which Handbook estimates are calculated. While there is substantial
overlap between LMAs and Handbook areas, there are notable differences. The
geographic scope of Handbook areas includes all 50 States and Puerto Rico, but
not the District of Columbia. The scope of LMAs includes the District. There are
2,357 Handbook areas, but only 2,352 Labor Market Areas. Of these, 2,305 are
common between the two area types. The following table provides detailed
information for Handbook areas that are not the geographic equivalent of an
LMA.
52 Handbook areas that are
not LMAs

47 LMAs that are not Handbook
areas

32 Metropolitan Divisions or
NECTA Divisions

10 Metropolitan Statistical Areas (MSAs) *

14 New England Expanded
Estimating Areas

8 Metropolitan NECTAs

1 Metropolitan NECTA

6 Micropolitan NECTAs
18 Isolated MCDs
4 Adjacent LMAs

6 Components of 2 Metropolitan
Divisions

2 MSAs *

* The 10 MSAs noted in the first category and the 2 MSAs noted in the third
overlap. Two of the 10 are comprised of both Metropolitan Divisions that are
Handbook areas and Metropolitan Divisions comprised of components that are
Handbook areas.

November 2010

LAUS Program Manual 8-11

Expanded Estimating Areas
Expanded Estimating Areas are unique to the New England States and were
created to facilitate estimation for very small LMAs. They are Handbook Areas
comprising one Metropolitan or Micropolitan NECTA and additional territory.
The additional territory inc luded in any given Expanded Estimating Area is either
one or more Isolated Minor Civil Divisions (Isolated MCDs) or one Adjacent
Small LMA. Of the 14 Expanded Estimating Areas, 8 include a Metropolitan
NECTA, while the remaining 6 include a Micropolitan NECTA. In addition to
the Metropolitan/Micropolitan NECTA, each Expanded Estimating Area includes
either an Adjacent Small LMA or one or more Isolated Minor Civil Divisions
(MCDs). These are areas that qualify as Small LMAs, but which were deemed
too small for effective Handbook estimation.
All of the MCD components of the Expanded Estimating Areas are disaggregated.
Estimates for the NECTA and (where applicable) the Adjacent Small LMA are
then created by aggregating the disaggregated MCD components.

November 2010

LAUS Program Manual 8-12

9 LAUS Estimation: Additivity
Introduction
inking substate labor force estimates to the CPS begins with a set of
independent Handbook employment and unemployment estimates. These
are prepared for all LMAs—that is, officially-designated Metropolitan
Areas (MAs), single-county labor market areas, multi-county areas, and
aggregations of cities and towns in New England—such that they exhaust all
geographic subdivisions of a State. Because of nonlinearity in the Handbook, the
LMA employment and unemployment estimates will not necessarily equal the
statewide total. Hence, an additivity adjustment must be performed. This process
introduces conformity between the Handbook and the statewide estimates by
making the sum of the exhaustive LMA estimates additive to State levels.

L

Usual statistical practice is to distribute aggregation differences proportionally
among the individual parts. In this manner, all components receive a
proportionate share of the difference between the sum of the parts and an
independent total. The LAUS program uses this simple linear additivity
adjustment method, referred to as the Handbook-Share technique, to adjust LMA
estimates to the State control totals. This method consists of distributing the
statewide estimates, based on the percentage share of each Handbook area
estimate, over the total of the Handbook estimates. This assumes a proportional
distribution throughout the State of the difference between the sum of the
independent Handbook estimates and the State control totals. This method is
applied to all areas for which an independent Handbook estimate is prepared and
to the intrastate portions of interstate areas. The adjustments for additivity are
performed on a current basis, and whenever the State estimates are revised.
After the Handbook estimates have been adjusted for additivity to the statewide
November 2010

LAUS Program Manual 8-1

estimates, LMA estimates are referred to as “LAUS” estimates rather than
Handbook estimates. These LAUS estimates are then disaggregated further into
smaller areas, such as single counties within multi-county LMAs, or sub-county
areas, such as cities and towns, for which estimates may be required by
legislation. Two methods for disaggregation exist based on the availability of UI
claims and decennial census data for apportioning LMA estimates to smaller
areas. As with the requirement for additivity of LMA estimates to statewide
totals, sub-LMA estimates produced by disaggregation are additive to the LMA
estimates. See Chapter 9 for a complete description of the disaggregation
process.

Adjustment to Independent Statewide Estimates—
The Handbook Share Method
The process of reconciling, or linking, LMA labor
force estimates to Statewide (model-based)
estimates begins with a set of geographically
exhaustive LMA Handbook employment and
unemployment estimates. A simultaneous
adjustment for additivity of all LMA estimates to
the statewide estimates is performed using the
percentage distribution of the substate Handbook
estimates, also known as the Handbook-Share
method. The Handbook-Share method of apportioning the State estimates of
unemployment and employment to areas assumes a proportional distribution
throughout the State of the difference between the sum of substate Handbook
estimates and the independent State estimates. This adjustment is performed for
both preliminary and revised estimates.
The Handbook-Share method should be followed by all States unless the State can
demonstrate and document why the linear additivity adjustment procedure is
inappropriate. The State must also be able to suggest a reasonable and equitable
alternative distribution. Reasons for alternative procedures may include
inconsistent quality of employment estimates or deficiencies in the Handbook
estimates for minor LMAs. Documentation should show how the alternative
procedure differs from linear adjustment in terms of the distribution of
employment and unemployment in the State. Linear additivity adjustments should
be reviewed annually and exception requests should be submitted to the Regional
Office before annual benchmarking.
The following worksheet illustrates simultaneous additivity and adjustments to
LMAs using the Handbook-Share method.

November 2010

LAUS Program Manual 8-2

Simultaneous Additivity of LMA Estimates
Using the Handbook-share Method

Area

Unemployment

Employment

Percent of
Summed
Handbook

Percent of
Summed
Handbook

Handbook

State

Statewide*

Handbook

49,300

Statewide*

562,800

MSA 1

18,500

0.394456

19,447

190,600

0.3481279

195,926

Major LMA 1

9,300

0.198294

9,776

107,100

0.1956164

110,093

Major LMA 2

8,700

0.185501

9,145

103,400

0.1888585

106,290

Minor LMA 1

2,300

0.049041

2,418

36,800

0.0672146

37,828

Minor LMA 2

1,900

0.045120

1,997

25,900

0.0493059

26,624

Intrastate Por-

6,200

0.132196

6,517

83,700

0.1528767

86,039

46,900

1.000000

49,300

547,500

1.000000

562,800

tion of Inter
state MSA 2
Sum of substate
Areas

*For the State, enter the model-based estimate. The substate data are the product of the
area’s Percent of Summed Handbook and the statewide estimate for unemployment or
employment.

Interstate Areas
For interstate areas, after the independent Handbook estimate is prepared by the
“controlling” State, the intrastate portions are calculated through the
disaggregation process. These intrastate portions are then adjusted for additivity
to the respective statewide levels by each State. The intrastate portions are then
summed to the total interstate area to obtain the employment and unemployment
estimates for the labor market area.

November 2010

LAUS Program Manual 8-3

November 2010

LAUS Program Manual 8-4

10

LAUS Estimation:
Disaggregation

Introduction

D

isaggregation techniques are used to obtain current
estimates of employment and unemployment for subareas within labor market areas. Disaggregation
involves prorating employment and unemployment for the
labor market area to sub-area jurisdictions. Since these
jurisdictions are within LMAs, independent employment and unemployment estimates
cannot be developed, as basic data are not always available and current LAUS estimating
procedures are not applicable.
Disaggregation methods are used to develop estimates for counties within multicounty
labor markets areas, cities within counties (either single -county LMAs or disaggregated
counties), and cities and towns within LMAs in New England.
Two methods of disaggregation are appropriate for LAUS use.
1.) The population-claims method uses current UI claims data by county or place (city or
town) of residence, 2000 census population data by age group, 2000 census employment
data, and the most recent Census Bureau population estimates. This method uses a
separate methodology for employment (employment/population indexed share) and
unemployment (claims-based unemployment disaggregation).
2.) The census-share method uses 2000 census employment and unemployment data.

A hierarchy of disaggregation techniques exists. For counties and county equivalents and
cities and towns in New England, the population-claims method of disaggregation is
required since the necessary residency-based claims data are available. Outside of New
England, for places within counties, such as cities, the population-claims method of
disaggregation was made mandatory in 2007 for States without approved Cooperative
Agreement variances. If census data employment and unemployment for the jurisdiction
are not available, contact the BLS Regional Office for assistance before proceeding.

November 2010

LAUS Program Manual 10-1

Note: Throughout this chapter, the term ‘county’ will be used synonymously with ‘cities
or towns’ in New England.
The starting point for disaggregation is the estimate of employment and unemployment
prepared for the LMA in accordance with Handbook instructions outlined in Chapter 7
and the directions on adjustment for additivity to statewide totals in Chapter 9.

November 2010

LAUS Program Manual 10-2

The Population-Claims Method of Disaggregation
Since current employment and unemployment estimates at the sub-LMA level are
required to implement numerous Federal economic assistance and employment and
training programs, methods of disaggregation which reflect current economic conditions
in these small areas are necessary. Other than the decennial census, there are very few
data series for small areas. Two exceptions are the monthly UI claims series and the
annual population estimates prepared by the Census Bureau.
Specifically, the current data used in LAUS disaggregation are UI continued claims by
place of residence for the week including the 12th of the month and total population
estimates prepared annually for counties and biennially for places pertaining to July 1 of
the given year. The procedure which incorporates the use of these data is known as the
population and claims-based disaggregation procedure.
Population-Based Employment Disaggregation for Counties

Early research showed that, for disaggregating labor market area employment, annually
prepared population estimates alone at the county and city/town level were superior to the
use of fixed decennial census ratios of total employment. The use of fixed census ratios
assumed no change in the ratios over the intercensal decade. The annual population
method allowed the ratios to change over time, but assumed that the
employment/population ratio in each subarea was equal to the employment/population
ratio of the labor market area as a whole. In many instances, this assumption proved
unrealistic.
Subsequent research devised a disaggregation procedure which allows for differences in
employment/population ratios within a LMA. This procedure utilizes the relationship of
the subarea’s employment/population ratio to that of the larger area, using decennial
census data. The assumption is that the relationship of the employment/population ratios
(the ratio of the subarea’s share of LMA employment to the subarea’s share of LMA
population) will hold relatively constant in the intercensal period.
This employment disaggregation method can only be used in conjunction with the
claims-based unemployment disaggregation.

November 2010

LAUS Program Manual 10-3

Employment Disaggregation Procedure and Sequence
A disaggregation procedure fitting the description above can be expressed as follows:

Where:
1.) E = total employment
2.) P = total population
3.) c = 2000 census
4.) i = designation of ith county in LMA (variables with no subscript pertain to the
LMA as a whole)
5.) t = reference period of the estimates (For employment, the reference period is
the month. For population, the reference period is the appropriate year).

When this procedure is used to disaggregate employment, additivity of the counties to the
LMA is not assured, because each county has its own employment/population index
share. Forcing additivity into the disaggregation yields the following modified formula,
known as the employment/population (E/P) Index share procedure:

Where E, P, c, i, and t are as defined above and h = the total number of counties in the
LMA. The additivity property can be verified by summing both the left-hand and righthand terms.
The Census Bureau prepares total population estimates pertaining to July 1 each year for
States and substate areas. Using decennial census data and the annually-prepared total
population estimates, an employment/population index share is calculated annually for
each county in a LMA.
Given the time lag in issuance of the population data from the Census Bureau, estimates
for the most recent year may not always be available at benchmark revision time. A
LAUS Technical Memorandum, typically issued in late fall each year, will advise the
States of the availability of population estimates.

November 2010

LAUS Program Manual 10-4

Applying the Employment Disaggregation Procedure
Each year, data are developed to produce the county employment/population indexshares, as follows:

Worksheet A.
Developing County Employment/Population Index-Shares
2000 CENSUS

County
A
B
C
LMA Total

Employmen
t
I
16,500
12,900
10,000
39,400

Population
II
32,000
25,300
22,000
79,300

CURRENT
Populatio
n
III
35,500
28,700
24,000
88,200

1st Stage
Employmen
t

E/P Indexshare

(I X III)/II
IV
18,305
14,634
10,909
43,847

IV ÷ Σ (IVs)
V
0.417464
0.333739
0.248797
1.00000

Step 1. Data from the 2000 census on total employment and population are entered in

Columns I and II for all counties (cities and towns in New England) in the LMA.
Step 2. The most recent July 1 population estimates (Current) are entered in Column III,

and the year of reference indicated.
Step 3. For each county, a first-stage employment level is calculated by moving 2000

census employment by the change in the county’s population since the census. Thus,
Column IV equals Column I times Column III divided by Column II.
Step 4. The county first-stage employment estimates in column IV are summed to a LMA

total (Column IV, LMA Total), Thus, the LMA total in column IV is a sum and should
not be calculated by applying the formula in the column heading to the LMA total
employment and population.
Step 5. The employment/population index-share is calculated for each county by dividing

the first-stage employment level in the county by the labor market area sum obtained in
Step 4. Thus, Column V equals Column IV divided by the Column IV LMA Total.

The sum of the employment/population index-shares should equal one, except for
rounding. If it does not precisely equal one, the largest share is rounded so that the sum of
the shares is exactly equal to 1. This will ensure that employment from the counties sums
to the LMA total employment. The shares are then used in the following Worksheet B,
Employment/Population Index-Share Approach to Disaggregating Total Employment, as
follows:

November 2010

LAUS Program Manual 10-5

Worksheet B.
Employment/Population Index-Share Approach to Disaggregating Total Employment
County

E/P Indexshare

I
A
B
C
LMA Total

II
0.417464
0.333739
0.248797
1.000000

January
Employmen
t
12,647
10,111
7,537
30,295

February
Employmen
t
12,891
10,306
7,683
30,879

….

November
Employmen
t

December
Employmen
t

14,814
11,843
8,829
35,485

15,087
12,061
8,992
36,140

III
…
…
…
...

Step 1. From Worksheet A, Column V, enter the employment/population index-shares in

Column II for each county listed in Column I.
Step 2. Enter the independent estimate of total employment for the LMA under the

appropriate month in Column III. The independent estimate for the labor market is the
estimate that results from the application of the Handbook estimating procedure and
adjustment for additivity to the statewide controls.
Step 3. The county employment/population index-shares are applied to the independent

LMA employment estimate to arrive at the disaggregated county employment estimates
for the month. The sum of the disaggregated county employment estimates may not add
to the LMA total because of rounding. If this is the case, the estimate for the largest
county should be adjusted so that the summed estimates equal the LMA total.

The same procedures are used to produce employment/population index-shares for cities:

Worksheet C.
Developing City Employment/Population Index-Shares
2000 CENSUS
City

CURRENT

1st Stage
Employment

E/P
Indexedshare

Employment

Population

Population

(I X III)/II

IV ÷ Σ (IVs)

I
18,300
14,000

II
38,000
29,500

III
42,000
33,000

IV
20,226
15,661

V
0.322077
0.249381

23,600

57,000

65,000

26,912

0.428542

55,900

124,500

140,000

62,800

1.00000

A
B
Balance of
County
County Total

Step 1. Data from the 2000 census on total employment and population are entered in

Columns I and II for all LAUS cities in the county. State specific cities should not be
included. Balance of county employment and population are derived by subtracting the
respective data for LAUS cities from the county totals.
Step 2. The most recent July 1 population estimates are entered in Column III, and the

year of reference indicated.
Step 3. For each city and for balance of county, a first-stage employment level is

calculated by moving 2000 census employment by the change in the city’s or balance of
county’s population since the census. Thus, Column IV equals Column I times Column III
divided by Column II.
November 2010

LAUS Program Manual 10-6

Step 4. The city and balance of county first-stage employment estimates are summed to a

county total (Column IV, County Total). Thus, the county total for column IV is a sum
and should not be calculated by applying the formula in the column heading to the
county total employment and population.
Step 5. The employment/population index-share is calculated for each city by dividing the

first-stage employment level in the city by the county sum obtained in Step 4. Thus,
Column V equals Column IV divided by the Column IV LMA Total.

The sum of the employment/population index-shares should equal one, except for
rounding. If it does not precisely equal one, the largest share is rounded so that the sum of
the shares is exactly equal to 1. This will ensure that employment from the cities sums to
the county total employment. The shares are then used in the following Worksheet D,
Employment/Population Index-Share Approach to Disaggregating Total Employment, as
follows:

Worksheet D.
Employment/Population Index-Share Approach to Disaggregating Total
City
I
A
B
County Total

E/P Indexshare
II
0.322077
0.249381
1.000000

January
Employmen
t

February
Employmen
t

19,582
15,162
60,800

20,049
15,524
62,250

….
III
…
…
...

November
Employmen
t

December
Employmen
t

21,013
16,270
65,243

21,674
16,782
67,295

Step 1. From Worksheet C, Column V, enter the employment/population index-shares in

Column II for each city listed in Column I.
Step 2. Enter the independent estimate of total employment for the county under the

appropriate month in Column III. The independent estimate for the county is the estimate
that results from the application of the Handbook estimating procedure and adjustment
for additivity to the statewide controls.
Step 3. The city employment/population index-shares are applied to the independent

county employment estimate to arrive at the disaggregated city employment estimates for
the month. The sum of the disaggregated city employment estimates may not add to the
county total because of rounding. If this is the case, the estimate for the largest city
should be adjusted so that the summed estimates equal the county total.

Claims-Based Unemployment Disaggregation
Research has shown that the use of current claimant information in disaggregating labor
market area unemployment to sub-areas is superior to decennial census based
disaggregation because it allows for seasonality during the course of the year and change
during the intercensal period. However, these studies have also shown that a strict
claimant allocation method is not appropriate for total unemployment because claimants
are not representative of the total group of unemployed. This is particularly true of the
entrant-reentrant segment, as these unemployed have a different seasonal pattern to their
joblessness. Disaggregation based solely on claims data generally underestimates urban
areas and inaccurately allocates blacks, youth, and older women.
In an attempt to correct for this, claims data by county of residence are used to distribute
the experienced unemployed component, i.e., those with recent job attachment.

November 2010

LAUS Program Manual 10-7

Decennial census age-group population data may not add to the LMA total because of
rounding. If this is the case, the estimate for the largest county should be adjusted so that
the summed estimates equal the LMA total.
Census population data are used in disaggregating unemployed entrants and reentrants,
under the assumption that the population distribution and age structure of the population
within the LMA do not shift drastically over time. Population aged 16 to 19 is one
element in the disaggregation; the other is population aged 20 and over. Note that these
age groups are those used to calculate the youth population ratio for estimating LMA
entrant and reentrant unemployment using the Handbook procedure. Entrant and
reentrant disaggregation ratios are calculated only once every 10 years. In addition,
differential migration will have an impact on the LMA’s distribution of population.
However, the lack of current data on migration by age group at the county level precludes
any attempt to correct for this.

Required Claims Data for Claims-Based Unemployment
Disaggregation
For all multi-county LMAs, the residency requirement for claims data is the coding and
tabulating of claimants by county (city or town in New England) of residence, within the
State paying the benefits or in border States if the claimant is filing under commuter
arrangements.
For interstate LMAs, the claims data used in disaggregation must be coded for residence
in counties (cities or towns in New England) in contiguous States where commuter
claimant arrangements exist, as well as within the State paying the benefits, in order to
use the claims-based method to disaggregate the intrastate portions of the interstate LMA.
If commuter claimant data are not available by county of residence, the census-share
method must be used to estimate unemployment and employment in each State’s portion
of the interstate LMA. However, in a given State’s intrastate portion, if the State has
claims data by county of residence, the claims-based unemployment disaggregation (and
the employment/population index share method) must be used to disaggregate to the
county level.
The geographic distribution of claimants filing continued claims under State UI and
UCFE certifying to unemployment in the week including the 12th of the month by county
of residence is used to disaggregate the LMA estimate of experienced unemployed to the
county level (city or town in New England). Claimants with any earnings due to
employment in the week includin g the 12th should be excluded from counts used in
disaggregation. Though used for Handbook estimation, Railroad Retirement Board
(RRB) claims should be excluded from the claims counts used in disaggregation.

November 2010

LAUS Program Manual 10-8

Unemployment Disaggregation Procedure and Sequence
The procedure and sequence for unemployment disaggregation is presented below, along
with an example. The example assumes the following Handbook data:
• unemployment, excluding entrants (line 11)=
5,000
• reentrant unemployment (line 13)=
1,600
• new entrant unemployment (line 15)=
400
• total unemployment (line 16)=
7,000
• total LMA claimants without earnings =
6,500
• independent estimate of LMA unemployment(LAUS) = 12,000
LMA Distribution of Population
Sub-area

Claimants

>20 yrs.

16-19 yrs.

1
2
3

2,500
2,250
1,750

25%
30%
45%

20%
35%
45%

Total

6,500

100%

100%

Step 1. For a LMA, determine the percent of Handbook unemployment that is accounted

for by the experienced unemployed, those jobless with recent job attachment, i.e.,
unemployment excluding entrants divided by total unemployment.

Example : 5,000 ÷ 7,000 = 0.71
If any approved atypical adjustment was made to the UI data so that a claims count was
removed from the Handbook claims line leading up to Unemployment Excluding
Entrants, but is added to Total Unemployment, then that figure should be added to
Unemployment Excluding Entrants for purposes of arriving at the experienced
unemployed proportion.
Step 2. Determine the proportion of LMA Handbook unemployment represented by

reentrants unemployment divided by total unemployment.

Example : 1,600 ÷ 7,000 = 0.23
Step 3. Determine the proportion of LMA Handbook unemployment represented by new

entrants unemployment divided by total unemployment.

Example : 400 ÷ 7,000 = 0.06
Note: The proportions obtained in steps 1, 2, and 3 should sum to one (100%).
Example : 0.71 + 0.23 + 0.06 = 1

November 2010

LAUS Program Manual 10-9

Step 4. Apply each of the proportions in steps 1, 2, and 3 to the independent LMA estimate

of total unemployed after additivity and adjustment to statewide controls. This results in a
disaggregation of total LMA unemployment into three parts:

A. experienced unemployed
B. reentrant unemployed
C. new entrant unemploye d.
Example :

A = 0.71 × 12,000 = 8,571.
B = 0.23 × 12,000 = 2,743
C = 0.06 × 12,000 =

686

Step 5. Allocate the LMA estimate of experienced unemployed (estimate A in Step 4) to all

counties (cities and towns in New England) based on the percent distribution of place-of
residence claims data.

Sub-area

Sub-area
Claims

1

2,500

2

2,250

3

1,750

LMA
Claims

÷
÷
÷

6,500
6,500
6,500

LMA
Exp
Unemp

Sub-area
Ratio

=
=
=

x
x
x

0.38
0.35
0.27

8,571
8,571
8,571

Sub-area
Exp
Unemp

=
=
=

3,297
2,967
2,308

Step 6. Allocate the LMA estimate of reentrant employment (estimate B in Step 4) to all

counties based on the percent distribution of the LMA’s population 20 years of age and
older from the 2000 census.

Sub-area

LMA
Reentrants

1

2,743

2

2,743

3

2,743

20+ Pop
Ratios

x
x
x

Sub-area
Reentrants

=
=
=

25%
30%
45%

686
823
1,234

Step 7. Allocate the LMA estimate of new entrant unemployment (estimate C in Step 4) to

all counties based on the percent distribution of the LMA’s population 16-19 years old
from the 2000 census.

November 2010

Sub-area

LMA New
Entrants

1

686

2

686

3

686

16-19 Pop
Ratios

x
x
x

20%
35%
45%

Sub-area
New Entrants

=
=
=

137
240
309

LAUS Program Manual 10-10

Step 8. Derive the total unemployment estimate for each county by summing the

county estimates derived in Steps 5, 6, and 7. The sum of the county
unemployment estimates should automatically equal the LMA total unemployed.
If they are not equal due to rounding, the data for the largest county is adjusted
accordingly
Sub-area

Step 5

1

3,297

2

2,967

Step 6

+
+
+

686
823

Step 7

+
+
+

137
240

3
2,308
1,234
309
Areas sum to independent LMA LAUS estimates.

November 2010

unemployment

=
=
=

4,120
4,030
3,851
12,000

LAUS Program Manual 10-11

Population-Claims Disaggregation of Interstate Areas
In interstate LMAs where all States have the necessary claims data by county of
residence, a “Handbook equivalent” for each intrastate portion is disaggregated from the
total LMA Handbook estimate using the population-claims method. The data for each
intrastate portion are adjusted for additivity to the respective statewide controls. Then the
adjusted intrastate portion is disaggregated to the county level by the population-claims
method. Employment (line 4) and experienced unemployed (line 11) are created at the
LMA level and are disaggregated into the each States’ component parts using the same
techniques discussed above. Reentrant unemployed (line 13) and new entrant
unemployed (line 15) do not get disaggregated since these estimates are already
developed at the component level.
The procedure and sequence for disaggregation is presented below, along with an
example. The example assumes the following Handbook data:
• employment (line 4)=
• unemployment, excluding entrants (line 11)=
• reentrant unemployment (line 13)=
• new entrant unemployment (line 15)=
• total unemployment (line 16)=
• total LMA claimants without earnings =

63,047
23,208
7,736
1,934
32,878
14,621

Interstate Area Employment Disaggregation Procedure
Each year, data are developed to produce the employment/population index-shares, as
follows:

Worksheet A.
Developing Employment/Population Index-Shares for an Interstate Area
2000 CENSUS
Component
Areas
In-State part
Out-of-state part
LMA Total

CURRENT

1st Stage
Employment

E/P
Indexedshare

Employment

Population

Population

(I X III)/II

IV ÷ Σ (IVs)

I
28,875
22,575
51,450

II
56,000
44,275
100,275

III
62,125
50,225
112,350

IV
32,033
25,609
57,642

V
0.555727
0.444273
1.00000

Step 1. Data from the 2000 census on total employment and population are entered in

Columns I and II for both the in-state and out-of-state parts in the LMA.
Step 2. The most recent July 1 population estimates (Current) are entered in Column III,

and the year of reference indicated.
Step 3. The first-stage employment levels are calculated for both the in-state and out-of-

state parts by moving 2000 census employment by the change in the county’s population
since the census. Thus, Column IV equals Column I times Column III divided by Column
II.

November 2010

LAUS Program Manual 10-12

Step 4. The first-stage employment estimates in column IV are summed to a LMA total

(Column IV, LMA Total), Thus, the LMA total in column IV is a sum and should not
be calculated by applying the formula in the column heading to the LMA total
employment and population.
Step 5. The employment/population index-share is calculated for each part of the interstate

by dividing the first-stage employment level in the county by the labor market area sum
obtained in Step 4. Thus, Column V equals Column IV divided by the Column IV LMA
Total.

The sum of the employment/population index-shares should equal one, except for
rounding. If it does not precisely equal one, the largest share is rounded so that the sum of
the shares is exactly equal to 1. This will ensure that employment from the counties sums
to the LMA total employment. The shares are then used in the following Worksheet B,
Employment/Population Index-Share Approach to Disaggregating Total Employment, as
follows:

Worksheet B.
Employment/Population Index-Share Approach to Disaggregating
Employment for an Interstate Area
E/P
Indexshare

Current
Employment

I
In-State part

II
0.555727

III
35,037

Out-of-state part
LMA Total

0.444273
1.000000

28,010
63,047

Component
Areas

Step 1. From Worksheet A, Column V, enter the employment/population index-shares in

Column II for each part listed in Column I.
Step 2. Enter the independent estimate of total employment for the LMA in Column III.

The independent estimate for the labor market is the estimate that results from the
application of the Handbook estimating procedure and adjustment for additivity to the
statewide controls.
Step 3. The employment/population index-shares are applied to the independent LMA

employment estimate to arrive at the disaggregated employment estimates for each part
for the month. The sum of the disaggregated employment estimates for the parts may
not add to the LMA total because of rounding. If this is the case, the estimate for the
largest county should be adjusted so that the summed estimates equal the LMA total.

November 2010

LAUS Program Manual 10-13

Interstate Area Claims-Based Unemployment Disaggregation
Step 1. For a LMA, determine the percent of claims for the in-state and out-of-state parts

by dividing the claims of each part by the total LMA claims.
Component Areas
In-State part
Out-of-state part
LMA Total

Claims

Claims
Ratio

10,381
4,240
14,621

0.71
0.29
1.00

Step 2. Apply each of the proportions in step 1 to the independent LMA estimate of total
unemployed to disaggregate the total LMA unemployment into the in-state and out-ofstate parts.
Component Areas

Claims Ratio

In-State part

0.71

Out-of-state part

0.29

Experienced
Unemployed

LMA Line 11

x
x

=
=

23208
23208

16,478
6,730

Unlike intrastate LMA disaggregation, interstate LMAs do not disaggregate unemployed
reentrants and new entrants because these estimates are already developed for the in-state
and out-of-state parts. These parts are summed to obtain the total reentrants and new
entrants estimates for the interstate LMA.
Component Areas
In-State part
Out-of-state part
LMA Total

Reentrants
Line 13
5,493
2,243
7,736

New
Entrants
Line 15
1,373
561
1,934

Step 3. For each part add the experienced unemployed estimate from step2 to the
reentrants and new entrants estimates to arrive at the total unemployment for each part.
The parts should sum to the interstate LMA total unemployment.
Component
Areas
In-State part
Out-of-state part

Step2
Line 11

New
Entrants
Line 15

Reentrants
Line 13

=
=

23,343

Areas sum to line 16 for the interstate LMA.

32,878

16,478
6,730

+
+

5,493
2,243

+
+

Unemployment
Line 16

1,373
561

9,535

In interstate areas where commuter claimant data are not available for all parts, the
census-share method must be used to estimate both employment and unemployment for
the intrastate portions of the interstate LMA. However, in a given intrastate portion, if the
portion is a multi-county area and the State has claims by county of residence (city and
town in New England), the population claims method must be used to disaggregate to the
county level. In this case, the census-share total unemployment ratio of the intrastate
portion to the whole LMA should be applied to unemployment, excluding entrants, B
November 2010

LAUS Program Manual 10-14

factor unemployment, A factor unemployment, and total unemployment to obtain a
Handbook “equivalent” estimate for the intrastate portion. After this intrastate portion is
adjusted for additivity to the statewide controls, the population-claims method must be
used to disaggregate to the county level.

Disaggregating Employment and Unemployment to
Incorporated Places Using the Population-Claims Method
Simple modifications of the employment/population index share employment
disaggregation and the claims-based unemployment disaggregation enable the
development of labor force estimates for units of local government as small as 2,500
population (according to the 2000 census data), provided claims data are available by
residence of the claimant in all such places in the State. The State may specify the
population level of the places to which this disaggregation method will apply.
In addition, the balance-of-county estimates (derived after subtracting the disaggregated
place estimates) must relate to a specifically defined geographic area. Census data for this
geographic area must be available for disaggregation to other places in the balance-ofcounty area. If such census data are not available, current claims and population data
cannot be used to disaggregate any place within the county. In places where commuter
claimant arrangements exist, further specification of the claims data is required.
Place estimates disaggregated by the population-claims method should be introduced the
first month for which residency claims data are available. Once this method is initiated, it
must be used for the rest of the calendar year. At benchmarking time, the State may opt to
return to census-sharing, in which case the full time series is revised using the censusshare method.
Specification of Population Size for Place Disaggregation

The place level to which the population-claims method is used is established based on the
last decennial population size of cities and towns and is adjusted over the intercensal
period by the annually prepared county and city population estimates.
In the intercensal period, the State should review the total population estimates for all
units of local government issued annually by the Census Bureau to determine whether
population changes have occurred which affect the composition of the size class for place
disaggregation. If, because of a reduction in population, a city falls below the size
specification for place disaggregation using the population-claims method, the State does
not have to revert to the census-share for that place, and may continue to use the
population-claims. In addition, the collection of residency-based claims data does not
have to be extended to other cities in the smaller size class.
A city can move into the size class specified for claims-population disaggregation of
places due to an increase in population. If a State is already using the population-claims
method at the place level, it has one year to develop the residency claims data needed for
disaggregation by coding claims by place of residence for the newly-added city. During
that year, the city would continue to be census-shared. If the residency data are not
developed for the newly-added city after one year, the State cannot use the populationclaims method and must revert to the census-share method for all cities in the size class.
Alternately, the State may avoid reverting to the census-share method by redefining
upward the size class for place disaggregation at the benchmarking time. Then, the

November 2010

LAUS Program Manual 10-15

population claims method can be used in a size class covering larger cities. If a city, due
to a change in population, moves either in or out of the size specified for populationclaims disaggregation of places, the State should notify BLS of this change and the
subsequent methodology changes.
Population-Based Employment Disaggregation

The Census Bureau issues total population estimates for all units of local government
annually. The latest population estimates and the employment/population index-shares
calculated from the census can be used to disaggregate LMA employment estimates
below the county level, using the procedure described in Worksheet A. Balance-ofcounty estimates must also be calculated to allow for the proper application of the
additivity adjustment. The case of disaggregating to a place from a single -county LMA is
straightforward. The employment/population index share procedure, which involves the
use of decennial census employment and population data and annually prepared
population estimates, is applied to the LMA total employment estimate for the month in
question to obtain the place estimate.
In the case of disaggregating to a place from a county in a multi-county LMA, the county
total employment estimate must first be prepared. Population data for counties and places
may not be available on the same time frame. For example, 2005 data may be available at
the county level, but for cities the most recent data may be from 2004. In this case, 2005
data would be used to disaggregate to the county level, and 2004 data for both the county
and cities would be used for disaggregating from the county level to cities within the
county. Thus, the disaggregated county employment is further broken down to the place
(city) by using the place’s most recent population, the county’s population for that same
year, and the place’s employment/population index-shares from the decennial census.
The employment/population ratio is then applied to the county employment estimate for
the month in question to obtain the place employment estimate.
In developing place data using the employment/population index-share approach, States
are reminded that they are to calculate index shares for all pla ces which meet the chosen
population specifications and not just for those which are reported to BLS. States are then
to calculate a rest-of county estimate by subtracting all disaggregated estimates from the
county total. The rest-of-county estimates must relate to a specific geographic area for
which census data exist, so that the census-share procedure can be used for
disaggregation. If this is not possible, the index share approach cannot be used.
Claims-Based Unemployment Disaggregation

Unemployment in a place of 2,500 population or more may be disaggregated directly
from the intrastate LMA (either single -county or multi-county) depending on the
existence of commuter claims arrangements and the availability of commuter claims
coded by city of residence and on the same reference period. Disaggregating directly
from the LMA cannot be done for interstate areas because interstate areas must first be
broken down into intrastate portions. Then, unemployment may be disaggregated directly
from those portions, based on the conditions described above.

November 2010

LAUS Program Manual 10-16

Geographic Basis of Claims Data Used to Distribute Experienced Unemployed

In the claims-based disaggregation, LMA unemployment is disaggregated into three basic
components: the experienced unemployed, unemployed entrants related to experienced
unemployed, and unemployed entrants related to the labor force. The experienced
unemployed component is distributed to areas based on the distribution of claims. In the
case of place disaggregation, if commuter claims arrangements exist and these claims are
coded and tabulated for city of residence, then the experienced unemployed distributor is
as follows:
Claimants residing in the city who file either in their own State or the border State, as a
percent of all residents of the county (in the case of single -county areas) or intrastate
portion (in the case of multi-county interstate areas and New England interstate LMAs)
who file in the State or in the border State.
If commuter claimant arrangements exist, but commuter claims are coded for county of
residence only and not city, then the experienced unemployed distributor is the following:
Claimants residing in the city who file in the State as a percent of all residents in the
county filing in the State.
That is, commuter claims are not used at all. Use of this modified ratio avoids distorting
the city’s share of the experienced unemployed, while allowing the county-coded
commuter claims to be used.
Unemployment Disaggregation Procedure for Cities or Towns

The following disaggregation is used in almost all cases. It is the same procedure for
claims-based county disaggregation described earlier, with the following modifications:
Step 5. Allocations are based on current claims data by city or town of residence.
Step 6. Allocations are based on population 20 years of age and older.
Step 7. Allocations are based on population 16-19 years of age for places with a 2000

population between 2,500 and 10,000.

Use of these ratios parallels the use of county to multi-county LMA ratios. The total
unemployment estimate for the place (Step 8) is then the sum of the disaggregated
experienced unemployment (Step 5) and the new and reentrants (Steps 6 and 7).
The procedure above can be used for places in:
1.) Single or multi-county LMAs not contiguous to a border State;
2.) Single or multi-county LMAs contiguous to a border State without commuter claimant
arrangements; and
3.) Single or multi-county LMAs with commuter claimant arrangements where such
claimants are also coded for city of residence.

Modifications to this procedure are required in (1) interstate areas and (2) single or multicounty LMAs contiguous to a State with commuter claimant arrangements where such
claimants are coded for county of residence only and not city.
In the case of an interstate area or single - or multi-county LMA contiguous to a State in
which commuter claims are coded for county of residence only, the Step 5 proportion
must be based on intrastate claims only, with claims data limited to residents of the
county and the city filing in the State. The proportion becomes the ratio of city residents
November 2010

LAUS Program Manual 10-17

filing in the State to county residents filing in the State. The ratio of city residents filing
in the State to county residents filing in the State and in the border State will
underestimate experienced unemployed in the city. For multi-county areas (including the
intrastate portion of interstate areas), it is necessary to first disaggregate to the county
level before disaggregating to the place in order to use the county-coded commuter
claimant data. Steps 6 and 7 are on the same geographic reference as Step 5, that is, the
city as a percent of the county. Step 8 is the sum of the disaggregated experienced
unemployed (Step 5) and new entrants and reentrants (Steps 6 and 7).

November 2010

LAUS Program Manual 10-18

Use of 2000 Census Data in Disaggregating Labor Force
Estimates—Census-Share Method
The use of 2000 census data for disaggregating labor force estimates may be used a State
has an approved variance in the Cooperative Agreement or when more current data for
disaggregation are not available. This typically will occur for administrative areas such as
Areas of Substantial Unemployment, unique geopolitical areas such as Indian
reservations, and very small areas such as parts of cities.
The census-share method uses employment and unemployment ratios. These ratios are
applied to independent single county LMA estimates after adjustment to State controls, or
to disaggregated sub-LMA levels which were based on those independent LMA
estimates.
When the claims-based unemployment disaggregation and population-based employment
disaggregation are used to disaggregate a place in a county, the balance-of-county area
must be a geographic area for which 2000 census data are available for disaggregating to
other places in the balance-of-county area.

The Census-Share Method of Disaggregation
The census-share method of disaggregation utilizes the ratios of employment and
unemployment in a subarea to the respective total for the larger area according to the
2000 census. These ratios are applied to the current total employment and unemployment
estimates for the larger area. This procedure is based on the assumption that the current
geographic distribution of employment (or unemployment) is the same as that in the
decennial census, or, equivalently, that employment (unemployment) in the subarea has
changed by the same proportion since the census as that in the larger area. The 2000
census-share procedure is used to disaggregate from the county to a subcounty area when
census labor force data are available and a State opts not to use the claims and
population-based disaggregation procedure at the city level. If census labor force data are
not available, contact the BLS regional office to make an atypical request to use the
population-share procedure.

Disaggregation Procedure and Sequence
The procedure and sequence disaggregation using the census-share method, along with an
example, is presented below. For the example, the following data are given:

Census data:
LMA Employment = 20,000
County 1
County 2
County 3

=
=
=

10,000
6,000
4,000

LMA Unemployment = 8,000
County 1 =
4,000
County 2 =
2,400
County 3 =
1,600

November 2010

LAUS Program Manual 10-19

•Independent estimate of LMA employment = 35,000
•Independent estimate of LMA unemployment = 7,000
Step 1. From the 2000 census data, obtain the number of employed in the county.
Step 2. From the 2000 census data, obtain the number of employed in the LMA containing

the county.
Step 3. Divide Step 1 by Step 2. The result is the ratio of the county employment to that of

the LMA as of April 2000

Example :
County 1 = 10,000 ÷ 20,000 = 0.5
County 2 = 6,000 ÷ 20,000 = 0.3
County 3 = 4,000 ÷ 20,000 = 0.2
Step 4. Apply the ratio developed in Step 3 to the total employment estimate for the LMA

for the relevant time period. This will yield the estimate of total employment in the county.

Example : Employment
County 1 = 35,000 × 0.5 = 17,500
County 2 = 35,000 × 0.3 = 10,500
County 3 = 35,000 × 0.2 = 7,000
Step 5. From the 2000 census data, obtain the number of unemployed in the county.
Step 6. From the 2000 census data, obtain the number of unemployed in the LMA

containing the county.
Step 7. Divide Step 5 by Step 6. The result is the ratio of the county unemployment to that

of the LMA as of April 2000.

Example :
County 1 = 4,000 ÷ 8,000 = 0.5
County 2 = 2,400 ÷ 8,000 = 0.3
County 3 = 1,600 ÷ 8,000 = 0.2
Step 8. Apply the ratio developed in Step 7 to the total unemployment estimate for the

LMA for the relevant time period. This will yield the estimate of total unemployment in the
county.

Example :
County 1 = 7,000 × 0.5 = 3,500
County 2 = 7,000 × 0.3 = 2,100
County 3 = 7,000 × 0.2 = 1,400

November 2010

LAUS Program Manual 10-20

11Annual Processing
Introduction
n the current LAUS methodology, Handbook-based and model-based labor force
estimates are revised annually to take advantage of the latest available information.
This process is known as Annual Processing or Annual Benchmarking. State model
performance is formally reviewed by State, regional and national office staff, and
adjustments are made to model specifications when necessary. Then new CPS
population controls, revised Handbook components, and revised State-supplied data are
incorporated into the State and substate estimates. In summary, annual processing
consists of model evaluation and performance review, incorporation of CPS population
controls, collection and incorporation of revised input data, re-estimation of State and
substate estimates, and benchmarking. The sections which follow discuss these processes
in detail.

I

Annual Model Review
A benefit of using a model-based estimation framework is the ability to adapt a State's
model to the changing nature of the State economy and data. The variables in a model are
based on the inter-relationships in the State’s economy, including seasonal patterns and
long-term trends, and the individual nature of the data sources available. The variable
coefficients of the signal-plus-noise models allow the models to adjust gradually to
structural changes in the economy and to discount unusual changes of input data, such as
those resulting from CPS sampling variability. However, for some types of events, such
as severe weather or spurious movement in the CPS, it is important to be able to review a
model’s performance and take direct corrective action. In some cases, intervention
variables are added to the model to restore model performance; in other cases, model
specifications are revised. (See Chapter 6 for a detailed discussion of intervention
variables and model specifications.)
The LAUS model evaluation and performance review is conducted in the fall of each
year. First, a technical memorandum is issued which requests State staff to review their
model performance and provide comments and evaluations to the national office. The
memorandum usually includes a list of suggested topics on model behavior for the States

November 2010

LAUS Manual 11-1

to consider while reviewing their model performance. In addition, States are asked to
provide information about their economy which might help to explain model behavior.
At the same time, the Statistical Methods Staff (SMS) review the model performance
using statistical tests for diagnostic evaluation and outlier detection. This review focuses
on statistical measures of model performance, and is in addition to the battery of
statistical tests which are run on the models each month. The tests help to determine
whether any changes to current model specifications or outlier interventions are
necessary. SMS shares the results of this research and their proposed actions with
States, either via technical memorandum or in presentations during the annual State/
regional meetings.
LAUS national office staff monitors model performance as part of their monthly duties.
In addition, they summarize State issues and concerns regarding model performance and
present resolutions and answers during the annual State meetings. Questions and issues
raised by the States are also responded to in a formal LAUS technical memorandum,
generally issued just after the conclusion of annual processing activities.

Population Controls
At the beginning of each year, new CPS population controls are introduced for use in
Division, State, and substate model estimation. These controls reflect both new data for
the most recent year and revisions to data for earlier years. The new and revised controls
are developed by the Census Bureau and delivered to BLS in late January.
Resident population at all levels of geography is estimated by updating a base population
from the census via estimates of the components of population change, consisting of
births, deaths, and migration. The CPS universe, defined as the civilian noninstitutional
population, is then estimated by subtraction of the resident military and institutional
populations, primarily nursing homes, prisons and jails, mental hospitals, and juvenile
facilities. Below the national level, this procedure is supplemented by direct estimates of
group quarters populations and, in some instances, student and military populations.
Population controlling occurs when the sample -based monthly CPS labor force estimates
are adjusted so that they are consistent with these independently derived population
estimates. Adjusting (controlling) the CPS sample -based labor force estimates to be
consistent with independently derived population estimates reduces the variability of the
CPS estimates, thus improving their quality.
There are several ways CPS population controls affect LAUS estimates. For model-based
estimates, the monthly impact is via the CPS inputs to the model estimates. For current
estimation, monthly CPS population controls are incorporated into the CPS estimates
through the second-stage ratio adjustment step of the CPS estimation process. (See
Chapter 2 for a description of the second-stage estimation process.) For substate
estimates, the monthly impact of CPS population controls is less direct, and occurs due to
the additivity adjustment of substate estimates to their respective statewide or balance-ofState totals.
Annual CPS population controls are incorporated into CPS estimates through the revision
of monthly CPS labor force estimates at the end of the year. During annual processing,

November 2010

LAUS Manual 11-2

the revised CPS data are incorporated into the LAUS estimates when the models are reestimated.

How Population Estimates are Calculated
Current estimates of the national population by age, sex, race, and Hispanic origin are
derived by quarterly updates of the resident population (enumerated in the last census)
using components of population change. This process uses the following simple formula
to update each category.
Revised Population = Enumerated Base Population

+ Births to U.S. resident women
- Deaths of U.S. residents
+ Net international migration
Births and deaths by sex, race, and Hispanic origin are obtained from the National Center
for Health Statistics (NCHS), generally through the calendar year two years prior to the
last July estimate date. Distribution by sex, race and Hispanic origin is projected to the
last July estimate date. The projected distribution is applied to a preliminary series of
births and deaths, also obtained from NCHS.
International migration, in its simplest form, is any change of residence across the
borders of the United States. The net international migration component of the
population estimates combines four parts: (1) net international migration of the foreign
born, (2) net migration of natives to and from the US, (3) net migratio n between the US
and Puerto Rico, and (4) net movement of the Armed Forces population to and from the
US. Net immigration of the foreign-born population is estimated in two parts,
immigration and emigration. The estimate of immigration utilizes information from the
American Community Survey (ACS) on the reported residence of the foreign-born
population in the prior year. The foreign born who reported living abroad in the year
prior to the survey are considered immigrants. For years 2001 – 2004, where ACS data
are unavailable, estimates are derived by linear interpolation between the 2000 Census
and the 2005 population estimates.
Emigration of the foreign born is estimated using a residual method. The foreign-born
household population from the census is “aged” using NCHS life tables to obtain the
expected population for the four years prior to the year being estimated (2005, 2006,
2007, and 2008 for the population estimates developed for 2009). This expected foreignborn estimate is then compared to the foreign-born population estimated by the ACS for
each of the years estimated. Subtracting the estimated from the expected population
produces a residual, which serves as the basis for emigration rates for specific time
periods.
Net international migration of the foreign-born population is estimated by subtracting the
number of emigrants from the number of immigrants. Age, sex, race, and Hispanic origin
information is estimated for each group separately using data from census 2000 and the
three-year ACS estimates from 2005 forward.
Net migration between the US and Puerto Rico is also estimated in two parts,
immigration and emigration. For 2005 and later years, the ACS and the Puerto Rico
Community Survey are the sources of information for the migration estimates.

November 2010

LAUS Manual 11-3

Net movement of Armed Forces and their dependents is estimated using data from the
Defense Manpower Data Center (DMDC) of the Department of Defense. DMDC
provides data by age, sex, and Hispanic origin; for race data, the Census Bureau applies
the race distribution by Hispanic origin from the Census 2000 active military population
to this Armed Forces movement overseas component.
The population of the States must be estimated using less direct methods than those used
to derive population estimates for the Nation as a whole because interstate migration, a
large component of the change in State populations, cannot be accounted for as directly
as births, deaths, and legal immigration. Administrative records are used to derive
estimates of domestic migration.
For the under age 65 population, person-level data from Federal tax returns are used to
identify how many tax filers and their dependents moved from one county to another
between tax filings. The ratio of IRS domestic migrants to IRS non-migrants is applied
to all potential migrants within each county to derive total domestic migration estimates.
Domestic migration in the age 65 and over population is derived using Medicare
enrollment data. Changes in Medicare enrollment are assumed to mirror changes in the
total age 65+ population. Year-to-year change in Medicare enrollment is used calculate
year-to-year change in the total population. The resulting estimates are compared to data
on total deaths and international migration in the age 65 and over; any residual is deemed
domestic migration.
To create population estimates to serve as controls for surveys such as the CPS, the total
population estimates are adjusted to remove armed forces personnel, the institutionalized
population, and persons under the age of 16.
(For more information on the development of national and State population estimates, see
the US Census Bureau documentation at
http://www.census.gov/popest/topics/methodology/2009-nat-meth.pdf and
http://www.census.gov/popest/topics/methodology/2009-st-co-meth.pdf.)

November 2010

LAUS Manual 11-4

State Annual Population Controls
Each January, the Census Bureau provides revised population estimates for each State.
These estimates use short-term projections, plus the previous year’s estimates with
revisions to both the previous vintage and the within-year monthly estimates. There are
three types of revisions.
1. Components of change are revised because of the availability of updated input data.
In addition, new projections are produced for the very recent dates for which no data
are available . These revisions are routine and generally affect only recent years’
estimates.
2. The Census Bureau may update the methods of estimating the components of
population change and special populatio n stock estimates (institutional and
noninstitutional group quarters and military) used in the estimates procedure. Such
revisions generally affect the data series cumulatively from the previous census
forward.
3. In addition to updating the method of estimating the components of population
change, the Census Bureau may revise the method of estimating population arising
from the components of change and special population stock estimates. This type of
change also affects the series from the census date forward.
Once BLS receives the population controls from the Census Bureau, they are used to
adjust the State's CPS labor force data. This is done by multiplying each month’s CPS
employment and unemployment estimates by the ratio of that month’s revised population
value divided by the original population value. Under normal circumstances, the CPS
unemployment rate is not affected because both the unemployment estimate (the
numerator of the rate calculation) and the labor force estimate (the denominator of the
rate calculation) have been adjusted by the same proportion.

Annual Population Revisions Affecting Substate Area Estimates
Substate area estimates benefit from the monthly and annual CPS population revisions
through the additivity process which assures consistency with the State totals. Substate
estimates utilize county and place total population data in the disaggregation of labor
market area estimates into smaller geographic entities. Total population estimates for
counties, incorporated places, and minor civil divisions are revised by the Census Bureau
each year.

November 2010

LAUS Manual 11-5

Annual Re-Estimation
Each year States are provided the schedule for benchmarking activities in a technical
memorandum. States are instructed to replace the model input data with revised Current
Employment Statistics (CES) employment, striker, unemployment insurance (UI)
claimant, and Unemployment Compensation for Federal Employees (UCFE) claimant
data for every period for which they have revisions or corrections. The CES nonfarm
wage and salary estimates should reflect the most recent Quarterly Census of
Employment and Wage (QCEW) benchmark and include any changes beyond the regular
two years of CES benchmarking. Claims counts for the current year should be updated
wherever possible.
At the end of each calendar year, the LAUS Census division models are re-estimated and
their not-seasonally adjusted estimates are forced to sum to the updated national CPS notseasonally-adjusted levels of employment and unemployment. After revised input data
have been provided, the statewide model-based estimates are re-estimated and
benchmarked to the newly revised Census division estimates. The selected area and
balance of State models are then also re-estimated and the not-seasonally adjusted
estimates are benchmarked to their newly revised statewide estimates.
Revised seasonally-adjusted estimates are also pro-rata adjusted. However, these
estimates are not controlled in the same manner as are the not-seasonally adjusted
estimates. Rather the same benchmark (or pro-rata) adjustment as was used for the notseasonally-adjusted estimates is applied directly to the seasonally adjusted estimates.
Seasonally adjusted estimates are therefore not additive to the estimates at higher levels
of geography.
State groupings for model-based annual processing are on a Census division basis
because real-time benchmarking is performed at the division level. Once all States in a
division have entered their input data and the national office has validated them, the
historical series for all States in the division are re-estimated and benchmarked.
Typically, five years of model-based estimates are revised during annual processing.
In re-estimation, the entire time series is used to re-estimate every observation. The
estimation process is run forward from the beginning of the time series, run backward so
that earlier observations benefit from later data, and then run forward again to the end of
the year. This is possible because LAUS models use a Forward Filter to modify each of
the model’s coefficients with the addition of each monthly observation.
The Forward Filter acts as weighting mechanism, which allocates how much a model’s
coefficients will change (and thus the estimates) with each new period’s data. Because
the Forward Filter works by evaluating each successive observation, one after another,
the models can produce estimates both forward and backward through time. During the
initial forward pass, each successive estimate incorporates all of the information from the
earlier months in the time series. At the end of the time series, the estimation process is
performed backward through time, so that each past month’s estimate can benefit from
the more recent data. Finally, the process is performed moving forward through time
again, so that information from the first two passes can be incorporated into the entire
time series.

November 2010

LAUS Manual 11-6

During annual processing, the smoothed seasonal adjustment process which creates the
official LAUS estimates uses a symmetric smoother of thirteen months, centered on the
current estimate. In months that do not have six subsequent observations – i.e., July
forward of the most recent production year – all future observations are used, with the
weights adjusted according to the number of available months remaining. December of
the most recent production year is smoothed using only the current and six previous
months’ data, identical to concurrent estimation. This method eliminates methodological
discontinuities between December and January estimates.

November 2010

LAUS Manual 11-7

Annual Processing of Substate Area Estimates
As in regular monthly estimation, substate area benchmarking is limited in data sources.
However, substate estimates are improved by incorporating updated source data, revising
prior inputs, adjusting for changes in UI procedures and coverage, incorporating changes
in geographic definitions, and adjusting the areas to revised State estimates through
additivity. In some cases, such as Handbook agricultural employment estimation, annual
benchmarking of the series is most critical, given the lack of ideal source data for
generating monthly changes at the labor market area level.
Benchmarking of substate area estimates is conducted during the first four months of
each year for the previous two year's estimates. Since five years of model-based
estimates are revised during annual processing, the remaining three years of substate
estimates are ratio-adjusted to conform to the new model-based control totals.
Occasionally, more than two years are revised, typically following changes in
methodology or large revisions to population or other input data. The requirements and
schedule for this activity are provided to the States via LAUS technical memoranda. In
addition, specific instructions relating to the LSS Plus software to be used in
benchmarking are provided to the regions and States, typically in January.

Incorporation of Substate Data Updates
The table below lists a number of Handbook inputs data, the reason for the revision or
update and the source of the updated.

Input

Reason for update

Source

Handbook Employment:
Agricultural factors

New agricultural data available

CES employment

Latest QCEW benchmark

States

BLS

Atypical

New data available

States

Handbook Unemployment:
UI and UCFE claims

Revised counts

States

RRB claims

Revised counts

BLS

New entrants distribution ratios

New age group populations

BLS

Statewide new entrants

Updated population controls

BLS

Reentrants distribution ratios

New age group populations

BLS

Statewide reentrants

Updated population controls

BLS

Additivity:
Statewide estimates

Controlled to CPS

BLS

Disaggregation:
Employment/Population Ratios
New entrant disaggregation
ratios

New population data

BLS

New age group populations

BLS

Reentrant disaggregation ratios

New age group populations

BLS

November 2010

LAUS Manual 11-8

The two most important Handbook updates are to the CES employment and the UI claims
data, both provided by the State.
The monthly sample -based employment estimates from the CES program are
benchmarked each year to more complete payroll counts from the QCEW program. CES
estimates are revised back to the previous year’s benchmark and brought forward, using
link-relative extrapolation to the end of the most recent calendar year. This means that
employment estimates used in LAUS estimation are typically revised for the preceding
twenty-one months.
For areas outside the scope of the CES program, QCEW data should be used for all time
periods available. For the months after the most recent QCEW update, States should use
the same projection procedures used to generates inputs for monthly estimation.
However, all estimates based on QCEW data (e.g., projection of QCEW data by
statistical modeling or chained monthly change factors) should be re-estimated using all
available QCEW data, rather than using the inputs created during monthly estimation.
UI and UCFE claimant data for substate areas are also revised at benchmarking time.
Continued claims without earnings are reviewed by State staff for the benchmarking
period. Area claims counts are corrected, finalized, and checked against the statewide
total. This is the opportunity to correct errors and omissions uncovered during the course
of the year. It is also the time to incorporate the effect that changes in UI law or practice
have had on the LAUS estimates.
The CPS State annual average data aggregated by agricultural region is used as the
annual benchmark for agricultural employment. An annual change factor is created using
the recently completed year average over the prior year annual average. Calculations of
the monthly factors use the current July CPS number and monthly agricultural factor as
the base in the equations because the majority of agricultural activity nationwide is
experienced in July. The monthly factors are created for the July-to-July period using the
July factor in the most recently completed calendar year; months after July are created
using the July numbers from the current year. Consequently, the series revision for
agricultural employment is for the July two years ago to the most recent July, and
extended for August to December of the most recent year.
The new entrants distribution ratios (L14) and the reentrants distribution ratios (L12),
which are needed to develop the handbook new entrants and reentrants estimates, are
revised each year using the latest intercensal population estimates available from the
Census Bureau. The L14 ratios, which allocate the Statewide new entrant estimates to
LMAs, are updated using the latest 16-19 year old age population data, while the L12
ratios, which apportion the Statewide reentrant estimates to the LMAs, are updated using
the latest 20 year and older population data. New L14 and 12 ratios are produced by the
national office for all LMAs and are made available to States during annual processing.
New entrants (L15) and reentrants (L13) estimates are benchmarked to incorporate new
Census population controls. Benchmarked entrants are created using a three step process.
First, the CPS new entrant and reentrant estimates are population controlled. Next, the
five year moving averages to compute LAUS entrant inputs are recalculated on the
population controlled estimates. Finally, the LAUS inputs are controlled to national
estimates of new entrants and reentrants and are provided to States.

November 2010

LAUS Manual 11-9

The employment-to-population ratios used for employment disaggregation are updated
with revised population data from the Census Bureau. Population data for counties,
incorporated places, and minor civil divisions are revised each year. The revised
population data are incorporated into employment/population indexed-share
disaggregation ratios (R01) by the national office. The revised ratios are then provided to
the States.
Ratios that disaggregate new entrants (R02) and reentrants (R03) to the county and city
levels are also revised by the national office. The revisions incorporate new age-group
population data from the Census Bureau.
In addition, annual processing of substate area estimates reflects the annual processing of
model-based LAUS estimates. The modeled series provides the control totals for
additivity adjustment of area estimates.

Incorporating Changes in Geographical Areas
Changes in population levels, both within metropolitan and non-metropolitan labor
market areas and in cities and towns, can impact LAUS geographically defined areas.
Population growth in areas can result in the creation of new a labor market area. Growth
in areas contiguous to an LMA can result in a redefined LMA. These changes are
identified by the US Office of Management and Budget near the end of the calendar
years. In addition, cities and towns newly identified as having populations of 25,000 or
more are also included as LAUS estimating areas during annual processing.

Substate Processing
BLS and States are each responsible for the creation and provision of certain data inputs,
as well as certain activities. The BLS provides the States with title code listings
(geographical listing of all areas a state re-estimates), new ratios for disaggregation, the
handbook benchmark control file (State revised estimates from STARS), revised
population data, annual agricultural factors, revised new entrant and reentrant statewide
estimates, and revised entrant distribution ratios. BLS staff also review the transmitted
data for accuracy and consistency.
States provide inputs for nonagricultural wage and salary employment, labor disputants,
unemployment insurance claimants (regular UI and UCFE) and final payment recipients.
States also enter inputs, generate revised estimates, and transmit revised estimates to
other States as part of interstate data exchange and to the national office for review.
Revised substate data undergo a thorough review to ensure accuracy and consistency. In
reviewing estimates prior to transmission, States should use the Benchmark Compare
edits in LSS Plus. These edits will compare the newly benchmarked estimates created in
the current annual processing with estimates for the same years produced during the prior
year, as a share of the statewide (or Balance-of-State) totals. States are urged to research
large changes displayed in the edit output and provide comments or correct the data,
whichever is appropriate. (See the appendix to this chapter “Guidelines for Reviewing
Benchmark Compare Edit Results” for more information on evaluating the results of the
Benchmark Compare edits.)

November 2010

LAUS Manual 11-10

Appendix to Chapter 11:
Guidelines for Reviewing Benchmark Compare Edit Results
Background
The Benchmark Compare edit is used to evaluate newly benchmarked estimates by
comparing them with the pre-benchmarked estimates for the latest year or with
previously benchmarked estimates for earlier years. For each year, the two sets of data
for the same time period and same geography are compared. Given the assumption that
the prior set of data is correct, changes in the new set of data are examined in the context
of the LAUS estimation methodology and the expected revisions to inputs.
During annual processing--that is, benchmarking--the statewide (and Balance-of-State
where appropriate) controls change due to new population controls, reestimation, and
benchmarking to new national CPS totals. However, since the Benchmark Compare edits
look at an area’s employment and unemployment le vels as shares of the State (or
Balance-of-State) totals, the changes at the statewide level have no impact on the
numerical values displayed in the edit results. In other words, if substate inputs do not
change at all, the percentage changes in this edit will all be zero regardless of the
magnitude of the end-of- year benchmark revision at the State (or Balance-of-State) level.
[This is similar to the monthly Questionable Data Edits for labor market areas (LMAs),
which also examine the change in an area’s share of the State (or Balance of State).]
Changes shown in the Benchmark Compare edit s arise from revisions to substate inputs
of different magnitudes (or directions) in different areas that affect both Handbook
estimation and disaggregation.
The edit results need to be evaluated in the context of the following: (1) the inputs that
are expected to change; (2) the normal expected magnitude of those changes; (3) any
information about abnormalities for a particular area or time period (such as major claims
data corrections or large population estimates revisions); (4) the size and nature of the
area (for example, a college town or agricultural center); (5) the methodology used to
derive the estimates for the area; and (6) the year in question.
Substate area input changes
A. Employment
Nonagricultural wage and salary employment (M01) generally is revised for the latest 1821 months. For substate areas for which Current Employment Statistics (CES) estimates
are created, earlier time periods may be corrected. The “Recent Corrections” webpage
maintained by the CES program lists corrections (that is, changes other than routine
revisions) at http://www.bls.gov/sae/saerevisions.htm. In metropolitan areas, the CES
revision is generally 2-3 percent or less. In micropolitan or small LMAs, the percentage
revision can be larger, though changes approaching 10 percent should be researched for
all but the very smallest LMAs. Changes may follow a quarterly pattern (as in the
QCEW) and, if large, usually should be consistently in one direction--up or down, not a

November 2010

LAUS Manual 11-11

mixture. This component affects LMAs directly and all disaggregated areas within them
indirectly. Changes to this component have a moderate impact, especially in small areas.
Agricultural employment changes from August in the second-to-last year in the series
forward, based on estimation factors provided by BLS. Note that a fair number of States
use atypical procedures for estimating agricultural employment and, thus, may show a
different pattern. This component affects LMAs directly and all disaggregated areas
within them indirectly. The impact of changes to this component can be substantial in
small, agricultural areas, but is virtually zero in large metropolitan areas.
Employment-population indexed-share disaggregation ratios are revised by BLS for all
counties in multi-county areas and cities that use this procedure. This input affects all
months of a given year by the same degree, with the impact varying noticeably by year in
relatively fast-growing (or rapidly declining) areas. This revision has a modest impact,
but it can be proportionately large, especially in small areas where the population is
changing at a pace very different from that of other counties in the LMA or cities in the
county. Changes tend to be largest in the latest two years, reflecting population data
availability for one additional year.
B. Unemployment
Continued claims inputs for the latest year should be based on a second or third, and
generally more complete, count of claims. This would generally raise Handbook
unemployment levels, which translates into some areas increasing and others decreasing
after additivity. The magnitude of the difference between first and second or second and
third counts of claims at the statewide level, available from STARS, can be used as a
guide. This input affects LMAs and claims-disaggregated areas directly and areas within
them indirectly. The impact is generally moderate, but may be large for small areas.
New statewide entrant totals generally are used for the latest two years. This should
affect the entrant component in all areas in the same direction, by the same proportion. It
has very little impact in most areas, though it has created large percentage differences in
extremely small areas.
Beginning in 2007, year-specific ratios were used for the distribution of unemployed
entrants from statewide totals to additivity areas and the disaggregation of entrants to the
county and city levels based on intercensal age-group population estimates produced by
the Census Bureau. The new ratios better reflect current population trends, account ing
for migration and mortality, as opposed to merely mortality in the prior approach. The
impact of annual revisions to this component is small in most areas.
Edit results suggesting problems
Large percentage changes often indicate problems. These include, but are not limited to,
changes of over 10 percent. Generally, the largest changes historically have been
employment revisions for the latest 18-21 months in very small areas, though

November 2010

LAUS Manual 11-12

unemployment revisions based on second counts of claimants in the most recent year
may be similarly large. Any of the following may indicate problems: (1) employment
changes of more than about 2 percent in metropolitan (CES) areas; (2) employment
changes of more than about 3 percent in any LMA for months before March of the
second-to- last year; (3) large changes in some months and very small changes in other
months of the same year (although estimates for December 2010 may have larger
revisions than prior months because the December inputs were not revised as part of
monthly processing); and (4) large variations in the pattern of differences, particularly for
disaggregated areas.
Widespread lack of change, when changes are expected, can indicate problems as well.
Examples include: (1) zero changes to LMA employment, especially for small areas-this actually happened when one State did not update nonfarm wage and salary
employment for small areas; (2) zero changes to disaggregated county- level employment-this may indicate the State mistakenly used the old disaggregation ratios rather than the
revised ones; (3) zero changes to just unemployment in the latest year--this may indicate
that the State neglected to use revised count s of claims data; and (4) zero changes to both
employment and unemployment in December [or any month(s)] of the latest year--this
actually happened as a result of a State mistakenly retransmitting an old “production”
LAUS data file and the data over-writing benchmarked data in LNS.
For the third-to- last (or earlier) year in the series--not required as part of 2010 annual
processing--no input revisions are generally expected, though, total nonfarm wage and
salary employment from the CES program may be corrected for certain areas.
Census-shared cities should follow the same pattern of the areas from which they are
disaggregated. If estimates for the derivation area change, it is not necessary (or
appropriate) to question the census-shared area. However, if a census-shared area does
not move in line with its derivation area, that indicates a problem and the census-share
ratios and geographic relationships should be checked carefully.
Notes
The Benchmark Compare edit output options consist of screens showing (1) an
employment change summary (“Employment” under the “Bmk Com” menu in LSS Plus),
(2) an unemployment change summary (“Unemployment”), and (3) employment and
unemployment change details (“Detailed Table”). While not required, States are
encouraged to enter comments for areas that appear questionable. The summaries are
useful for obtaining an overview of the changes, including the pattern by month for an
area, while the detail is useful to see the specific employment and unemployment values,
both before and after benchmarking. The summaries also include a separate window that
shows details for a single month and area. In LNS (but not LSS Plus), the Benchmark
Compare detail edit output can be sorted, which allows the analyst to easily identify the
largest changes in each direction in relative terms. In LSS Plus, the Detailed Table
output can be saved (File > Save As …) in numerous formats, including Excel, and the
resulting file can be sorted in a similar manner.

November 2010

LAUS Manual 11-13

12 STARS Output Tables
Introduction

A

ll States use the monthly web-based system called STARS (State Time Series
Analysis and Review System) each month to produce their labor force estimates
and to transmit the estimates to the BLS national office.

In producing estimates, States provide input data such as CES employment, strikers, and
unemployment insurance claims. These data are used not only to produce the estimates,
but are also stored in a national office database and are available through STARS, along
with data from the CPS, for use in labor force analysis , model interpretation, and other
research.
Each time STARS is run, it provides both BLS and State analysts with output containing
a series of tables and graphs with essential information for studying employment trends,
preparing releases, or understanding the nature of a month-to-month change in model
estimates.
The primary functions of STARS are to:
• integrate State/BLS data entry.
• calculate estimates of labor force, employment, unemployment, and the unemployment
rate for the current month and revised estimates for the previous month for States and
selected areas. (See Chapter 8 for geographic information on area and balance of State
models.)
• provide error measures, analytical charts and tables.
• allow transmittal of the estimates to BLS.
CPS data are loaded into STARS at the national office as soon as the monthly national
press release is issued. State analysts can enter their inputs when they become available,
review their listings, and redo them if an error is found. States have the option to run
estimates with preliminary numbers before the actual data are available without

November 2010

LAUS Program Manual 12-1

transmitting the estimates to BLS, and, when they are correct and verified, final LAUS
estimates are transmitted to BLS through the STARS system.

A STARS User's Guide is available to assist users in creating and updating their monthly
model-based estimates using the web-based STARS interface. The user’s guide is
designed to introduce new users to the STARS interface and provide them with the basic
skills required for operating STARS.
The latest version of the STARS User's Guide is available at the STARS website under
the Help link on the login screen menu. Users are not required to log into STARS to
access it. The user’s guide can be viewed online, printed, or downloaded by individual
chapter or in its entirety in PDF format. The guide can also be referenced while a user is
logged into the system.

November 2010

LAUS Program Manual 12-2

STARS Review Estimates
The Review Estimates link in the STARS Monthly Processing Menu
enables users to review current, finalized estimation output tables, as
well as historical estimation output tables for the selected State, substate areas (i.e., metropolitan areas, balance of State) and Census
Division. Official STARS estimation output tables are available from
January 2005 through the most recent reference month. The output
tables can be viewed online, downloaded, or printed. (See pages 3-24
through 3-29 of the STARS User’s Guide.)
Below is the header page that precedes the estimation output table s. It
provides basic information about the estimation run. Starting at the top, it displays the
reference month/year and the State. Then it lists the dates and times for when the inputs
were entered, when the division estimates were created, when the inputs were finalized
and when the division was finalized. Next, it provides a quick look at the inputs that
were entered for the current and previous months. There is also a State Comments
section that allows a State to document their run in the output. In this example the
analyst commented on a possible employment input error. States are encouraged to use
this section to identify their runs. These comments are also very helpful to BLS analysts.

November 2010

LAUS Program Manual 12-3

Following the header page are 12 tables and 6 sets of charts.
•

Table 1 includes the basic estimates which include levels and ratios for the labor
force, employment and unemployment, along with the error ranges for the
unemployment rate and the over-the-month changes.

•

Table 3 provides the over-the-year changes in the estimates along with the
standard errors for the over-the-year-changes.

•

Tables 2, 7 and 8 provide error measures for the estimates and the model
components.

•

Tables 4-6 display the changes in the model components and the CPS.

•

Tables 8 and 10 provide information on the data inputs.

•

Tables 9 and 10 show the seasonal factors. Table 11 shows the model
diagnostics and prediction error.

•

Table 12 displays the pro-rata benchmark factors.

•

Figure 1 charts the unemployment rate.

•

Figure 2 graphs the unemployment level and figure 3 shows the employment
level. Figures 4-5 are dedicated to the seasonal factors for unemployment and
employment estimates and their inputs.

•

Figure 6 displays the CPS population.

Explanations and examples of the tables and figures are provided on the following pages.

November 2010

LAUS Program Manual 12-4

STARS Table 1: Year-to-Date Model Estimates
Table 1 is comprised of two sets of tables that present year-to-date information on the
LAUS model estimates for the Smoothed Seasonally Adjusted (SSA) series and the not
seasonally adjusted (NSA) series. Tables 1a (SSA) and 1b (NSA) provide a quick
comparison of current estimates to earlier estimates.
Tables 1a and 1b indicate when the monthly changes are significant at the 5 percent level
and the 10 percent level. The error range for the unemployment rate at the 90 percent
confidence interval is also displayed. Developing trends and month-to-month changes
can be observed. A comparison of the smoothed seasonally adjusted and not seasonally
adjusted series can be made. Table 1a is shown below.

** Significant change at 5% level
* Significant change at 10% level
+ 90% Confidence Interval

Tables 1c (SSA) and 1d (NSA) show the over-the month level changes and the over-the
month percent changes for the labor force, employment, unemployment and the
unemployment. Table 1c is shown below.

November 2010

LAUS Program Manual 12-5

STARS Table 2: Standard Errors for Year-to-Date Model
Estimates
This table shows the monthly standard errors for each of the labor force components
displayed in Table 1. The standard error refers to the variability of an estimate and is
used in the construction of confidence intervals.

November 2010

LAUS Program Manual 12-6

STARS Table 3: Over-the-Year Changes
Table 3 consists of four tables. The first two tables show over-the-year changes for the
each of the basic types of labor force estimates for the smoothed seasonally adjusted
series (Table 3a) and the unadjusted series (Table 3b). Both the level of change and the
percent change are given. The data for all years prior to the current year reflect the
annual updating. The over-the-year changes give an indication of the state's labor force
trends.

The second set of tables provides the standard errors for the over-the-year-change
associated with the above tables. The standard errors for the over-the-year-changes in the
smoothed seasonally adjusted series are found in Table 3c and the standard errors for the
unadjusted series are in Table 3d.

November 2010

LAUS Program Manual 12-7

Tables 4-6: Components of Change
These tables show the components of the unemployment rate, unemployment and
employment of both the model and the CPS and how they changed.
Table 4 displays the components of change for the unemployment rate. Table 4a contains
the level, trend, seasonal, and smoothed seasonal change for the model. Table 4b
contains the changes for the CPS, the signal and the noise. The same items are shown for
the unemployment level in Table 5 and the employment level in Table 6.
Since each model estimate is the sum of its variable components, the analyst can see the
influence that each variable has on the total estimate by examining the components of
change. It may be that one input variable is the primary influence in the current over-themonth change in the estimate.
The analyst can determine if the variable components are behaving in their "normal" way
for this time of year by comparison with their past behavior. Historical components of
change can be computed using data from the Extract Data link in the STARS Monthly
Processing Menu.

** Significant change at 5% level
* Significant change at 10% level

November 2010

LAUS Program Manual 12-8

** Significant change at 5% level
* Significant change at 10% level

** Significant change at 5% level
* Significant change at 10% level

November 2010

LAUS Program Manual 12-9

Tables 7-8: Standard Errors
Table 7 shows the standard error for the model components of change for both the
smoothed seasonally adjusted and not seasonally adjusted series.

Table 8a lists the monthly CPS estimates and table 8b shows the standard error for
the CPS levels and changes.

November 2010

LAUS Program Manual 12-10

Tables 9-10: State Data, Trend and Seasonal Factors
Table 9 displays the seasonal factors that are applied to the unemployment rate,
the unemployment level, the employment level and the employment-population
ratio to create the benchmarked seasonally adjusted estimates, which are inputs to
the official smoothed seasonally estimates.

In table 10 the levels, trends and seasonal factors of the UI claims and CES inputs
are displayed.

November 2010

LAUS Program Manual 12-11

Table 11: Diagnostics, Prediction Error and State Inputs
Table 11 is useful for checking new data for unusual values. Before receiving
new data into the model, values are predicted from the accumulated historical
experience of the CPS data and State inputs. Observations that are far from the
predicted value can identify outliers, which are unusually large changes; inliers,
which are unusually small changes; or incorrect data due to mistakes.

November 2010

LAUS Program Manual 12-12

Table 12: Pro-Rata Benchmark Factors
State levels are pro-rated to sum to the Division totals. Table 12 displays the prorata factors applied to the State model estimates.

November 2010

LAUS Program Manual 12-13

Figures 1-6
Also included in the STARS output are charts that visually display the LAUS
estimates, the input data and the seasonal factors.
Figure 1 Unemployment Rate
Figure 1a charts the smoothed seasonally adjusted LAUS unemployment rates,
the benchmarked seasonally adjusted LAUS unemployment rates, the
unbenchmarked seasonally adjusted LAUS unemployment rates and the
seasonally adjusted claims rates. Figure 1b displays the unadjusted LAUS
unemployment rates, the CPS unemployment rates and the claims rates.
Figure 2 Unemployment
Figure 2a charts the smoothed seasonally adjusted LAUS unemployment levels,
the benchmarked seasonally adjusted LAUS unemployment levels, the
unbenchmarked seasonally adjusted LAUS unemployment levels and the
seasonally adjusted claims levels. Figure 2b displays the unadjusted LAUS
unemployment levels and the CPS unemployment levels and the claims levels.
Figure 3 Employment
Figure 3a charts the smoothed seasonally adjusted LAUS employment levels, the
benchmarked seasonally adjusted LAUS employment levels the unbenchmarked
seasonally adjusted LAUS employment levels and the seasonally adjusted CES
employment levels. Figure 3b displays the unadjusted LAUS employment levels,
the CPS employment levels and the CES employment levels.
Figures 4-5 Seasonal Factors
Seasonal factors indicate the expected seasonal variation in the series. Often
differences in the seasonal patterns help to explain the difference between the
CPS and State inputs series in their direction and magnitude of change.
Figure 4a shows the seasonal factors and the seasonal means for the LAUS
unemployment series. Figure 4b shows the seasonal factors and the seasonal
means for the claims series.
Figure 5a shows the seasonal factors and the seasonal means for the LAUS
employment series. Figure 5b shows the seasonal factors and the seasonal means
for the CES employment series.
Figure 6 CPS Population
Figure 6 charts the CPS population estimate.

November 2010

LAUS Program Manual 12-14

Figure 1: Unemployment Rate

November 2010

LAUS Program Manual 12-15

Figure 2: Unemployment Level

November 2010

LAUS Program Manual 12-16

Figure 3: Employment Level

November 2010

LAUS Program Manual 12-17

Figure 4: Unemployment Seasonal Factors

November 2010

LAUS Program Manual 12-18

Figure 5: Employment Seasonal Factors

November 2010

LAUS Program Manual 12-19

Figure 6: CPS Population

November 2010

LAUS Program Manual 12-20

Appendix
Notes for the STARS Tables
What are the STARS Tables?
o They are tools for understanding the LAUS model-based labor force estimates and relating these
estimates to the State’s and/or area’s economy.
o They provide information about the estimates.
o They provide information about the strengths and limitations of the estimates.
o They relate various pieces of information regarding the CPS and model-based estimates to one
another so that a more comprehensive understanding of the estimates is achieved.
What are the STARS Tables used for?
o They provide guidance for relating the estimates to the area’s overall economy.
§ They provide a snapshot of the current calendar year’s estimates.
§ They provide a snapshot of the current calendar year’s estimates as they relate to the
previous month and year.
§ They provide a snapshot of the volatility of the current year’s estimates using statistical
measures of reliability.
§ They provide error measures for the estimates – the point estimates themselves and the
over-the- month and change in the estimates.
o They provide guidance for relating the estimates to their component parts.
§ They provide a monthly decomposition of the over-the- month change in the CPS
estimates into its principle components – signal and noise.
§ They provide a monthly decomposition of the over-the- month change in the Model
estimates into their principle components – signal, trend, and seasonal.
§ They provide a monthly decomposition of the over-the- month change in the trend into
change in the smoothed estimates and a residual.
o They provide guidance regarding normal and abnormal seasonal movements in the model
estimates and the state-supplied model inputs.
§ Seasonal factors for model employment and unemployment estimates
§ Seasonal factors for CES and UI claims state-supplied model inputs.
o They provide guidance regarding possible extreme values for CPS and state-supplied model
inputs.
§ Prediction errors for CPS and state-supplied model inputs.
§ Standardized prediction errors for CPS and state-supplied model inputs.
o They provide guidance regarding the impact of real-time benchmarking.
§ Unusually large changes in the pro-rata benchmarking factors from one month to the
next.
o They provide graphs of the last three-plus years for CPS, Model, State-supplied model inputs and
CPS population so that the user can place the behavior of the current calendar year’s data in a
larger context. This provides a quick an easy way to relate the current month’s estimates with
those of the recent past.
§ The impact of unusual CPS and/or state-supplied model inputs on the model-based
estimates can be observed and evaluated.
§ Overall trends, as well as departures from these trends can be observed.
§ Concurrent seasonal factors can be compared with the historical norms for a given month.

November 2010

LAUS Program Manual 12-21

Appendix
Notes for the STARS Tables
Using the STARS Tables
o Table 1a. This table provides a snapshot of the current year’s smoothed seasonally adjusted
estimates, including indicators of significant over-the- month change.
§ This table provides the topside labor force characteristics (LF, LFP, EM, EP, UN, UR)
with the normal seasonal fluctuations removed.
§ This table provides the error range for the unemployment rate estimate at 90%
confidence. The user can say with 90% confidence that the true seasonally adjusted
unemployment rate for the general population should fall within this range.
§ This table allows the user to see if there was a statistically significant change in any of
the topside labor force estimates.
§ This table allows user to see if there is a trend in the movement of the year-to-date
smoothed seasonally adjusted estimates.
o Table 1b. This table provides a snapshot of the current year’s unadjusted estimates, including
indicators of significant over-the- month change.
§ This table provides the topside labor force characteristics (LF, LFP, EM, EP, UN, UR)
with their normal monthly seasonal fluctuations.
§ This table provides the error range for the unemployment rate estimate at 90%
confidence. The user can say with 90% confidence that the true unadjusted
unemployment rate for the general population should fall within this range.
§ This table allows the user to see if there was a statistically significant change in any of
the topside labor force estimates. (used with 1d, below)
§ Combined with table 1a, this table allows the user to see if the mo nth-to-month
movements in the unadjusted estimates are consistent with past seasonal behavior.
o Table 1c. This table displays the monthly over-the- month change in the topside smoothed
seasonally adjusted estimates for the current year in both level and percent terms.
§ This tables shows the user what the over-the month change was in direction, numeric, and
percent terms, and whether there is a pattern in the month-to- month changes – such as,
are they consistently positive or negative.
§ This table allows the user to evaluate whether the trend in the smoothed seasonally
adjusted estimates is consistent with past behavior and/or expected behavior.
§ When used in combination with table 1a, the user can tell whether a change was
statistically significant.
§ When used in combination with figures 1a, 2a, and 3a, it provides the user with context
for the current year’s estimates – with the caveat that the current year’s forward-filter
estimates will display more volatility than the historical estimates.

November 2010

LAUS Program Manual 12-22

Appendix
Notes for the STARS Tables
o Table 1d. This table displays the monthly over-the- month change in the topside unadjusted
estimates for the current year in both level and percent terms.
§ This tables shows the user what the over-the month change was in direction, numeric, and
percent terms, and whether there is a pattern in the month-to- month changes – such as,
are they consistently positive or negative.
§ This table allows the user to evaluate whether the month-to- month movement in the
unadjusted estimates is consistent with expected seasonal behavior.
§ When used in combination with table 1b, the user can tell whether a change was
statistically significant.
§ When used in combination with figures 1b, 2b, and 3b, it provides the user with context
for the current year’s estimates – with the caveat that the current year’s forward-filter
estimates will display more volatility than the historical estimates.
o Tables 2a. and 2b. These tables provide users with year-to-date standard errors for smoothed
seasonally adjusted and unadjusted model estimates. The standard errors displayed on these
tables are for one standard error (one sigma) The measures can be used to describe the reliability
of the model estimates and to calculate error ranges and confidence intervals for the model
estimates.
§ The tables provide a means fo r conveying to data users that model estimates are not
precise measures of labor force “truth,” but rather are imprecise “estimates” of the labor
force characteristics around which ranges can be constructed that include, with varying
degrees of confidence, the unobserved “true” labor force value.
§ The standard errors in these tables, together with common adjustment factors (1.0 = 68%,
1.282 = 80%, 1.645 = 90%, 1.960 = 95%, 2.326 =98%, and 2.576 = 99%), can be used to
calculate error measures – error ranges, confidence intervals, and coefficients of variation
– at various level of reliability for the seasonally adjusted and unadjusted model
estimates.
§ These error measures allow the users to convey the critical concept of imprecision
associated with all forms of estimates. Just like political poll numbers, CPS estimates
and model labor force estimates have uncertainty associated with them.
o Tables 3a. and 3b. These tables provide users with year-to-date over-the- year change – level and
percent – for smoothed seasonally adjusted and unadjusted estimates. These tables can be used
to evaluate possible cyclical behavior in the model estimates.
§ If the over-the-year changes in the smoothed seasonally adjusted and unadjusted
estimates show consistent increases in employment and decreases in unemployment and
unemployment rates, this is an indication that the labor market is in a positive phase of
the business cycle and is expanding (and possibly tightening).
§ If the over-the-year changes in the smoothed seasonally adjusted and unadjusted
estimates show consistent decreases in employment and increases in unemployment and
unemployment rates, this is an indication that the labor market is in a negative phase of
the business cycle and is contracting.
§ If the over-the-year changes in the smoothed seasonally adjusted and unadjusted
estimates show inconsistent employment, unemployment, and unemployment rate
changes, this is an indication that the labor market conditions are inconclusive – they
may be stagnant or at a turning-point in the business cycle.

November 2010

LAUS Program Manual 12-23

Appendix
Notes for the STARS Tables
o Tables 3c. and 3d. These tables provide users with year-to-date over-the- year change standard
errors for smoothed seasonally adjusted and unadjusted model estimates. The standard errors
displayed on these tables are for one standard error (one sigma). The measures can be used to
describe the reliability of the model estimates and to calculate error ranges and confidence
intervals for the model estimates.
§ The tables provide a means for conveying to data users that model estimates are not
precise measures of labor force “truth,” but rather are imprecise “estimates” of the labor
force characteristics around which ranges can be constructed that include, with varying
degrees of confidence, the unobserved “true” labor force value.
§ The standard errors in these tables, together with common adjustment factors (1.0 = 68%,
1.282 = 80%, 1.645 = 90%, 1.960 = 95%, 2.326 =98%, and 2.576 = 99%), can be used to
calculate error measures – error ranges, confidence intervals, and coefficients of variation
– at various level of reliability for the seasonally adjusted and unadjusted model
estimates.
§ These error measures allow the users to convey the critical concept of imprecision
associated with all forms of estimates. Just like political poll numbers, CPS estimates
and model labor force estimates have uncertainty associated with them.
o Table 4a. This table provides a year-to-date decomposition of the over-the- month change in
model unadjusted unemployment rate estimates. The model estimates are decomposed into their
signal change, trend change, and seasonal change components.
§ This table allows the user to determine which component is responsible for the overall
over-the- month change in the model unemployment rate.
§ The signal change is the overall change in the unadjusted estimate and is equal to the
trend change plus the seasonal change.
§ The smooth change is the overall change in the official smoothed seasonally adjusted
estimate.
§ The residual change is the change in the trend which is not accounted for in the smooth
change. Smooth change and residual change sum to the total trend change.
§ The trend change is the overall change in the seasonally adjusted estimate and is equal to
the signal change minus the seasonal change.
§ The seasonal change is equal to the over-the- month change in the seasonal factor and is
equal to the signal change minus the trend change.
o Table 4b. This table provides a year-to-date decomposition of the over-the- month change in CPS
unemployment rate estimates. The CPS estimates are decomposed into actual change, the signal
change, and noise change components.
§ This table allows the user to determine how much noise change was removed from the
overall over-the- month change in the unadjusted model unemployment rate.
§ The signal change is the overall change in the unadjusted estimate and is equal to the
actual change minus the noise change.
§ The signal change in table 4b will equal the signal change in table 4a.
§ This table allows the user to see how much change is being passed to the model from an
unusual change in the CPS unemployment rate.

November 2010

LAUS Program Manual 12-24

Appendix
Notes for the STARS Tables
o Table 5a. This table provides a year-to-date decomposition of the over-the- month change in
model unadjusted unemployment level estimates. The model estimates are decomposed into their
signal change, trend change, and seasonal change components.
§ This table allows the user to determine which component is responsible for the overall
over-the- month change in the model unemployment level.
§ The signal change is the overall change in the unadjusted estimate and is equal to the
trend change plus the seasonal change.
§ The smooth change is the overall change in the official smoothed seasonally adjusted
estimate.
§ The residual change is the change in the trend which is not accounted for in the smooth
change. Smooth change and residual change sum to the total trend change.
§ The trend change is the overall change in the seasonally adjusted estimate and is equal to
the signal change minus the seasonal change.
§ The seasonal change is equal to the over-the- month change in the seasonal factor and is
equal to the signal change minus the trend change.
o Table 5b. This table provides a year-to-date decomposition of the over-the- month change in CPS
unemployment level estimates. The CPS estimates are decomposed into actual change, the signal
change, and noise change components.
§ This table allows the user to determine how much noise change was removed from the
overall over-the- month change in the unadjusted model unemployment.
§ The signal change is the overall change in the unadjusted estimate and is equal to the
actual change minus the noise change.
§ The signal change in table 5b will equal the signal change in table 5a.
§ This table allows the user to see how much change is being passed to the model from an
unusual change in CPS unemployment.
o Table 6a. This table provides a year-to-date decomposition of the over-the- month change in
model unadjusted employment estimates. The model estimates are decomposed into their signal
change, trend change, and seasonal change components.
§ This table allows the user to determine which component is responsible for the overall
over-the- month change in the model employment.
§ The signal change is the overall change in the unadjusted estimate and is equal to the
trend change plus the seasonal change.
§ The smooth change is the overall change in the official smoothed seasonally adjusted
estimate.
§ The residual change is the change in the trend which is not accounted for in the smooth
change. Smooth change and residual change sum to the total trend change.
§ The trend change is the overall change in the seasonally adjusted estimate and is equal to
the signal change minus the seasonal change.
§ The seasonal change is equal to the over-the- month change in the seasonal factor and is
equal to the signal change minus the trend change.
o Table 6b. This table provides a year-to-date decomposition of the over-the- month change in CPS
employment estimates. The CPS estimates are decomposed into actual change, the signal change,
and noise change components.

November 2010

LAUS Program Manual 12-25

Appendix
Notes for the STARS Tables
§
§
§
§

This table allows the user to determine how much noise change was removed from the
overall over-the- month change in the unadjusted model employment.
The signal change is the overall change in the unadjusted estimate and is equal to the
actual change minus the noise change.
The signal change in table 6b will equal the signal change in table 6a.
This table allows the user to see how much change is being passed to the model from an
unusual change in CPS employment

o Table 7. This table provides users with year-to-date standard errors for the over-the- month (o-tm) change (i.e., components of change) in smoothed seasonally adjusted and unadjusted
estimates. The standard errors displayed on these tables are for one standard error (one sigma).
The measures can be used to describe the reliability of the o-t-m change in model estimates and
to calculate error ranges and confidence intervals for the over-the- month change in model
estimates.
§ The tables provide a means for conveying to data users that o-t-m changes in model
estimates are not precise measures of “true” labor force change,” but rather are imprecise
“estimates” of the labor force characteristics around which ranges can be constructed that
include, with varying degrees of confidence, the unobserved “true” change in the labor
force value.
§ The standard errors in these tables, together with common adjustment factors (1.0 = 68%,
1.282 = 80%, 1.645 = 90%, 1.960 = 95%, 2.326 =98%, and 2.576 = 99%), can be used to
calculate error measures – error ranges, confidence intervals, and coefficients of variation
– at various level of reliability for o-t- m changes in smoothed seasonally adjusted and
unadjusted model estimates.
§ These error measures allow the users to convey the critical concept of imprecision
associated with all forms of estimates.
o Table 8a. This table provides a list of the year-to-date CPS estimates. It provides a quick and
easy reference for examining the month-to-month behavior of the CPS estimates.
§ This table allows the user to see any unusual month-to- month movements in the current
year’s estimates.
§ This table allows the user to see if there is a trend (i.e., consistent positive or negative
month-to-month changes) in the CPS estimates and assess whethe r the CPS behavior is
consistent with expectations.
o Table 8b. This table provides users with standard errors for year-to-date CPS estimates – for
levels and over-the- month (o-t- m) change. The standard errors displayed on these tables are for
one standard error (one sigma). These measures can be used to describe the reliability of the
CPS estimates and to calculate error ranges and confidence intervals for the estimates and o-t-m
changes in the estimates.
§ The error measures produced from these standard errors provide a means for conveying
to data users that CPS estimates are not precise measures of labor force “truth,” but rather
are imprecise “estimates” of the labor force characteristics around which ranges can be
constructed that include, with varying degrees of confidence, the unobserved “true” labor
force value, or, o-t-m change in that value.
§ The standard errors in these tables, together with common adjustment factors (1.0 = 68%,
1.282 = 80%, 1.645 = 90%, 1.960 = 95%, 2.326 =98%, and 2.576 = 99%), can be used to

November 2010

LAUS Program Manual 12-26

Appendix
Notes for the STARS Tables
§

calculate error measures – error ranges, confidence intervals, and coefficients of variation
– at various level of reliability for the CPS estimates and o-t-m change in those estimates.
These error measures allow the users to convey the critical concept of imprecision
associated with all forms of survey estimates.

November 2010

LAUS Program Manual 12-27

Appendix
Notes for the STARS Tables
o Table 9. This table provides a list of year-to-date model-based concurrent seasonal factors for the
benchmarked seasonally adjusted model estimates, which are inputs to the official smoothed
seasonally adjusted estimates. These factors are produced as part of the classical (TCSI) series
decomposition that is part of the model estimation process. They are not produced by using an
ex-post- facto arithmetic process such as X-12 ARIMA.
§ The table can be used to calculate the model’s seasonal component of change, which is
equal to the o-t-m change in the seasonal factor.
§ The table can be used in conjunction with figures 4a and 5a to assess how much
difference there is in the current seasonal factor compared to the historical average for the
month, and by extension, how much the model’s seasonal component of change is being
impacted by adjustments to the current month’s seasonal factor.
o Table 10. This table provides year-to-date state-supplied CES and UI claims data – unadjusted
values (level), model-based smoothed seasonally adjusted estimates (trend), and seasonal factors.
It provides a quick and easy reference for examining the month-to- month behavior of the statesupplied model inputs.
§ This table allows the user to see any unusual month-to- month movements in the current
year’s model input variables.
§ The table allows users to assess whether month-to-month movements in the unadjusted
values conform to expected seasonal behavior.
§ Using the smoothed seasonally adjusted input variables (trend), the user can evaluate the
trend in their model input data – and evaluate whether the trend is consistent with their
expectations and other data sources.
o Table 11a. This table provides year-to-date prediction errors – levels and standardized – for CPS
unemployment and employment estimates. These error are used to evaluate how unusual (i.e.,
different from the expected value) the CPS estimates are. Large (more than 2.0) and very large
(more than 3.0) standardized prediction errors are a warning sign of a potential outlier, which
may, or may not, be smoothed out during re-estimation at the end of the year.
§ This table is used to evaluate how unusual (different from the predicted value) the current
CPS estimate is.
§ Typically, unusual CPS values directly impact the current model estimates.
§ Frequently however, if it is a one-time occurrence, the impact will be smoothed out of the
model estimates when the estimates are re-estimated.
o Table 11b. This table provides year-to-date prediction errors – levels and standardized – for State
CES and UI claims data. These error are used to evaluate how unusual (i.e., different from the
expected value) the state-supplied inputs are. Large (more than 2.0) and very large (more than
3.0) standardized prediction errors are a warning sign of a potential outlier, which may, or may
not, be smoothed out during re-estimation at the end of the year.
§ This table is used to evaluate how unusual (different from the predicted value) the current
state-supplied data are.
§ While the impact on the model estimates is typically less than the CPS, extreme values
can have a directly impact the current model variables.
§ Frequently, like unusual CPS estimates, their impact is smoothed out of the model
estimates when the estimates are re-estimated.

November 2010

LAUS Program Manual 12-28

Appendix
Notes for the STARS Tables
o Table 12. This table provides year-to-date pro-rata real-time benchmarking factors for model
unemployment and employment level estimates.
§ This table is used to evaluate the impact of real- time benchmarking on the model
estimates.
§ Calculating the over-the-month change in the pro-rata factor and multiplying that amount
times the model estimate (UE or EM) will produce a reasonable estimate of the impact of
real-time benchmarking.
§ Because the pro-rata factors are decimals, multiplying the change in the pro-rata factors
by 100 will convert the change into a percentage, which may convey the potential impact
more clearly.
o Figure 1a. This graph displays the last three-plus years of smoothed seasonally adjusted model
unemployment rate, benchmarked seasonally adjusted rate, the unbenchmarked seasonally
adjusted rate and seasonally adjusted UI claims rate.
§ This graph can be used to obtain a visual sense of the trend of the series, whether they are
moving in a similar fashion.
§ The graph is useful in spotting any unusual movements in the smoothed seasonally
adjusted model or claims rate estimates.
§ The graph is limited, as noted at the top of the graph, by the fact that past years’ model
estimates are re-estimated and therefore typically much less volatile than current forwardfiler estimates.
o Figure 1b. This graph displays the last three-plus years of unadjusted model unemployment rate,
CPS unemployment rate, and unadjusted UI claims.
§ This graph can be used to obtain a visual sense of the trend of the three series, whether
they are moving in a similar fashion.
§ The graph is useful in spotting any unusual movements in the CPS, model, or claims rate
estimates.
§ The graph can be used to evaluate whether the month-to-month movements in the various
data series follow expected seasonal patterns.
§ The graph is limited, as noted at the top of the graph, by the fact that past years’ model
estimates are re-estimated and therefore typically much less volatile than current forwardfiler estimates.
o Figure 2a. This graph displays the last three-plus years of smoothed seasonally adjusted model
unemployment, benchmarked seasonally adjusted unemployment, the unbenchmarked seasonally
adjusted unemployment and seasonally adjusted UI claims.
§ This graph can be used to obtain a visual sense of the trend of the series, whether they are
moving in a similar fashion.
§ The graph is useful in spotting any unusual movements in the seasonally adjusted model
or claims estimates.
§ The graph is limited, as noted at the top of the graph, by the fact that past years’ model
estimates are smoothed and therefore typically much less volatile than current forwardfiler estimates.

November 2010

LAUS Program Manual 12-29

Appendix
Notes for the STARS Tables
o Figure 2b. This graph displays the last three-plus years of unadjusted model unemployment, CPS
unemployment, and unadjusted UI claims.
§ This graph can be used to obtain a visual sense of the trend of the series, whether they are
moving in a similar fashion.
§ The graph is useful in spotting any unusual movements in the CPS, model, or claims data.
§ The graph can be used to evaluate whether the month-to-month movements in the various
data series follow expected seasonal patterns.
§ The graph is limited, as noted at the top of the graph, by the fact that past years’ model
estimates are re-estimated and therefore typically much less volatile than current forwardfiler estimates.
o Figure 3a. This graph displays the last three-plus years of smoothed seasonally adjusted model
employment, benchmarked seasonally adjusted employment, unbenchmarked seasonally
adjusted employment and seasonally adjusted CES employment.
§ This graph can be used to obtain a visual sense of the trend of the two series, whether
they are moving in a similar fashion.
§ The graph is useful in spotting any unusual movements in the smoothed seasonally
adjusted model or CES employment estimates.
§ The graph is limited, as noted at the top of the graph, by the fact that past years’ model
estimates are re-estimated and therefore typically much less volatile than current forwardfiler estimates.
o Figure 3b. This graph displays the last three-plus years of unadjusted model employment, CPS
employment, and CES employment estimates.
§ This graph can be used to obtain a visual sense of the trend of the series, whether they are
moving in a similar fashion.
§ The graph is useful in spotting any unusual movements in the CPS, model, or CES
employment data.
§ The graph can be used to evaluate whether the month-to-month movements in the various
data series follow expected seasonal patterns.
§ The graph is limited, as noted at the top of the graph, by the fact that past years’ model
estimates are re-estimated and therefore typically much less volatile than current forwardfiler estimates.
o Figure 4a. This graph displays the monthly concurrent seasonal factors for model unemployment
estimates for the past ten years – with each month’s seasonal mean.
§ This graph can be used to determine if the seasonal factors are fixed – there will be no
difference from the monthly mean.
§ This graph can be used to determine if the seasonal factors are changing over time – the
plot of the seasonal factors will produce a line that crosses the mean. The more vertical
the line, the greater the amount of variability from the seasonal mean.

November 2010

LAUS Program Manual 12-30

Appendix
Notes for the STARS Tables
o Figure 4b. This graph displays the monthly concurrent seasonal factors for UI claims for the past
ten years – with each month’s seasonal mean.
§ This graph can be used to determine if the seasonal factors are fixed – there will be no
difference from the monthly mean.
§ This graph can be used to determine if the seasonal factors are changing over time – the
plot of the seasonal factors will produce a line that crosses the mean. The more vertical
the line, the greater the amount of variability from the seasonal mean.
o Figure 5a. This graph displays the monthly concurrent seasonal factors for model employment
estimates for the past ten years – with each month’s seasonal mean.
§ This graph can be us ed to determine if the seasonal factors are fixed – there will be no
difference from the monthly mean.
§ This graph can be used to determine if the seasonal factors are changing over time – the
plot of the seasonal factors will produce a line that crosses the mean. The more vertical
the line, the greater the amount of variability from the seasonal mean.
o Figure 5b. This graph displays the monthly concurrent seasonal factors for CES employment for
the past ten years – with each month’s seasonal mean.
§ This graph can be used to determine if the seasonal factors are fixed – there will be no
difference from the monthly mean.
§ This graph can be used to determine if the seasonal factors are changing over time – the
plot of the seasonal factors will produce a line that crosses the mean. The more vertical
the line, the greater the amount of variability from the seasonal mean
o Figure 6: This graph displays the monthly population estimates from the CPS.
§

This graph can be used to determine if changes in the labor forces levels are attributed to
changes in the CPS 16 years and over noninstitutionalized population.

November 2010

LAUS Program Manual 12-31

Draft
Glossary
Additional Benefits (AB) See State Additional Benefits.
Additional Claim An additional claim is a notice of new unemployment filed at the beginning of
a second or subsequent series of claims within a benefit year or within a period of eligibility when
there has been intervening employment. This is one of three types of initial claims.
Additivity Adjustment The procedure which forces the exhaustive Handbook estimates to equal
the State estimate is known as additivity adjustment. The process is usually linear unless an
atypical procedure is in effect. The linear additivity adjustment is accomplished through the
Handbook share procedure of linking LMAs to the CPS-based State estimate.
Agent State The State in which a claimant files an interstate claim for compensation against
another (liable) State where wages were earned is the agent State. Usually, this is the claimant’s
State of residence.
All Other Nonagricultural Employment This includes self-employed, unpaid family workers, and
domestics in private households.
American Community Survey (ACS) A household survey developed by the Census Bureau to
replace the long form of the decennial census program. The ACS is a large demographic survey
collected throughout the year using mailed questionnaires, telephone interviews, and visits from
Census Bureau field representatives to about 3 million household addresses annually
Annual Processing (AP) A series of activities conducted annually which results in benchmarked
State and substate estimates. These activities include the State submission of revisions to model
inputs, revisions to substate inputs, incorporation of revised population controls, model
re-estimation and smoothing, benchmarking, and seasonally adjusting the revised series.
Areas of Substantial Unemployment (ASU) This is defined under JTPA as an area of at least
10,000 population with an average of 6.5 percent or more unemployment in the previous 12
months. It is used for determining eligibility for employment and training programs.
Autocorrelation Identifies whether the error terms in a regression equation are not independent
over time. If this is not accounted for in the equation for the regression line, poor coefficients and
predicted values may result. All State e models have coefficients adjusted to reflect
autocorrelation.
Autocorrelation Coefficient or ρ (rho) A mathematically determined value that measures the
relationship or correlation between successive error terms of the same series. A value of "0"
means that there is no correlation and a value of "1" indicates strong positive autocorrelation.
Base Period (Base Year) A base period is a specified period of twelve consecutive months (or in
some States, 52 weeks preceding the beginning of a benefit year) during which an ind ividual must
have the required employment and/or wages in order to establish entitlement to compensation or
allowances under an applicable program.
Benchmark This is a point of reference (either an estimate or a count) from which measurement
can be made or upon which adjustments are based.
Benefit Year A period, generally a 52-week period, during which individual claimants may
receive their maximum potential benefit s UI amount.

Glossary

Bias The difference between the expected value of the estimate from a probability sample and the
true value of the population parameter.
Birth and Adoption Unemployment Compensation An experiment program conducted
2000-2003 that was designed to provide unemployment compensation to employees on approved
leave following the birth or adoption of a child.
Bureau of Labor Statistic s (BLS) Established in 1884 and now part of the U.S. Department of
Labor, this Federal agency functions as the principal data-gathering agency of the Federal
Government in the field of labor economics. BLS collects, processes, analyzes, and disseminates
data relating to employment, unemployment, the labor force, productivity, prices, family
expenditures, wages, industrial relations, and occupational safety and health.
Bureau of the Census (BOC) The BOC is a bureau of the U.S. Department of Commerce. It
conducts censuses of population and housing every 10 years and of agriculture, business,
governments, manufacturers, mineral industries, and transportation at five year intervals. It also
conducts the Current Population Survey (CPS) for the BLS.
Census A count or enumeration (as opposed to a sample or an estimate) of a specified population
or some other characteristics in a given area (housing, industry, etc.)
Census Share A method used to disaggregate LMA employment and unemployment estimates to
subareas by assigning to the areas the same proportion of the monthly, independent LMA estimate
as was evidenced in the most recent census.
Census Tracts Census tracts are small, relatively permanent statistical subdivisions of a county
that provide comparable population and housing census tabulations. Tracts are designed to be
relatively similar in population characteristics, economic status, and living conditions. The
average tract has about 4,000 inhabitants. Census tract boundaries are recommended by local
census tract committees and approved by the Bureau of the Census.
Certification (Certifying) The process and form by which a claimant states and attests to facts which
will determine eligibility for UI benefits for a given week. These include, for example, a search for work
and availability for work.

Civilian Labor Force The sum of all employed and unemployed persons excluding persons under
16 years of age, inmates of institutions, and members of the Armed Forces.
Claim A claim is a notice of unemployment filed by an individual to request a determination of
unemployment insurance eligibility and the amount of benefit entitlement, or to claim benefits or
waiting-period credit.
Claimant A person who files either an initial claim or a continued claim under (1) any State or
Federal unemployment compensation program or (2) any other program administered by the State
agency.
Claims-Based Unemployment Disaggregation A method for disaggregating LMA
unemployment to subareas by us ing (1) claims by county of residence to distribute Handbook
experienced unemployment and (2) CPS-based data to allocate Handbook new and reentrant
unemployment. It is used in conjunction with the population-based indexed share employment
disaggregation.
Class of Worker There are three classes of workers: (1) wage and salary workers who receive
LAUS Program Manual G-2

Glossary

wages, salary, commission, tips, or pay in kind from an employer; (2) self-employed persons who
work for profit or fees in their own business, profession, or trade, or on their own farms; and (3)
unpaid family workers who work without pay for 15 or more ho urs a week on a farm or in a
business operated by a household member to whom they are related by birth, marriage, or
adoption.
Coefficients The values of the intercept and slope in the formula for the regression line.
Coefficients are estimated by a mathematical formulation which calculates coefficients by
minimizing the squares of the differences between the actual values (Y) and the predicted values
(Y'). They represent (mathematically) the relationship of the independent variable to the
dependent variable and how the changes in one variable can be related to another. In the case by
a of a variable coefficient model, the coefficients are allowed to change over time to reflect
changes that are occurring in the relationships of the dependent and the independent variables.
Coefficient of Variation (CV) The measure of relative dispersion of data. The standard
deviation divided by the arithmetic mean times 100 yields the coefficient of variation.
Combined Statistical Area A geographic entity consisting of two or more adjacent Core Based
Statistical Areas (CBSA) linked through commuting ties.
Commutation Regular travel of a person from the place of residence to the job location or to the
place of filing for UI benefits is referred to as commutation.
Commuter Claimant Under the Intrastate Benefit Payment plan, a worker who travels regularly
across a State line from home to work, and by mutual agreement between States, files in the State
where the individual last worked when employed, and is treated as a resident of that State.
Compositing An estimating technique which combines information from different sources,
taking into account the relative accuracy of each source. In the LAUS regression models, the
Kalman Filter technique can be thought of as a type of compositing. It combines CPS and model
estimates using their variances as a measure of the accuracy of the data.
Continued Claim A claim filed after the initial claim, by mail, telephone, or in person, for
waiting-period credit or payment for a certified week of unemployment.
Core A densely settled concentration of population, comprising either an urbanized area (of
50,000 or more population) or an urban cluster (of 10,000 to 49,999 population) defined by the
Census Bureau, around which a Core Based Statistical Area is defined.
Core Based Statistical Area (CBSA) A statistical geographic entity consisting of the county or
counties associated with at least one core (urbanized area or urban cluster) of at least 10,000
population, plus adjacent counties having a high degree of social and economic integration with
the core as measured through commuting ties with the counties containing the core. Metropolitan
and Micropolitan Statistical Areas are the two categories of Core Based Statistical Areas.
Correlation A statistical term which ind icates a structural, functional, qualitative correspondence
between comparable entities. Correlation can be either positive (simultaneous increase or
decrease in both variables) or in negative (increase in the value of one and decrease in the value of
the other variable).
Correlation Coefficient A mathematically determined value that measures the relationship or
correlation between two time series. As with the autocorrelation coefficient, a value of "0"
LAUS Program Manual G-3

Glossary

indicates no correlation and a value of "1" indicates a strong positive relationship. A value of "-1"
indicates strong negative relationship, or meaning that as one series increases, the other series
decreases.
Covered Employment Those jobs covered by the unemployment compensation programs are
considered covered employment. At this time, those not covered include some agricultural
workers, employees of religious and small nonprofit organizations, household workers, and
self-employed workers.
Current Employment Statistics (CES) program A BLS monthly survey of about 140,000
businesses and government agencies, representing approximately 410,000 individual worksites
that yields estimates of nonagricultural wage and salary employment, hours, and earnings by
industry. These statistics are prepared monthly for the nation as a whole, and by cooperating
State agencies for all 50 States, the District of Columbia, Puerto Rico, the Virgin Islands, and
about 400 metropolitan areas and divisions.
Current Population Survey (CPS) A monthly survey conducted by the Bureau of the Census of
approximately 60,000 assigned households of which 50,000 are eligible for interview. This
survey of the civilian noninstitutional population of the United States provides monthly statistics
on employment, unemployment, demographic characteristics, and related subjects which are
analyzed by the Bureau of Labor Statistics.
Denial of Benefits An action imposed by a State agency after a nonmonetary determination or an
appeals decision which cancels, reduces, or postpones a claimant's benefit rights.
Dependent Variable The variable for which estimates are desired, usually termed the "Y"
variable. In the LAUS models, the dependent variable used in constructing the model is the
monthly CPS estimate.
Determination An official decision by the State UI agency regarding the unemployment claim of a
person. (See monetary and nonmonetary determination.)

Disaggregation A method to divide a statistic into its component parts. For example, the LMA
unemployment is divided into each component county or city.
Disaster Unemployment Assistance (DUA) A program that provides unemployment assistance
to individuals whose unemployment is a direct result of a major disaster as declared by the
President of the United States.
Discouraged Workers Individuals not in the labor force who want and are available for a job and
who have looked for work sometime in the past 12 months (or since the end of their last job if they
held one within the past 12 months), but are not currently looking, because they believe there are
no jobs available or there are none for which they would qualify.
Dynamic Residency Adjustment Ratios (DRR) A method to adjust CES employment data for
resident employment in an area that accounts for the relationship between employed residents and
jobs in that area and in other areas within commuting distance and job growth within these areas.
Earnings Disregarded The amount prescribed by State unemployment compensation laws that a
claimant may earn without any reduction in weekly benefit amount fo r a week of total
unemployment are earnings disregarded. This is also referred to as the forgiveness level for
earnings. The amounts vary by State.
LAUS Program Manual G-4

Glossary

Earnings Due to Employment These are earnings, either from the regular employer or from odd
jobs, which a UI claimant may receive while certifying to a week of unemployment. The
existence of these earnings classified the claimant as employed, even when earnings are less than
the State's forgiveness level.
Emergency Unemployment Compensation (EUC08) A 100 percent federally- funded temporary
program that provided up to 34 weeks of benefits to eligible jobless workers in every state, and up
to 19 additional weeks in states with “high unemployment” (for a maximum of 53). The EUC08
program was effective from July 2008 through June 2, 2010.
Employed In the CPS, those individuals 16 years of age or older who worked at least one hour for
pay or profit or worked at least 15 unpaid hours in a family business during the reference week are
considered employed. Ind ividuals are also counted as employed if they had a job but did not work
because they were ill, on vacation, in a labor dispute, prevented from working because of bad
weather, or taking time off for personal reasons.
Employment/Population Ratio The proportion of the civilian noninstitutional population who are
classified as employed.
Employment and Training Administration (ETA) Agency under the Department of Labor that
administers federal government job training and worker dislocation programs, federal grants to
states for public employment service programs, and unemployment insurance benefits. These
services are primarily provided through state and local workforce development systems.
Enumeration Districts (EDs) Administrative units used in the Census d, are referred to as
enumeration districts. They contain, on the average, about 750 people. The EDs provide a list of
addresses for housing units which is used to help set up the sample file for the CPS.
Error See Standard Error.
Establishment An economic unit which produces goods or services, is generally found at a single
physical location, and is primarily engaged in one type of economic activity.
Exhaustees Individuals who have exhausted all of the unemployment insurance benefits to which
they are entitled within a benefit year and cannot establish a new benefit year.
Extended Benefits (EB) The supplemental program, established by Public Law 91-373, that pays
extended compensation during a period of specified high unemployment to individuals for weeks
of unemployment after they have exhausted regular compensation. The program is financed
equally from Federal and State funds and becomes operative at the State level. The State
determines benefits and certain restrictions.
Extrapolate A method to project values of a variable in an unobserved interval from values within
an already observed interval.
Federal Information Processing Standards (FIPS) Standards for information processing that
comprise a geographically exhaustive five digit code system wherein areas such as State, counties,
territories, and metropolitan areas are uniquely identified. FIPS codes were officially replaced
with the Geographic Names Information System (GNIS) as the Federal and national standard for
geographic nomenclature. However, FIPS codes are still maintained by the Census Bureau and
are referred to as either Census codes of Federal codes.
LAUS Program Manual G-5

Glossary

Final Payment The last payment to a claimant which exhausts the individual's maximum
potential benefit entitlement under a specific program is referred to as a final payment.
Forgiveness Level See Earnings Disregarded.
Gain A weighting factor used in the Kalman Filter in determining the current month coefficients.
Using this factor, a portio n of the difference between the current month's CPS and the preliminary
model estimate is added or subtracted from the previous month's coefficient value. This is used to
produce the current month's coefficient.
Geographic Names Information System (GNIS) The Geographic Names Information System
(GNIS) is the Federal and national standard for geographic nomenclature. The GNIS contains
information about physical and cultural geographic features of all types in the United States,
associated areas, and Antarctica, current and historical, but not including roads and highways.
The GNIS Feature ID has superseded the Federal Information Processing Standard (FIPS) 55
Place Code as the Federal feature identifier.
Handbook Method A building-block estimation method that uses data from several
sources—including the Current Population Survey, the Current Employment Statistics program,
and unemployment insurance program—to produce labor force estimates at the substate level.
Estimates for Labor Market Areas (LMAs), including both metropolitan and micropolitan areas
and small LMAs, are produced using this methodology.
Henderson Trend Filter (H13) A filtering procedure, based on moving averages, to remove the
irregular fluctuations from the seasonally-adjusted series, leaving the trend. It is part of a set of
trend filters developed by Robert Henderson (1916) for use in actuarial work that are used
extensively by seasonal adjustment packages such as X-12-ARIMA.
Household As defined by the Bureau of the Census, a household is all persons who occupy a
housing unit. A housing unit is a room or group of rooms intended for occupancy as separate
living quarters and consists of either a separate entrance or complete cooking facilities for the
exclusive use of the occupants.
ICON (Interstate Connection) A centralized computerized system of reporting and exchanging
unemployment insurance claims information between States.
Independent Variables Variables used in the regression equation to predict the dependent
variable, "Y". The independent variables are usually termed d the "X" variables.
Information Technology Support Center (ITSC) Established Department of Labor to assist all
state Unemployment Insurance agencies in the area of Unemployment Insurance Information
Technology.
Initial Claim Any notice of unemployment filed by an individual to initiate (1) a determination of
entitlement to an d eligibility for compensation (a new claim), (2) a subsequent period of
unemployment within a benefit year or period of eligibility (an additional claim), or (3) a new
claim filed to request a determination of eligibility and establishment of a new benefit year within
an existing spell of unemployment (transitional claim).
Institutional Population Persons residing in CPS-defined institutions, such as prisons, nursing
homes, juvenile detention facilities, or residential mental hospitals. Persons residing outside of
these institutions constitute the non- institutional population.
LAUS Program Manual G-6

Glossary

Insured Unemployment Unemployment during a week for which waiting period credit or
benefits are claimed under the regular unemployment insurance compensation programs,
supplemental extended benefit programs, or the railroad unemployment insurance program, is
considered insured.
Insured Unemployment Rate (IUR) The rate computed by dividing Insured Unemployed for the
current quarter by Covered Employment for the first four of the last six completed quarters.
Intercept The value of "Y" (dependent variable) where the regression line crosses the "Y" axis is
the intercept. The intercept is usually denoted by β 0 .
Interpolate A method to estimate values of a function between two known values.
Interstate claim A claim filed in one (agent) State based on monetary entitlement to
compensation in another (liable) State. The agent State is usually the claimant’s State of
residence. The liable State is the location of the establishment in which wage credits were earned.
Intrastate Claim A claim filed in the same State in which the individual's wage credits were
earned. A nonresident of the State s, filing an intrastate claim is called a commuter claimant.
Job Leavers Individuals who quit or otherwise terminate their employment voluntarily and
immediately begin looking for work.
Job Losers Individuals on layoff and those whose employment ended involuntarily and who
immediately begin looking for work.
Kalman Filter A statistical technique used in the Signal- in plus-Noise models to adjust the model
coefficient. The coefficients are updated each month with new information using the Kalman
Filter technique. This technique combines information from the model and CPS when making
the new model estimate by taking into account the relative accuracy of each.
Labor Force The total of all civilians classified as employed and unemployed. The labor force,
in addition, includes members of the armed forces stationed in the United States.
Labor Market Area (LMA) An economically integrated geographical unit within which workers
may readily change jobs without changing their place of residence. All States are divided into
exhaustive LMAs, which include a county or a group of contiguous counties, except in New
England where cities and towns are used. Independent Handbook estimates of employment and
unemployment are made monthly for each LMA and form the basis for the LAUS estimates.
Labor Surplus Area Defined under the Defense Manpower Policy No. 4A as an area with at least
120 percent of the national unemployment rate. (There is a variable floor and ceiling rate of 6%
and 10 %.)
LAUS Estimate The official BLS-published employment and unemployment estimates. For
States, they are based on the signal-plus-noise models. For areas, they are developed using the
Handbook procedures and are controlled to the State levels.
LAUS Redesign A multi-year, multi-project initiative implemented with January 2005 estimates
that improved labor force estimates for State and substate areas. The redesign included improved
time-series models, models for six additional metropolitan areas, real- time benchmarking,
enhanced procedures for developing substate data, the implementation of 2000-Census based
configurations for metropolitan areas, metropolitan divisions, micropolitan areas, and small labor
market areas as well as the incorporation of 2000-Census inputs and updates in the methodology.
LAUS Program Manual G-7

Glossary

Least Squares A basic regressio n technique used to "fit" (calculate) a model equation to a time
series of data. There are several different types of least square calculations but all are based on
minimizing the sum of the squared differences between the data points and a regression line.
Liable/Agent Data Transfer (LADT) The record format used for the exchange of statistical data
via the Interstate Connection (ICON). The LADT record format was developed to accommodate
the exchange of data pertaining to: a) interstate weeks claimed; b) intrastate commuter weeks
claimed; c)interstate initial claims (new, additional and transitional); and, d) interstate reopened
claims and claim transfers by ICON for the exchange of interstate UI claims data among States.
Liable State Any State against which a worker files a claim for compensation through the
facilities of another (agent) State is the liable State. The State location of the establishment in
which wage credits are earned is the liable State.
Link Relative Technique A method for employment estimation that involves, for each estimating
cell, comparing the ratio of all employees in one month to all employees in the preceding month.
The all employee estimate for each month is obtained by multiplying this ratio by the all employee
estimate for the previous month. The technique is used in the CES estimating methodology.
Local Area Unemployment Statistics (LAUS) The Federal/State cooperative program under
which employment and unemployment estimates for States and local areas are developed. These
estimates are prepared by State Employment Security Agencies in accordance with BLS
definitions and procedures. They are used for planning and budgetary purposes, as an indication
of need for employment and training programs, and to allocate Federal funds under JTPA, FEMA,
etc.
Mass Layoff Statistics (MLS) A Federal-State cooperative program which uses a standardized,
automated approach to identify, describe, and track the effects of major job cutbacks, using data
from each State's unemployment insurance database.
Mass Layoff Event A layoff in which 50 initial claims or more have been filed against an
establishment during a five-week period, with the separations expected to last longer than 30 days.
Mean Square Error (MSE) A measure of the total error that can arise in an estimate. It is equal
to the variance plus the bias squared. Mean square error is a more comprehensive measure of
estimation error than is variance and is an important statistical and analytical tool.
Metropolitan Division (MD) A county or group of counties within a CBSA that contains a core
with a population of at least 2.5 million. A Metropolitan Division consists of one or more
main/secondary counties that represent an employment center or centers, plus adjacent counties
associated with the main county or counties through commuting ties.
Metropolitan Statistical Area (MA) A CBSA associated with at least one urbanized area that has
a population of at least 50,000. The Metropolitan Statistical Area comprises the central county or
counties containing the core, plus adjacent outlying counties having a high degree of social and
economic integration with the central county as measured through commuting.
Micropolitan Area (MC) A CBSA associated with at least one urban cluster that has a population
of at least 10,000, but less than 50,000. The Micropolitan Statistical Area comprises the central
county or counties containing the core, plus adjacent outlying counties having a high degree of
social and economic integration with the central county as measured through commuting.
LAUS Program Manual G-8

Glossary

Migration The permane nt movement of an individual's residence from one location to another.
Model A mathematical equation which relates different variables and data. In time series, this
relationship is computed over time. In the LAUS signal-plus-noise models, the monthly State
CPS labor force as estimates are related to different independent variables and data that show
strong correlations to the monthly estimates.
Monetary Determination A written notice is issued to inform an individual whether or not the
individual meets the employment and wage requirements necessary to establish entit lement to
compensation under a specific unemployment insurance program. If an individual is entitled, the
weekly and maximum benefit amounts the individual may receive are also determined.
Months for Cyclical Dominance (MCD) An estimate of the time span required to identify
significant cyclical movements in a monthly economic time series. The MCD indicates the
shortest span of months over which changes in the series are dominated by cyclical rather than
irregular or erratic movements.
Moving Average A continuous process that uses a series of calculations made by initially taking
the simple average, or arithmetic mean, of a consecutive number of items, and then dropping the
first item and adding the next item in sequence and averaging, so that the number of items in the
series remains constant.
New Claim The first initial claim filed in person, by mail, telephone, or the Internet to request a
determination of entitlement to and eligibility for compensation. This is one of three types of
initial claims.
New England City and Town Area (NECTA) A statistical geographic entity that is defined using
cities and towns as building blocks and that is conceptually similar to the Core Based Statistical
Areas in New England (which are defined using counties as building blocks).
New England City and Town Area Division A city or town or group of cities and towns within a
NECTA that contains a core with a population of at least 2.5 million. A NECTA Division consists
of a main city or town that represents an employment center, plus adjacent cities and towns
associated with the main city or town, or with other cities and towns that are in turn associated with
the main city or town, through commuting ties.
New Entrants Individuals who enter the labor market for the first time and do not find jobs.
They include students entering the labor market after graduation from school and others who have
not previously held a full- time job lasting two weeks or longer.
Nonagricultural Wage and Salary Employment In the CES program, this is a count of jobs by
place of work on nonagricultural establishment payrolls (including employees on paid sick leave,
paid holiday, or paid vacation) for any part of the pay period including the 12th of the month. It
does not include proprietors, self-employed, unpaid volunteer or family workers, domestic
workers in households, military personnel, and persons who are laid off, on leave without pay, or
on strike for the entire reference period.
Noninstitutional Population See Institutional Population.
Nonmonetary Determination A process that determines whether a claimant meets legal criteria
other than wage credits under State UI law. It is usually concerned with: (1) reason claimant left
job (separation issues); and (2) job search (able, available, and actively seeking work).
LAUS Program Manual G-9

Glossary

Not in the Labor Force All persons 16 years of age or older who are neither employed nor
unemployed are considered not in the labor force. Some examples are students, housewives,
retirees, etc.
Place-of-Residence Adjustment of Employment Establishment-based data, which are on a
place-of-work basis, are adjusted to reflect the place of residence of the employed. The current
adjustment also corrects for multiple jobholding in the place-of-work series. See Dynamic
Residency Ratio.
Population-Based Indexed Share Employment Disaggregation A method that uses the annually
prepared total population estimates and data from the Census to disaggregate labor market area
total employment to the county or city level. This method is used only in conjunction with the
claims-based unemployment disaggregation.
Population Controls Refers to population data developed from various independent sources, such
as vital statistics on births, deaths, migration, school enrollment, persons living in group quarters,
inmates in institutions, etc., which are used in Current Population Survey estimation procedures to
independently adjust sample-based labor force levels. Population controls are updated annually
by the Bureau of the Census and provided to the Bureau of Labor Statistics.
Population Estimates Annual population estimates prepared by the Census Bureau that entails
updating population information from the most recent census with information found in the annual
administrative records such as tax records, Medicare records and some vital statistics information.
Predicted Value The value of Y' (Y prime) that one obtains by “plugging in" values of the
independent variables into the formula for the regression line is the predicted value. The
coefficients have already been determined by a mathematical formulation.
Prediction Period A period of time which is outside the sample period. Coefficients for the
regression line derived from the sample period are used to make predictions in subsequent periods.
It is also called the "outside sample" period.
Primary Metropolitan Statistical Area (PMSA) A geographic designation used prior to the LAUS
2005 Redesign for a subarea defined within an area that meets the requirements to qualify as an
MSA and also has a population of one million or more.
Primary Sampling Unit (PSU) The first stage of CPS sampling involves dividing the United
States into primary sampling units, most of which comprise a metropolitan area, a large county, or
a group of smaller counties with homogeneous demographic and economic characteristics.
Program for Measuring Insured Unemployment Statistics (PROMIS) A stand-alone PC-based
system that stores all claimant information, including socioeconomic characteristics, and generates
the UI inputs to the LAUS and Mass Layoff Statistics (MLS) programs and, potentially, other
programs. PROMIS operates as the clearinghouse for multi-purpose input data, allowing
flexibility to provide a more complete picture of the unemployment situation at substate levels.
Quarterly Census of Employment and Wages Program (QCEW) A federal/State cooperative
program that produces a comprehensive tabulation of employment and wage information for
workers covered by State unemployment insurance (UI) laws and Federal workers covered by the
Unemployment Compensation for Federal Employees (UCFE) program.
Railroad Retireme nt Board (RRB) The RRB is an independent agency in the executive branch of
LAUS Program Manual G-10

Glossary

the U.S. government which administers a comprehensive social insurance system for the nation's
railroad workers and their families, providing protection against the loss of income resulting from
old age, disability, death, unemployment, and temporary sickness.
Raking This is the process which forces additivity among components to the aggregate estimate.
It is performed on an iterative basis in the CPS.
Real-Time Benchmarking A tiered approach to estimation in which the census division estimates
are benchmarked to the national levels of employment and unemployment on a monthly basis.
The benchmarked division model estimates are then used as the benchmark for the States within
each division. The distribution of the monthly benchmark adjustment to the States is based on
each State's monthly model estimate. In this manner, the monthly State employment and
unemployment estimates will add to the national level. Substate estimates are then revised and
forced to add to the new S tate estimates. In the past, this was done annually because the state data
were benchmarked to the CPS annual average for each state. Under this approach, benchmarking
occurs monthly, while annual processing will continue to be done at the beginning of each
calendar year on the previous year's estimates.
Reentrants In the CPS, persons who previously worked at a full-time job at least two weeks but
who were out of the labor force for two weeks or more prior to beginning to look for work.
Reference Week The week for which data are collected. For the CPS, the reference week is the
calendar week including the 12th of the month. For UI data, it is the certification period. In most
States, the reference week fo r UI certifications is the calendar week including the 12th.
Exceptions are States with flexible benefit weeks and New York, whose week is a
Monday-through-Sunday week.
Regression A statistical tool which utilizes the relation between two or more variables so that one
variable can be predicted from the other(s).
Regression Equations The basic formula for a regression equation is shown below. In this
example, the equation has an intercept (β 0 ), independent variables (X1 and X2) with coefficients (β 1
and β 2 r espectively). The equation's error term is ε.
Y= β 0 + β 1 (X1 ) + β 2 (X2 ) + ε
Regression Line A line fitted to the points in the scatter plot to summarize the relationship
between the variables being studied. When it slopes down (from top left to bottom right), this
indicates a negative or inverse relationship between the variables; when it slopes up (from bottom
right to top left), a positive or direct relationship is indicated.
Reopened claim A claim filed after a break in claimed weeks during a benefit year. This break could be
caused by illness, disqualification, unavailability, or failure to report for any reason other than job
attachment. It is not a break resulting from other employment.

Residency Adjustment Factor A method formerly used to convert the CES nonfarm job count to a
person count by place of residence. It was replaced in 2005 with the Dynamic Residency
Adjustment Ratio (DRR) as part of the LAUS Redesign.
Residency Adjustment System (RAS) A National Office software system that assists States in
correctly coding the residency of their UI claims records. The system verifies and corrects
erroneous addresses and assigns geocodes including State, county, city/town, longitude and
LAUS Program Manual G-11

Glossary

latitude and census tract and block. RAS facilitates the city claims disaggregation process and is
required for the use of PROMIS.
Rotation Group One of eight systematic subsamples which comprise the total CPS sample. A
rotation group is in the sample for four consecutive months 1 year, leaves the sample during the
following eight months, and then returns for the same four calendar months of the next year.
Sample A subset of a statistical population selected for the purpose of making generalized
statements about the whole.
Sample Period A period of time which is used to derive coefficients for the regression line. It is
also called the "inside sample" period.
Sampling Error The measure of sampling variability, that is, the natural variations that might
occur by chance because only a sample of the population is surveyed.
Sample Regression A type of regression in which the dependent variable is calculated from a
sample survey. Consequently there is an additional error (sampling error) to be considered.
Sampling Ratio The proportion of units needed to be sampled to provide data of a specified level
of statistical reliability is the sampling ratio. Sampling ratios vary by cell, depending on the
degree of variability of the measured item.
Scatter Plot A graph which plots the values of the dependent variable (Y) against the values of
one of the independent variable (X). By convention, the "X" variable is plotted agains t the
horizontal scale and the "Y" variable is plotted against the vertical scale.
Self-Employment Assistance (SEA) An optional program to help unemployed workers to create their
own jobs by starting small businesses. To be eligible for the program an individual must be eligible
for unemployment compensation, have been permanently laid off from his/her previous job and
identified through the profiling system as likely to exhaust his/her benefits, and must participate in
self-employment activities including entrepreneurial training and business counseling.

Seasonal Adjustment A statistical technique that eliminates the influences of weather, holidays,
the opening and closing of schools, and other recurring seasonal events from economic time series.
This permits easier observation and analysis of cyclical, trend, and other nonseasonal movements
in the data.
Separation Issue, Nonmonetary Determination Situations of nonmonetary determination in which the
claimant acted in the termination of the employment relationship. For example, voluntary quit without good
cause, or voluntary quit for personal reasons.

Series Break An interruption in a time series caused either by a change in definition or in
methodology which makes it improper to compare data from after the change with data from
before the change.
Short-time compensation A program, commonly known as work-sharing, that allows an
employer, faced with the need for layoffs because of reduced workload, to reduce the number of
regularly scheduled hours of work for all employees rather than incur layoffs. This program
provides partial UI benefits to individuals whose work hours are reduced from full- time to
part-time on the same job.
Signal-Plus-Noise Models Econometric models used by the LAUS program to produce State
LAUS Program Manual G-12

Glossary

labor force statistics. The models measure the true labor force value contained in the monthly
CPS estimates (the signal) by extracting the noise associated with CPS sampling error.
Slope A value that tells how much change in the dependent variable (Y) results from a change in
one of the independent variables (X). It is defined as the change in "Y" divided by the change in
"X".
Smoothed Seasonal Adjustment (SSA) Seasonally-adjusted estimates that incorporate a long-run
trend smoothing procedure. A smoothed-seasonally adjusted series was introduced in 2010 to
reduce the number of spurious turning points in the former estimates. The estimates are smoothed
using the Henderson Trend Filter (H13) that suppresses irregular variation in real time. This new
approach addresses longstanding issues related to end-of- year revision and enhances the analytical
utility of the estimates.
Smoothing In the time series regression, one month's data are used in estimating another and the
best estimate is made when data from all the other months are incorporated. The process of
forward-back-forward model re-estimation is referred to as smoothing because of its impact on
monthly estimates. In LAUS, smoothing is part of the annual benchmarking processing to update
the model estimates series.
Standard Deviation A measure of dispersion around the mean value of a population frequently
denoted by sigma (σ). It is the positive square root of the variance.
Standard Error The term "Standard Error" can be used in many contexts. In general, it refers to
the variability of an estimate. In sampling, it usually refers to the confidence interval of the
sample estimate -the probability of including the true value with repeated sample. One standard
error is about 68 percent confidence; and 1.645 times the standard error is the more commonly
used 90 percent confidence. The model estimates also have confidence intervals. These relate to
the variability of the estimate relative to the regression line.
State Additional Benefits (AB) Solely State- financed programs for extending the potential
duration of benefits during periods of high unemployment for claimants in approved training who
exhaust benefits, or for a variety of other reasons. Although some state laws call these programs
“extended benefits,” this publication uses the term “additional benefits” to avoid confusion with
the federal-state EB program.
State Employment Security Agency (SESA) A generic name for the State agency usually
responsible for the following three activities: (1) The Unemployment Insurance Program which
includes UI tax collection, administration, and determination and payment of unemployment
benefits. (2) The Employment or Job Service Program which is an exchange for workers so
seeking work and employers seeking workers. (3) Research and Analysis which includes
collection, analysis, and publication of labor market information.
Statistical Population A group of entities or individuals that are of concern to a statistician for a
particular investigation. This is sometimes referred to as simply a "population".
Stochastic A term used to denote the randomness of a variable or process. A stochastic, or
random, variable is one whose value changes. In the case of the LAUS regression models, the
values of the model variables change from month to month.
Survey The process used to collect data for the analysis of some aspect of a group or area.
Time Series A consecutive set of observations over a specified period of time.
LAUS Program Manual G-13

Glossary

Time Series Independence A condition present when successive values of a time series are
nonrelated or noncorrelated.
Trade Readjustment Allowances (TRA) Benefits provided to individuals who were laid off or
had hours reduced because their job was adversely affected by increased imports from other
countries.
Transitional Claim A new claim filed to request a determination of eligibility and establishment
of a new benefit year within an existing spell of unemployment. This is one of three types of
initial claims.
Unemployment Compensation for Ex-Servicemen (UCX) This federal program provides
unemployment benefits to ex-servicemen.
Unemployed In the CPS, those individuals considered unemployed must be 16 years of age or
older who do not have a job but are available for work and are actively seeking work during the
reference week (the week including the 12th of the month). The only exceptions to these criteria
are individuals who are waiting to be recalled from a layoff and individuals waiting to report to a
new job within 30 days. They are also considered unemployed.
Unemployed Disqualified Persons who are able to work and are available for work but are
disqualified from receiving benefits for separation issues or other nonmonetary reasons.
Unemployment Compensation for Federal Employees (UCFE) This federal program provides
benefits to federal employees.
Unemployment Insurance (UI) Insurance premiums collected by the State and Federal
governments from which unemployment compensation is paid.
Unemployment Rate The number of persons unemployed, expressed as a percentage of the
civilian labor force.
Variable An entity that can take on a number of different values. It is frequently denoted by
letters such as "X" or "Y". Examples of variables would be CPS unemployment rate and CES
employment.
Variable Coefficient Model (VCM) A type of sample regression model in which the model's
coefficients are allowed to change over time.
Variance A mathematical measure of the dispersion of the values of a variable around its mean.
The variance may arise from a sampling of the population under study, or may just measure the
variability of population values around its means. The variance is frequently denoted as sigma
squared (σ2 ).
Waiting Week A period of unemployment during which a claimant may not draw benefits and
during which certain requirements essential to the establishment of claimant eligibility for benefits
must be met.
Weeks Claimed The number of weeks of benefits claimed, including weeks for which a waiting
period or fixed disqualification period is being served. Interstate claims are counted by State of
residence.
Worksharing See short-time compensation.
LAUS Program Manual G-14