Attachment I CPS-FSS Nonresponse Bias Report

Attachment I
Census Report “Evaluating Nonresponse Bias in the 2015 Food
Security Supplement to the Current Population Survey”

MEMORANDUM FOR

David Smallwood
Chief, Food Assistance Branch
US Department of Agriculture Economic Research Service

Through:

Lisa Clement
Survey Director, Current Population & American Time Use Surveys

From:

James Farber
Acting Chief, Demographic Statistical Methods Division

Subject:

Evaluating Nonresponse Bias in the 2015 Food Security Supplement
to the Current Population Survey

The purpose of this memorandum is to report on analysis of various nonresponse estimates
computed for the 2015 Food Security Supplement to the Current Population Survey and to
provide nonresponse bias analysis tables for the 2011 through 2014 Food Security Supplements.
If you have any questions about this document, please contact KeTrena Farnham at
301.763.5931 or [email protected] or contact David Hornick at 301.763.4183 or
[email protected].
Attachments (4):
Attachment A:
Attachment B:
Attachment C:
Attachment D:
Attachment E:

Nonresponse Bias Analysis Tables for the 2014 Food Security Supplement
Nonresponse Bias Analysis Tables for the 2013 Food Security Supplement
Nonresponse Bias Analysis Tables for the 2012 Food Security Supplement
Nonresponse Bias Analysis Tables for the 2011 Food Security Supplement
Summary of Key Differences Between the Tables for 2011-2015

cc:
Alisha Coleman-Jensen
Kyra Linse
Tim Marshall
Greg Weyland
David Watt
Lorelei De Vos

(USDA ERS)
(ADDP)

David Hornick
Yang Cheng
KeTrena Farnham
Mei Li
Tim Trudell
Rebecca Hoop
Nghiep Huynh
Anne Johnson

(DSD)

1

(DSMD)

Current Population Survey – Food Security Supplement
Demographic Statistical Methods Division
U.S. Census Bureau

Evaluating Nonresponse Bias in the 2015
Food Security Supplement to the Current
Population Survey

KeTrena Farnham
December 15, 2017
1

Table of Contents
1.

Executive Summary ................................................................................................................ 3

2.

Introduction ............................................................................................................................. 4

3.

Overview of the Current Population Survey ........................................................................... 4

4.

Overview of the 2015 Food Security Supplement to the Current Population Survey ............ 4

5.

Discussion of Nonresponse in the 2015 Food Security Supplement to the Current
Population Survey ................................................................................................................... 5

6.

Methods ................................................................................................................................... 6

7.

Limitations............................................................................................................................... 7

8.

Response Rates ........................................................................................................................ 7

9.

Respondent Distributions ...................................................................................................... 14

10. Discussions and Conclusions ................................................................................................ 20
11. References ............................................................................................................................. 21

List of Tables

Table 1:
Table 2:
Table 3:
Table 4:

2015 Food Security Unit Response Rates........................................................................ 8
Response Rates for December 2015 CPS Households .................................................... 9
Response Rates for 2015 FSS Households .................................................................... 11
Response Rates for 2015 FSS Households for Characteristics Only Available for
Responding CPS Households ........................................................................................ 13
Table 5: Respondent and Nonrespondent Distributions for December 2015 CPS Households .. 15
Table 6: Respondent and Nonrespondent Distributions for 2015 FSS Households .................... 17
Table 7: Respondent and Nonrespondent Distributions for 2015 FSS Households for
Characteristics Only Available for Responding CPS Households ................................ 19

2

1. Executive Summary
This report presents results of an analysis of various nonresponse estimates computed for the
2015 Food Security Supplement (FSS) to the Current Population Survey (CPS). The sample
included all households who completed a CPS interview. A nonresponse bias analysis was
conducted to determine whether nonresponse among different demographic groups may have
biased estimates. We investigated overall response rates, demographic subgroup response
rates, and demographic respondent and nonrespondent distributions.
Attached are similar tables found within this report using the data from the 2011, 2012, 2013,
and 2014 FSSs.
Our key findings for the 2015 FSS are:
•

The CPS household weighted response rate was 86.34 percent. The FSS household
weighted response rate was 75.04 percent. Thus, the overall weighted response rate
for the FSS sample was 64.79 percent.

•

For the CPS household estimates, excluding blanks (no responses), there are
significant differences in the response rates and respondent distributions for each of
the variables that we investigated except gender of reference person. Excluding the
blanks and missing values, the largest difference in response rates is seen for the age
of reference person. The largest differences between respondent and nonrespondent
distributions are within race, Hispanic origin, and age of reference person.

•

For the FSS household estimates, excluding blanks, there are significant differences
in the response rates and respondent distributions for each of the variables that we
investigated except type of living quarters and gender of reference person. The
largest difference in response rates is seen for the race of reference person. The
largest difference between respondent and nonrespondent distributions is within age
of reference person.

•

For FSS household estimates for characteristics only available for CPS respondents,
there are significant differences in the response rates and respondent distributions for
each of the variables that we investigated. The largest differences in response rates,
as well the largest difference in respondent and nonrespondent distributions, are seen
within income. Note: the largest difference in response rates for measure of labor
force of reference person is not significantly different from the largest difference in
response rates for income.

3

2. Introduction
The Office of Management and Budget (OMB) provides guidelines for conducting a
nonresponse bias study when the expected unit response rate of a survey is below 80 percent
(OMB, 2006). The CPS household response rates have historically been above 80 percent,
but the overall supplement response rates (which are the product of the CPS household and
FSS household response rates) are below this threshold.
This document provides results from our evaluation of nonresponse in the 2015 FSS to the
CPS and to provide nonresponse bias analysis tables for the 2011 through 2014 FSS. Its
purpose is to determine the existence of potential nonresponse bias in the 2015 FSS.
3. Overview of the Current Population Survey
The monthly CPS collects primarily labor force data about the civilian noninstitutionalized
population living in the United States. The institutionalized population, which is excluded
from the population universe, is composed primarily of the population in correctional
institutions and nursing homes (98 percent of the 4.0 million institutionalized people in
Census 2010). Interviewers ask questions concerning labor force participation about each
member 15 years old and over in sample households. For December 2015, the week
containing the nineteenth of the month is the interview week. The week containing the
twelfth is the reference week (i.e., the week about which the labor force questions are asked).
The CPS uses a multistage probability sample based on the results of the decennial census,
with coverage in all 50 states and the District of Columbia. The sample is continually
updated to account for new residential construction. When files from the most recent
decennial census become available, the Census Bureau gradually introduces a new sample
design for the CPS.
Every ten years, the CPS first-stage sample is redesigned 1 reflecting changes based on the
most recent decennial census. In the first stage of the sampling process, primary sampling
units (PSUs) 2 were selected for sample. In the 2010 sample design, the United States was
divided into 1,987 PSUs. These PSUs were then grouped into 852 strata. Within each
stratum, a single PSU was chosen for the sample, with its probability of selection
proportional to its population as of the most recent decennial census. In the case of strata
consisting of only one PSU, the PSU was chosen with certainty.
4. Overview of the 2015 Food Security Supplement to the Current Population Survey
In December 2015, in addition to the basic CPS questions, interviewers asked supplementary
questions of how much households spent for food, their use of Federal and community food
assistance programs, and whether they were able to afford enough food. The universe for
this supplement are households eligible for the basic CPS. This supplement allows for proxy
response. However, if at all possible, interviewers should interview the person within the
1
2

For detailed information on the 2010 sample redesign, please see (BLS, April 2014).
The PSUs correspond to substate areas (i.e., counties or groups of counties) that are geographically contiguous.

4

household who is responsible for buying or preparing food for the household. Households
with incomes below 185 percent of the poverty threshold are asked all supplement questions,
whereas households with incomes over 185 percent of the poverty threshold are asked only a
few questions, unless their answers identify them as “food insufficient or experiencing some
degree of food hardship” and make them eligible for the entire supplement (U.S. Census
Bureau, 2016).
The key estimates include:
• Concern about food adequacy
• Money for food
• Access to food
• Use of emergency food
• Food assistance program participation
• Food intake reductions or hunger
Key domains for this supplement are:
• Households
• Families
• Unrelated Individuals
• Age
5. Discussion of Nonresponse in the 2015 Food Security Supplement to the Current
Population Survey
Some degree of nonresponse bias and variance is a normal feature of almost all statistical
surveys. The FSS produces food security estimates using the answers from responding
households and persons. These food security estimates will be biased if answers from
respondents differ from the potential answers of nonrespondents. The magnitude of the bias
is a function of the response rate and differences between respondents and nonrespondents.
There were two ways that a household could be a nonrespondent to the FSS:
• The entire household did not respond to CPS (the occupants were not found at home
after repeated calls or are unavailable for some other reason).
• The household responded to CPS but did not respond to the FSS interview.
Because the FSS is directly linked to the CPS response rate, the CPS and FSS attempt to
minimize nonresponse bias by increasing response rates and adjusting weights for potential
differences between respondents and nonrespondents. We try to increase response rates
within CPS by conducting personal visit interviews for new and returning sample units,
mailing advance letters for all sample units, providing a Spanish language questionnaire for
potential respondents who do not speak English, allowing interpreters for potential
respondents who do not speak English or Spanish, training field representatives to gain
respondent cooperation, allowing proxy respondents in special circumstances, and mailing
follow-up letters to nonresponding households. We also help minimize nonresponse bias by
reducing respondent burden for the FSS by limiting the length of the survey.
5

We reduce the effects of respondent/nonrespondent differences through noninterview
weighting adjustments. These adjustments group respondents and nonrespondents into
adjustment cells, and the weights of the nonrespondents are reallocated to the respondents
within the adjustment cells.
CPS noninterview adjustment cells are formed by noninterview cluster (NICL) and central
city status. The NICLs are created based on sample PSUs that are similar in metropolitan
status and population size within the same state (U.S. Census Bureau, 2006). Metropolitan
status is defined as metropolitan or nonmetropolitan. Within metropolitan PSUs, a further
breakdown into “central city” and “not central city” is defined. This results in 127 NICLs
and 214 adjustment cells. These variables were chosen for the noninterview adjustment cells
because they are thought to be correlated with the CPS variables of interest.
FSS noninterview adjustment cells are defined to be the same as the CPS noninterview
adjustment cells.
Despite the measures taken to reduce nonresponse bias, there is likely still some amount of
nonresponse bias that we cannot correct without knowing the food security of the
nonrespondents.
6. Methods
Data
The data for this nonresponse bias analysis are from the December 2015 FSS to the CPS and
the December 2015 CPS. The U.S. Census Bureau conducts the CPS every month, although
this file has only December data. The December survey uses two sets of questions, the basic
CPS and a set of supplemental questions. The CPS, sponsored jointly by the Census Bureau
and the U.S. Bureau of Labor Statistics, is the country’s primary source of labor force
statistics for the entire population. The U.S. Department of Agriculture sponsors the
supplemental questions for December.
For a small number of variables, we had complete household information for all sample
households, including respondents and nonrespondents. These variables were primarily
limited to geographic and sampling data. There are also some variables with partial
information for the nonrespondents. Normal CPS processing uses previous responses to
demographic questions (when available) and does not re-ask those that are unlikely to change
from interview to interview. Any variables that have never been answered are imputed using
the hot deck imputation method. Hot deck imputation assigns a value collected for a person
with similar characteristics to the missing value. Where possible, we did not use allocated or
imputed values because we thought it would be better to use actual values for the
nonrespondents when comparing respondents to nonrespondents. Edited values were only
used for tenure and presence of children while investigating nonresponse bias within FSS
households by domain for characteristics that were only available for CPS respondents.

6

Weights
In the detailed weighting process for the CPS, baseweights were adjusted with the weighting
control factor (WCF), which accounts for subsampling in the field but does not include any
nonresponse/noninterview or population coverage adjustments. This subsampling-adjusted
baseweight is the weight used throughout this report for household calculations for CPS.
When computing rates and distributions for FSS households, the FSS baseweights, which are
the noninterview-adjusted weights from CPS, were used. Note that FSS weights are higher
than CPS weights because they include the CPS noninterview adjustment, which inflates
weights back up to the eligible weighted CPS household sample.
All numbers presented in the report are weighted unless otherwise noted.
Universe for the Estimates
We analyzed nonresponse for households. Since one person in each household responded for
the entire household, our analysis focused on household nonresponse within reference person
demographics, HU characteristics, and geography.
7. Limitations
There are some limitations to this analysis which may affect the results. In particular:
1. Using past data to assign subgroup variables to nonrespondents is not necessarily
accurate for households. Due to inmovers and outmovers, it is possible for demographic
variables that we get from past data to be out of date. However, we do not believe our
results need to be 100% accurate in order to show major differences between respondents
and nonrespondents. This assumes that the demographics of neighborhoods do not
change much in one and a half years.
2. Nonrespondents for CPS are never given the opportunity to respond to the FSS.
8. Response Rates
The response rates tell us the percentage of eligible sample cases that responded to the CPS
and the FSS. It is useful to compare response rates for different subgroups to understand the
magnitude of potential biases.
We produced weighted and unweighted response rates for the 2015 FSS by key domains and
variables. The overall FSS response rate is the product of CPS household response and FSS
household response rate.
Response rates are defined as:
𝑅𝑅𝑅𝑅 =

∑𝑖𝑖∈𝑠𝑠 𝑤𝑤𝑖𝑖 𝑅𝑅𝑖𝑖 𝐷𝐷𝑖𝑖
∑𝑖𝑖∈𝑠𝑠 𝑤𝑤𝑖𝑖 𝐷𝐷𝑖𝑖
7

where:
𝑤𝑤𝑖𝑖 = the appropriate weight (1 if unweighted) for the response rate calculation
𝑅𝑅𝑖𝑖 = the response indicator (1 for respondents, 0 for nonrespondents)
𝐷𝐷𝑖𝑖 = the domain indicator (1 if within domain of interest, 0 otherwise)
𝑠𝑠 = the set of all eligible households

Eligible households are all sample housing units (HUs) that did not receive Type B or Type
C (out-of-scope) outcome codes. Persons within group quarters (GQs) are treated as
individual HUs. The CPS interview data contains all eligible and non-eligible HUs, and the
FSS interview data contains only eligible HUs to the CPS.
For the December 2015 CPS, there were 61,356 occupied HUs eligible for the household
analysis. Of the 61,356 occupied households, 52,868 were interviewed. Of the 52,868
households that were interviewed for CPS, 39,948 also responded to the FSS.
Table 1 shows that the weighted percentage of households where at least one person
responded to CPS is 86.34 percent. From those responding households, 75.04 percent of the
weighted households responded to FSS, which is a comparable response rate to what is seen
for other CPS supplements. This results in an overall weighted FSS response rate of 64.79
percent.
Attachments A, B, C, and D contain Tables 1 through 7 using U.S. Census Bureau internal
data from CPS interviews for December 2014, December 2013, December 2012, and
December 2011, respectively.
Table 1: 2015 Food Security Unit Response Rates
Response Category

Count

Weighted
Sum 3*
136,720,437
114,896,602
99,206,383
114,896,602
86,218,073

Sampled CPS Households
73,942
Eligible CPS Households
61,356
CPS Household Response
52,868
Food Security Households
52,868
Food Security Household Response
39,948
Overall Food Security Response
Source: U.S. Census Bureau internal data from December 2015 CPS interviews.
*May not sum to totals due to rounding.

Response Rates
Unweighted*
Weighted*
86.17%

86.34%

75.56%
65.11%

75.04%
64.79%

Table 2 shows weighted response rates for all CPS households by domain. The standard
error column shows the standard error of the response rate. Standard errors are conditional
on the sample and represent expected variability in the response process, rather than
traditional sampling error. Replicate weights were used to calculate the standard errors to
account for the sample design.
For CPS households, CPS household weights prior to noninterview adjustments were used. For FSS households,
the FSS baseweights, which are the noninterview-adjusted weights from CPS, were used. Note that FSS weights are
higher than CPS weights because they include the CPS noninterview adjustment, which inflates weights back up to
the eligible weighted CPS household sample. The CPS Household Response row and Food Security Households
row are the same set of households but are presented twice to show the difference in weights used.

3

8

Table 2: Response Rates for December 2015 CPS Households
Unweighted
Weighted
Characteristic
Households
Households*

Weighted
Response
Rate (%)

Standard
Error
(%)

Significance
Grouping×

Type of Living Quarters
Housing Unit
58,065
109,401,649
86.21%
0.161%
B
Non-Housing Unit 4
3,249
5,417,526
90.22%
0.631%
A
Blank 5
42
77,427
6.98%
4.880%
C
Principal City Status
Principal City within
19,925
38,047,814
84.82%
0.289%
C
CBSA/MSA
Not Part of a Principal City
29,029
60,032,533
86.40%
0.239%
B
within CBSA/MSA
Outside of a CBSA/MSA
12,402
16,816,255
89.58%
0.415%
A
Region
Northeast
10,550
20,386,340
83.87%
0.453%
C
Midwest
12,096
25,494,863
87.89%
0.332%
A
South
22,866
43,098,059
86.33%
0.294%
B
West
15,844
25,917,339
86.79%
0.313%
B
Urban/Rural Status
Urban
46,752
91,564,538
85.77%
0.181%
B
Rural
13,635
21,645,150
89.04%
0.383%
A
Missing
969
1,686,914
83.11%
1.534%
C
Race of Reference Person
White Only
44,250
81,988,642
93.83%
0.129%
A
Black Only
6,337
12,528,960
91.51%
0.426%
B
Asian Only
2,255
4,608,554
93.29%
0.574%
A
Other Race/Two or More Races
1,571
2,295,193
92.65%
0.742%
A, B
Blank
6,943
13,475,254
32.53%
0.732%
C
Gender of Reference Person
Male
28,493
53,302,951
92.91%
0.176%
A
Female
28,493
53,527,431
92.81%
0.157%
A
Blank
4,370
8,066,219
0.07%
0.048%
B
Hispanic Origin of Reference Person
Hispanic
6,014
13,195,689
92.54%
0.357%
B
Non-Hispanic
50,295
92,330,420
93.55%
0.117%
A
Blank
5,047
9,370,493
6.62%
0.442%
C
Age of Reference Person
15-29
6,392
11,559,776
91.20%
0.397%
E
30-39
9,163
17,123,353
92.04%
0.327%
D, E
40-49
9,284
17,792,399
92.61%
0.297%
C, D
50-59
10,780
20,309,507
93.25%
0.276%
C
60-69
9,619
17,717,095
95.17%
0.251%
B
70+
9,136
17,140,926
97.28%
0.201%
A
Blank or Less than 15
6,982
13,253,545
29.81%
0.690%
F
Overall
61,356
114,896,602
86.34%
0.156%
Source: U.S. Census Bureau internal data from December 2015 CPS interviews.
* May not sum to totals due to rounding. For weighted percent of total sample, see Table 5.
×
Within each characteristic, response rates identified with the same letter are not significantly different at the
α=0.10 level. A indicates the highest response rates, B indicates the next highest rates, etc.
Non-HUs include quarters within rooming or boarding homes; non-permanent units in transient hotels, motels,
etc.; unoccupied sites for mobile homes, trailers, or tents; group quarters in school dormitories; and other units
that are not defined to be housing units.
5
Blank indicates that the living quarters type was either not identified or was identified with an invalid code.
4

9

Excluding the blanks and missing values, largest difference in response rates for the CPS
subgroups is 6.08 percent, seen for the age of reference person, where age group 70+ has a
response rate of 97.28 percent versus 91.20 percent for age group 15-29. Additionally,
households in non-HUs have a higher response rate than HUs, midwest had the highest
response rate among the regions, rural has a higher response rate than urban, Asian only and
White only have higher response rates than Black only (none of which are significantly
different from other race/two or more races), non-Hispanic has higher response rates than
Hispanic, and households with references person aged 70+ have the highest response rates
among the age groups.
The response rate for blanks within the demographic subgroups is so low because these
demographic items are collected during the interview, resulting in a large portion of the
household nonrespondents falling within these blank categories instead of the categories
where they belong. Any household with a blank value within the demographic subgroups
above indicates that the household has not previously responded to the CPS or never
provided responses to those demographic questions in previous interviews. The nonresponse
in the non-blank demographic categories are from households which had previously
responded to the CPS and provided a valid response (non-blank) within the demographic
category.
Table 2 shows standard errors which facilitate hypothesis testing of differential response
rates. However, the practical significance of response rate differences is usually driven more
by the magnitude of the difference. Therefore, excluding blanks, if the nonrespondents are
different from respondents, principal city status and age of reference person have the most
potential for bias.
Table 3 shows weighted response rates for all FSS households by domain.

10

Table 3: Response Rates for 2015 FSS Households
Unweighted
Weighted
Characteristic
Households
Households*

Weighted
Response
Rate (%)

Standard
Error (%)

Significance
Grouping×

Type of Living Quarters
Housing Unit
49,950
109,311,325
75.09%
0.299%
A
Non-Housing Unit 6
2,916
5,579,233
73.98%
1.111%
A
Blank 7
2
6,043
100.00%
0.000%
N+
Principal City Status
Principal City within
16,853
38,013,576
73.57%
0.489%
C
CBSA/MSA
Not Part of a Principal City
25,064
60,032,533
75.22%
0.358%
B
within CBSA/MSA
Outside of a CBSA/MSA
10,951
16,850,493
77.71%
0.909%
A
Region
Northeast
8,811
20,378,347
72.72%
0.707%
C
Midwest
10,608
25,494,863
78.65%
0.541%
A
South
19,707
43,106,053
74.17%
0.453%
B
West
13,742
25,917,339
74.76%
0.612%
B
Urban/Rural Status
Urban
40,040
91,418,066
74.54%
0.312%
B
Rural
12,014
21,866,967
77.17%
0.713%
A
Missing
814
1,611,569
74.31%
2.267%
A, B
Race of Reference Person
White Only
41,455
88,884,720
76.60%
0.313%
B
Black Only
5,793
13,373,411
69.59%
0.681%
C
Asian Only
2,106
5,046,915
70.04%
1.235%
C
Other Race/Two or More Races
1,452
2,455,425
80.03%
1.215%
A
Blank
2,062
5,136,131
64.70%
1.232%
D
Gender of Reference Person
Male
26,456
57,316,521
75.21%
0.375%
A
Female
26,410
57,573,238
74.88%
0.343%
A
Blank
2
6,842
53.06%
39.217%
A+
Hispanic Origin of Reference Person
Hispanic
5,562
14,276,574
73.85%
0.763%
B
Non-Hispanic
46,974
99,904,757
75.63%
0.304%
A
Blank
332
715,271
16.68%
2.924%
C
Age of Reference Person
15-29
5,826
12,229,482
75.72%
0.679%
C
30-39
8,424
18,300,118
76.61%
0.544%
B, C
40-49
8,577
19,100,087
75.72%
0.530%
C
50-59
10,042
21,909,070
75.70%
0.546%
C
60-69
9,159
19,486,260
78.25%
0.445%
A
70+
8,864
19,283,847
77.36%
0.555%
A, B
Blank or Less than 15
1,976
4,587,738
37.59%
1.202%
D
Overall
52,868
114,896,602
75.04%
0.294%
Source: U.S. Census Bureau internal data from December 2015 CPS interviews.
* May not sum to totals due to rounding. For weighted percent of total sample, see Table 6.
N indicates that the estimate is not comparable.
+
Exercise caution: The sample size is extremely small, leading to unreliable estimates.
×
Within each characteristic, response rates identified with the same letter are not significantly different at the
α=0.10 level. A indicates the highest response rates, B indicates the next highest rates, etc.
6
7

See footnote 4.
See footnote 5.

11

For the FSS household estimates, we have no significant differences among the gender
response rates, and excluding the blanks, the response rates between the type of living
quarters are not significantly different. Excluding the blanks and missing values, the largest
difference in response rates for the FSS subgroups is 10.44 percent, seen for the race of
reference person, where other race/two or more races has a response rate of 80.03 percent
versus 69.59 percent for black only (not significantly different from Asian only).
Additionally, households outside the CBSA/MSA have the highest response rate within
principal city status, rural has higher response rates than urban, non-Hispanic has higher
response rates than Hispanic, and households with reference person aged 60-69 have the
highest response rates among the age groups (not significantly different from reference
person aged 70+).
Again, although Table 3 shows standard errors which facilitate hypothesis testing, the
practical significance of response rate differences is driven more by the magnitude of the
difference than the sample size. Therefore, excluding blanks, if the nonrespondents are
different from respondents, race of reference person has the most potential for bias.
Table 4 shows weighted response rates for all FSS households by domain for characteristics
that were only available for CPS respondents.

12

Table 4: Response Rates for 2015 FSS Households for Characteristics Only Available for Responding
CPS Households
Unweighted
Weighted
Weighted
Standard Significance
Characteristic
Households
Households*
Response Error (%)
Grouping×
Rate (%)
Tenure (Edited)
Owned or Mortgage
35,107
75,519,119
76.02%
0.308%
A
Rented for Cash
17,006
37,938,239
73.11%
0.505%
B
No Cash Rent
755
1,439,245
74.60%
2.236%
A, B
Family Income
Less than $10,000
2,912
6,062,107
79.12%
1.025%
C
$10,000-$19,999.99
4,431
9,268,125
82.06%
0.669%
B
$20,000-$29,999.99
4,703
10,144,735
81.80%
0.697%
B
$30,000-$39,999.99
4,386
9,461,316
82.64%
0.630%
B
$40,000-$49,999.99
3,373
7,180,189
83.07%
0.710%
A, B
$50,000-$59,999.99
3,241
6,921,466
82.57%
0.735%
B
$60,000-$74,999.99
3,952
8,586,429
82.54%
0.713%
B
$75,000-$99,999.99
4,435
9,594,610
84.32%
0.622%
A
$100,000-$149,999.99
4,570
10,187,341
82.96%
0.598%
A, B
$150,000+
3,740
8,589,036
83.16%
0.740%
A, B
Blank or Don’t Know
2,060
4,586,740
50.86%
1.333%
D
Refused
11,065
24,314,509
53.12%
0.653%
D
Household Type
Husband/Wife Primary Family
25,608
55,941,698
75.79%
0.338%
A
Unmarried Householder
8,798
19,635,736
73.33%
0.575%
C
Primary Family
Primary Individual
18,380
39,173,946
74.81%
0.438%
B
Group Quarters with Family
13
19,468
56.38%
17.506%
A, B, C+
Group Quarters without Family
69
125,754
82.28%
7.907%
A, B, C
Child(ren) Present (Edited)
No
39,222
84,863,020
74.76%
0.321%
B
Yes
13,646
30,033,581
75.83%
0.475%
A
Measure of Labor Force of Reference Person
Employed
31,385
68,285,842
75.05%
0.349%
B
Unemployed
1,326
2,917,619
77.98%
1.245%
A
Not in Labor Force
19,756
42,849,040
75.20%
0.426%
B
Blank
401
844,101
56.25%
3.062%
C
Overall
52,868
114,896,602
75.04%
0.294%
Source: U.S. Census Bureau internal data from December 2015 CPS interviews.
* May not sum to totals due to rounding. For weighted percent of total sample, see Table 7.
+
Exercise caution: The sample size is small, leading to unreliable estimates.
×
Within each characteristic, response rates identified with the same letter are not significantly different at the
α=0.10 level. A indicates the highest response rates, B indicates the next highest rates, etc.

For the FSS household estimates of characteristics only available for CPS respondents,
excluding the blanks, refusals, and GQ groups, the largest difference in response rates for any
of the subgroups is 5.20 percent 8, seen for the income variable, where the rates range from
79.12 percent for income less than $10,000 to 84.32 percent for income $75000-$99,999.99.
(Note: Within household type, while the difference between husband/wife primary family
and GQs with family is 25.90 percent, it is not significantly different from 0, which is why
the largest difference is found within the income variable.) For the tenure category,
Excluding blanks, this difference is not significantly different from 2.93 percent, seen for the measure of labor
force of reference person, where the rates range from 75.05 percent for employed to 77.98 percent for unemployed.

8

13

own/mortgage has a higher response rate than cash rent (neither are significantly different
from no cash rent), and within the children present category, households having children had
a higher response rate than households without children. Among the non-GQ households,
husband/wife primary families had the highest response rate. For the measure of labor force
of reference person category, unemployed reference persons had a higher response rate than
reference person who were unemployed or not in the labor force.
Again, although Table 4 shows standard errors which facilitate hypothesis testing, the
practical significance of response rate differences is driven more by the magnitude of the
difference than the sample size. Therefore, excluding blanks, if the nonrespondents are
different from respondents, among the categories in Table 4, income has the most potential
for bias.
9. Respondent Distributions
Respondent and nonrespondent distributions show the relative percent of members of a
domain subset within respondents and nonrespondents separately. This is different than the
response rates, which are the relative percent of respondents within the different domain
subsets. We used chi-square tests to determine if the respondent and nonrespondent
distributions differed.
Respondent distributions are defined as:
𝑅𝑅𝑅𝑅 =

∑𝑖𝑖∈𝑠𝑠 𝑤𝑤𝑖𝑖 𝑅𝑅𝑖𝑖 𝐷𝐷𝑖𝑖
∑𝑖𝑖∈𝑠𝑠 𝑤𝑤𝑖𝑖 𝑅𝑅𝑖𝑖

This definition assumes the same eligibility criteria, weights, and indicators as the response
rate calculations in the previous section. Nonrespondent distributions use the same formula,
but with the 𝑅𝑅𝑖𝑖 variable indicating nonrespondents instead of respondents. The chi-square
test statistics were calculated using replicate weights to account for the sample design.
Table 5 shows the percent of total sample distribution as well as comparisons of respondent
and nonrespondent distributions for CPS households within the different domain subgroups.

14

Table 5: Respondent and Nonrespondent Distributions for December 2015 CPS Households
Household
Household
% of Total
Chi-Square
Characteristic
Respondent Nonrespondent
P-value
Sample*
Statistic (df)
%*
%*
Type of Living Quarters
Housing Unit
95.22%
95.07%
96.16%
187.5428 (2)
< 0.0001
Non-Housing Unit
4.72%
4.93%
3.38%
[28.9847 (1) < 0.0001]
Blank
0.07%
0.01%
0.46%
Principal City Status
Principal City within
33.11%
32.53%
36.80%
CBSA/MSA
Not Part of a Principal City
73.2449 (2)
< 0.0001
52.25%
52.28%
52.03%
within CBSA/MSA
Outside of a CBSA/MSA
14.64%
15.18%
11.17%
Region
Northeast
17.74%
17.23%
20.96%
Midwest
22.19%
22.59%
19.67%
55.7301 (3)
< 0.0001
South
37.51%
37.51%
37.54%
West
22.56%
22.67%
21.83%
Urban/Rural Status
Urban
79.69%
79.16%
83.07%
56.9957 (2)
< 0.0001
Rural
18.84%
19.43%
15.11%
[47.8369 (1)
< 0.0001]
Missing
1.47%
1.41%
1.82%
Race of Reference Person
White Only
71.36%
77.55%
32.23%
Black Only
10.90%
11.56%
6.78%
Asian Only
4.01%
4.33%
1.97%
14,363.5781 (4)
< 0.0001
Other Race/
[40.2352 (3)
< 0.0001]
2.00%
2.14%
1.08%
Two or More Races
Blank
11.73%
4.42%
57.94%
Gender of Reference Person
Male
46.39%
49.92%
24.09%
22,700.5337 (2)
< 0.0001
Female
46.59%
50.07%
24.54%
[0.1877 (1)
0.6649]
Blank
7.02%
0.01%
51.37%
Hispanic Origin of Reference Person
Hispanic
11.48%
12.31%
6.27%
21,621.5060 (2)
< 0.0001
Non-Hispanic
80.36%
87.06%
37.96%
[7.7121 (1)
0.0055]
Blank
8.16%
0.63%
55.77%
Age of Reference Person
15-29
10.06%
10.63%
6.49%
30-39
14.90%
15.89%
8.68%
40-49
15.49%
16.61%
8.38%
16,161.2963 (6)
< 0.0001
50-59
17.68%
19.09%
8.73%
[270.7361 (5) < 0.0001]
60-69
15.42%
17.00%
5.45%
70+
14.92%
16.81%
2.97%
Blank or Less than 15
11.54%
3.98%
59.29%
Overall
100%
100%
100%
Source: U.S. Census Bureau internal data from December 2015 CPS interviews.
* May not sum to totals due to rounding.
[] The values within brackets are the chi-square statistic, df, and p-value when the blanks/missings are excluded
from the chi-square test.

The chi-square tests for CPS households showed significant differences (at the α=0.10 level)
between respondent and nonrespondent distributions for all variables. Simply looking at the
distributions for the race, gender, Hispanic origin, and age of reference person, you can tell
15

that there are big differences between the respondent and nonrespondent distributions, which
corresponds to the magnitude of the chi-square test statistics (14,363.58, 22,700.53,
21,621.51, and 16,161.30, respectively). However, when you exclude the blanks from the
chi-square test, the gender of the reference person no longer has a significant difference
between the two distributions. Note: The chi-square tests only indicate that the distributions
of respondents and nonrespondents differ but do not necessarily indicate a nonresponse bias
problem. These differences will only cause bias if the respondents and nonrespondents
report differing rates of food security.
Even though there are significant differences between the respondents and nonrespondents,
the differences might not be large enough to cause meaningful differences in estimates.
Furthermore, weighting adjustments might also minimize the impact of some differences.
Because the CPS noninterview adjustments take NICL and central city status into account,
the principal city status and region differences may be reduced within those adjustments.
Table 6 shows the percent of total sample distribution as well as comparisons of respondent
and nonrespondent distributions for FSS households within the different domain subgroups.

16

Table 6: Respondent and Nonrespondent Distributions for 2015 FSS Households
Household
Household
% of Total
Chi-Square
Characteristic
Respondent Nonrespondent
P-value
Sample*
Statistic (df)
%*
%*
Type of Living Quarters
Housing Unit
95.14%
95.21%
94.94%
N/A (2)
N/A
Non-Housing Unit
4.86%
4.79%
5.06%
[1.0051 (1)
0.3161]
Blank
0.01%
0.01%
0.00%
Principal City Status
Principal City within
33.09%
32.44%
35.04%
CBSA/MSA
Not Part of a Principal City
18.8838 (2)
< 0.0001
52.25%
52.38%
51.87%
within CBSA/MSA
Outside of a CBSA/MSA
14.67%
15.19%
13.10%
Region
Northeast
17.74%
17.19%
19.38%
Midwest
22.19%
23.26%
18.98%
54.4022 (3)
< 0.0001
South
37.52%
37.08%
38.83%
West
22.56%
22.47%
22.81%
Urban/Rural Status
Urban
79.57%
79.04%
81.15%
12.9933 (2)
0.0015
Rural
19.03%
19.57%
17.41%
[11.1872 (1)
0.0008]
Missing
1.40%
1.39%
1.44%
Race of Reference Person
White Only
77.36%
78.97%
72.52%
Black Only
11.64%
10.79%
14.18%
Asian Only
4.39%
4.10%
5.27%
243.4677 (4)
< 0.0001
Other Race/
[143.6142 (3)
< 0.0001]
2.14%
2.28%
1.71%
Two or More Races
Blank
4.47%
3.85%
6.32%
Gender of Reference Person
Male
49.89%
50.00%
49.55%
1.1310 (2)
0.5681
Female
50.11%
50.00%
50.43%
[0.6275 (1)
0.4283]
Blank
0.01%
Z%
0.01%
Hispanic Origin of Reference Person
Hispanic
12.43%
12.23%
13.02%
316.5346 (2)
< 0.0001
Non-Hispanic
86.95%
87.63%
84.90%
[5.2768 (1)
0.0216]
Blank
0.62%
0.14%
2.08%
Age of Reference Person
15-29
10.64%
10.74%
10.35%
30-39
15.93%
16.26%
14.92%
40-49
16.62%
16.77%
16.17%
1,385.6153 (6)
< 0.0001
50-59
19.07%
19.24%
18.56%
[23.0130 (5)
0.0003]
60-69
16.96%
17.68%
14.78%
70+
16.78%
17.30%
15.22%
Blank or Less than 15
3.99%
2.00%
9.98%
Overall
100%
100%
100%
Source: U.S. Census Bureau internal data from December 2015 CPS interviews.
Note: The chi-square test for type of living quarters could not be calculated due to a frequency of 0 within the
blank/nonrespondent cell.
Z: value rounds to zero but is not zero.
* May not sum to totals due to rounding.
[] The values within brackets are the chi-square statistic, df, and p-value when the blanks/missings are
excluded from the chi-square test.

17

The chi-square tests for FSS households showed significant differences (at the α=0.10 level)
for the distributions of all variables except type of living quarters and gender of reference
person. Simply looking at the distributions for age of the reference person, you can tell that
there are big differences between the respondent and nonrespondent distributions, which
correspond to the magnitude of the chi-square test statistic (1,385.62).
As mentioned for CPS household respondent distributions, the chi-square tests only indicate
that the distributions of respondents and nonrespondents differ but do not necessarily indicate
a nonresponse bias problem. Furthermore, weighting adjustments might minimize the impact
of some differences. Because the FSS noninterview adjustments take NICL and central city
status into account, the principal city status and region differences may be reduced within
those adjustments.
Table 7 shows the percent of total sample distribution as well as comparisons of respondent
and nonrespondent distributions for FSS households within the different domain subgroups
for characteristics that were only available for CPS respondents.

18

Table 7: Respondent and Nonrespondent Distributions for 2015 FSS Households for Characteristics Only
Available for Responding CPS Households
Household
Household
% of Total
Chi-Square
Characteristic
Respondent Nonrespondent
P-value
Sample*
Statistic (df)
%*
%*
Tenure (Edited)
Owned or Mortgage
65.73%
66.58%
63.15%
31.0653 (2) < 0.0001
Rented for Cash
33.02%
32.17%
35.57%
No Cash Rent
1.25%
1.25%
1.27%
Family Income
Less than $10,000
5.28%
5.56%
4.41%
$10,000-$19,999.99
8.07%
8.82%
5.80%
$20,000-$19,999.99
8.83%
9.63%
6.44%
$30,000-$19,999.99
8.23%
9.07%
5.73%
$40,000-$19,999.99
6.25%
6.92%
4.24%
$50,000-$19,999.99
6.02%
6.63%
4.21%
3,577.4282 (11) < 0.0001
$60,000-$19,999.99
7.47%
8.22%
5.23%
[28.3240 (9)
0.0008]
$75,000-$19,999.99
8.35%
9.38%
5.25%
$100,000-$19,999.99
8.87%
9.80%
6.05%
$150,000-$19,999.99
7.48%
8.28%
5.04%
Blank or Don’t Know
3.99%
2.71%
7.86%
Refused
21.16%
14.98%
39.75%
Household Type
Husband/Wife Primary
48.69%
49.18%
47.22%
Family
Unmarried Householder
17.09%
16.70%
18.26%
Primary Family
17.3372 (4)
0.0017
Primary Individual
34.10%
33.99%
34.41%
{15.0907 (3)
0.0017}
Group Quarters with
0.02%
0.01%
0.03%
Family
Group Quarters without
0.11%
0.12%
0.08%
Family
Child(ren) Present (Edited)
No
73.86%
73.59%
74.69%
4.4606 (1)
0.0347
Yes
26.14%
26.41%
25.31%
Measure of Labor Force of Reference Person
Employed
59.43%
59.44%
59.42%
Unemployed
2.54%
2.64%
2.24%
56.9167 (3)
< 0.0001
Not in Labor Force
37.29%
37.37%
37.06%
[4.3552 (2)
0.1133]
Blank
0.73%
0.55%
1.29%
Overall
100%
100%
100%
Source: U.S. Census Bureau internal data from December 2015 CPS interviews.
* May not sum to totals due to rounding.
[] The values within brackets are the chi-square statistic, df, and p-value when the blanks and refusals are
excluded from the chi-square test.
{} The values within braces are the chi-square statistic, df, and p-value when the GQ groups are combined.

The chi-square tests for FSS household estimates of characteristics only available for CPS
respondents showed significant differences (at the α=0.10 level) for the distributions of all
variables. Simply looking at the distributions for family income, you can tell that there are
big differences between the respondent and nonrespondent distributions, which correspond to
the magnitude of the chi-square test statistic (3,577.43). Even after excluding the blanks and
refusals, there is still a significant difference between the respondent and nonrespondent
distributions for income. However, when you exclude the blanks from the chi-square test,
19

the unedited measure of the labor force of the reference person no longer has a significant
difference between the two distributions.
As mentioned previously, the chi-square tests only indicate that the distributions of
respondents and nonrespondents differ but do not necessarily indicate a nonresponse bias
problem. Furthermore, weighting adjustments might minimize the impact of some
differences.
10. Discussions and Conclusions
This analysis found evidence of potential nonresponse bias for both CPS and FSS
households. For CPS, there is potential nonresponse bias for all investigated characteristics
except possibly for gender of reference person. For FSS, there is potential nonresponse bias
for all investigated characteristics except gender of reference person and type of living
quarters.
Excluding the blanks and missing values, the largest difference in response rates for the CPS
subgroups is 6.08 percent, seen for the age of reference person, where age group 70+ has a
response rate of 97.28 percent versus 91.20 percent for age group 15-29. For respondent and
nonrespondent distributions within CPS households, the largest differences are seen within
race, Hispanic origin, and age of reference person.
Excluding the blanks and missing values, the largest difference in response rates for the FSS
subgroups is 10.44 percent, seen for the race of reference person, where other race/two or
more races has a response rate of 80.03 percent versus 69.59 percent for black only (not
significantly different from Asian only). For respondent and nonrespondent distributions
within FSS households, the largest difference is seen within age of reference person.
Among the estimates for the FSS households for characteristics only available for CPS
respondents, the largest differences in response rates 9 and between the respondent and
nonrespondent distributions are seen within income.
Using the information learned from this analysis, discussions should be had with the sponsor
regarding enhancements to the weighting process. The findings suggest that research be
done into the possible inclusion of other geographic and demographic characteristics into the
household noninterview adjustments for the FSS. Research could be conducted into whether
the nonresponse adjustment should include the geographic and demographic characteristics
that were investigated in this report to determine if they can help reduce the nonresponse
bias. Some other potential characteristics that may be related to food security to consider
including in the noninterview adjustment may be block and/or tract planning database
variables, which would include geographic and demographic variables based on the location
of the sampled household, such as percent of population that is Hispanic, percent of HUs
where no one lives regularly (vacant HUs), percent of population that is below the poverty
level. Note: planning database variables are estimates using American Community Survey or
Census data.
9

See footnote 8.

20

11. References
Bureau of Labor Statistics. (April 2014). “Redesign of the Sample for the Current Population
Survey.” http://www.bls.gov/cps/sample_redesign_2014.pdf
Office of Management and Budget. (2006). “Standards and Guidelines for Statistical Surveys.”
https://georgewbushwhitehouse.archives.gov/omb/inforeg/statpolicy/standards_stat_surveys.pdf
U.S. Census Bureau. (2006). Current Population Survey: Design and Methodology. Technical
Paper 66. Washington, DC: Government Printing Office.
http://www.census.gov/prod/2006pubs/tp-66.pdf
U.S. Census Bureau. (2016). “Current Population Survey Food Security Supplement.”
https://catalog.data.gov/dataset/current-population-survey-food-security-supplement
All online references last accessed: October 31, 2017

21

Attachment A
Key differences between these 2014 tables and the 2015 tables are:
•

For December 2015 CPS, there is no significant difference in response rates for
gender of reference person, excluding blanks (Table 2). This is not the case for
December 2014 CPS (Table A-2).

•

For 2015 FSS, there is no significant difference in response rates for gender of
reference person, excluding blanks (Table 3). This is not the case for 2014 FSS
(Table A-3).

•

For 2015 FSS within responding CPS households, there is a significant difference in
response rates for child(ren) present (Table 4). This is not the case in 2014 FSS
within responding CPS households (Table A-4).

•

For December 2015 CPS, the respondent and nonrespondent distributions (excluding
blanks/missings) are not significantly different at the α=0.10 level for gender of
reference person (Table 5). This is not the case for December 2014 CPS (Table A5).

•

For 2015 FSS, the respondent and nonrespondent distributions (excluding
blanks/missings) are not significantly different at the α=0.10 level for gender of
reference person (Table 6). This is not the case for 2014 FSS (Table A-6).

•

For 2015 FSS within responding CPS households, the respondent and nonrespondent
distributions are significantly different at the α=0.10 level for child(ren) present
(Table 7). This is not the case for 2014 FSS within responding CPS households
(Table A-7).

•

For 2015 FSS within responding CPS households, the respondent and nonrespondent
distributions (excluding blanks/missings) are not significantly different at the α=0.10
level for measure of labor force of reference person (Table 7). This is not the case
for 2014 FSS within responding CPS households (Table A-7).

Table A-1: 2014 Food Security Unit Response Rates
Response Category

Count

Weighted
Sum 10*
132,791,561
112,013,896
98,076,311
112,013,896
89,013,820

Sampled CPS Households
73,842
Eligible CPS Households
61,677
CPS Household Response
53,901
Food Security Households
53,901
Food Security Household Response
43,253
Overall Food Security Response
Source: U.S. Census Bureau internal data from December 2014 CPS interviews.
* May not sum to totals due to rounding.

10

See footnote 3.

A-1

Response Rates
Unweighted*
Weighted*
87.39%

87.56%

80.25%
70.13%

79.47%
69.58%

Attachment A
Table A-2: Response Rates for December 2014 CPS Households
Unweighted
Weighted
Characteristic
Households
Households*

Weighted
Response
Rate (%)

Standard
Error
(%)

Significance
Grouping×

Type of Living Quarters
Housing Unit
58,553
106,822,202
87.43%
0.187%
B
Non-Housing Unit 11
3,087
5,119,044
91.07%
0.683%
A
Blank 12
37
72,651
25.43%
9.151%
C
Principal City Status
Principal City within
19,597
36,295,447
86.52%
0.315%
C
CBSA/MSA
Not Part of a Principal City
28,678
57,067,389
87.45%
0.231%
B
within CBSA/MSA
Outside of a CBSA/MSA
13,402
18,651,060
89.90%
0.484%
A
Region
Northeast
11,457
20,072,141
83.79%
0.440%
B
Midwest
12,842
25,253,261
88.53%
0.394%
A
South
22,041
41,786,724
88.09%
0.298%
A
West
15,337
24,901,770
88.72%
0.306%
A
Urban/Rural Status
Urban
44,512
84,304,411
87.18%
0.198%
B
Rural
13,139
20,326,339
89.72%
0.397%
A
Missing
4,026
7,383,146
85.94%
0.646%
C
Race of Reference Person
White Only
44,935
80,788,809
94.55%
0.134%
A
Black Only
6,323
12,245,388
92.58%
0.415%
C
Asian Only
2,295
4,421,669
94.10%
0.577%
A, B
Other Race/Two or More Races
1,540
2,208,697
93.31%
0.701%
B, C
Blank
6,584
12,349,334
33.47%
0.762%
D
Gender of Reference Person
Male
28,924
52,495,216
93.91%
0.158%
A
Female
28,702
52,200,009
93.44%
0.162%
B
Blank
4,051
7,318,671
0.03%
0.021%
C
Hispanic Origin of Reference Person
Hispanic
6,033
12,892,470
93.09%
0.388%
B
Non-Hispanic
50,925
90,636,797
94.32%
0.121%
A
Blank
4,719
8,484,630
6.96%
0.433%
C
Age of Reference Person
15-29
6,633
11,678,587
91.83%
0.389%
E
30-39
9,177
16,883,352
93.07%
0.308%
D
40-49
9,583
17,784,729
93.60%
0.265%
D
50-59
11,157
20,225,358
94.34%
0.243%
C
60-69
9,346
16,680,016
96.04%
0.236%
B
70+
9,160
16,528,193
97.36%
0.186%
A
Blank or Less than 15
6,621
12,233,662
31.05%
0.617%
F
Overall
61,677
112,013,896
87.56%
0.185%
Source: U.S. Census Bureau internal data from December 2014 CPS interviews.
* May not sum to totals due to rounding. For weighted percent of total sample, see Table A-5.
×
Within each characteristic, response rates identified with the same letter are not significantly different at the
α=0.10 level. A indicates the highest response rates, B indicates the next highest rates, etc.

11
12

See footnote 4.
See footnote 5.

A-2

Attachment A
Table A-3: Response Rates for 2014 FSS Households
Unweighted
Weighted
Characteristic
Households
Households*

Weighted
Response
Rate (%)

Standard
Error (%)

Significance
Grouping×

Type of Living Quarters
Housing Unit
51,089
106,758,410
79.39%
0.261%
A
Non-Housing Unit 13
2,803
5,234,116
80.89%
0.955%
A
Blank 14
9
21,370
100.00%
0.000%
N+
Principal City Status
Principal City within
16,892
36,315,277
78.06%
0.443%
B
CBSA/MSA
Not Part of a Principal City
25,063
57,061,201
78.80%
0.308%
B
within CBSA/MSA
Outside of a CBSA/MSA
11,946
18,637,418
84.25%
0.742%
A
Region
Northeast
9,564
20,072,141
77.94%
0.519%
C
Midwest
11,386
25,253,261
81.45%
0.621%
A
South
19,420
41,786,724
79.63%
0.457%
B
West
13,531
24,901,770
78.41%
0.570%
C
Urban/Rural Status
Urban
38,756
84,235,496
78.76%
0.277%
B
Rural
11,699
20,533,685
82.84%
0.614%
A
Missing
3,446
7,244,716
78.15%
0.980%
B
Race of Reference Person
White Only
42,408
87,060,508
80.73%
0.271%
A
Black Only
5,843
13,039,573
76.52%
0.666%
B
Asian Only
2,162
4,802,349
75.94%
1.084%
B
Other Race/Two or More Races
1,443
2,346,335
81.12%
1.236%
A
Blank
2,045
4,765,132
67.24%
1.183%
C
Gender of Reference Person
Male
27,101
56,293,466
78.92%
0.368%
B
Female
26,798
55,718,397
80.02%
0.297%
A
Blank
2
2,034
0.00%
0.000%
N+
Hispanic Origin of Reference Person
Hispanic
5,599
13,742,852
77.95%
0.592%
B
Non-Hispanic
47,964
97,599,219
80.10%
0.268%
A
Blank
338
671,826
18.27%
2.556%
C
Age of Reference Person
15-29
6,070
12,265,994
79.79%
0.689%
B
30-39
8,537
17,988,267
80.01%
0.539%
B
40-49
8,950
19,011,844
80.16%
0.519%
B
50-59
10,515
21,786,929
80.29%
0.521%
B
60-69
8,957
18,273,900
82.99%
0.454%
A
70+
8,910
18,345,606
82.22%
0.480%
A
Blank or Less than 15
1,962
4,341,357
42.68%
1.298%
C
Overall
53,901
112,013,896
79.47%
0.261%
Source: U.S. Census Bureau internal data from December 2014 CPS interviews.
* May not sum to totals due to rounding. For weighted percent of total sample, see Table A-6.
N indicates that the estimate is not comparable.
+
Exercise caution: The sample size is extremely small, leading to unreliable estimates.
×
Within each characteristic, response rates identified with the same letter are not significantly different at the
α=0.10 level. A indicates the highest response rates, B indicates the next highest rates, etc.
13
14

See footnote 4.
See footnote 5.

A-3

Attachment A
Table A-4: Response Rates for 2014 FSS Households for Characteristics Only Available for Responding
CPS Households
Unweighted
Weighted
Weighted
Standard Significance
Characteristic
Households
Households*
Response Error (%)
Grouping×
Rate (%)
Tenure (Edited)
Owned or Mortgage
35,771
73,474,852
79.96%
0.304%
B
Rented for Cash
17,357
37,097,582
78.34%
0.422%
C
No Cash Rent
773
1,441,462
83.36%
1.612%
A
Family Income
Less than $10,000
3,275
6,718,702
83.03%
0.742%
D
$10,000-$19,999.99
4,827
9,874,916
85.82%
0.597%
B, C
$20,000-$29,999.99
4,970
10,117,035
87.58%
0.551%
A
$30,000-$39,999.99
4,808
10,086,378
87.13%
0.616%
A, B
$40,000-$49,999.99
3,565
7,232,210
85.97%
0.704%
B, C
$50,000-$59,999.99
3,344
6,772,380
86.71%
0.701%
A, C
$60,000-$74,999.99
4,079
8,334,511
87.45%
0.606%
A
$75,000-$99,999.99
4,636
9,369,489
87.16%
0.635%
A, B
$100,000-$149,999.99
4,615
9,528,595
86.65%
0.568%
A, C
$150,000+
3,687
8,056,046
85.56%
0.647%
C
Blank or Don’t Know
1,931
4,170,815
58.44%
1.426%
E
Refused
10,164
21,752,820
55.99%
0.708%
E
Household Type
Husband/Wife Primary Family
26,259
54,789,600
79.76%
0.325%
B
Unmarried Householder
8,912
18,952,790
78.27%
0.535%
C
Primary Family
Primary Individual
18,671
38,170,621
79.61%
0.395%
B
Group Quarters with Family
8
12,053
100.00%
0.000%
N+
Group Quarters without Family
51
88,832
90.45%
5.032%
A
Child(ren) Present (Edited)
No
39,772
82,033,372
79.48%
0.301%
A
Yes
14,129
29,980,524
79.44%
0.364%
A
Measure of Labor Force of Reference Person
Employed
32,144
66,515,098
79.35%
0.352%
B
Unemployed
1,547
3,289,593
85.04%
0.994%
A
Not in Labor Force
19,796
41,387,810
79.65%
0.350%
B
Blank
414
821,395
56.91%
3.124%
C
Overall
53,901
112,013,896
79.47%
0.261%
Source: U.S. Census Bureau internal data from December 2014 CPS interviews.
* May not sum to totals due to rounding. For weighted percent of total sample, see Table A-7.
N indicates that the estimate is not comparable.
+
Exercise caution: The sample size is small, leading to unreliable estimates.
×
Within each characteristic, response rates identified with the same letter are not significantly different at the
α=0.10 level. A indicates the highest response rates, B indicates the next highest rates, etc.

A-4

Attachment A
Table A-5: Respondent and Nonrespondent Distributions for December 2014 CPS Households
Household
Household
% of Total
Chi-Square
Characteristic
Respondent Nonrespondent
Sample*
Statistic (df)
%*
%*
Type of Living Quarters
Housing Unit
95.37%
95.23%
96.33%
107.0932 (2)
Non-Housing Unit
4.57%
4.75%
3.28%
[21.2686 (1)
Blank
0.06%
0.02%
0.39%
Principal City Status
Principal City within
32.40%
32.02%
35.11%
CBSA/MSA
Not Part of a Principal City
36.6337 (2)
50.95%
50.89%
51.37%
within CBSA/MSA
Outside of a CBSA/MSA
16.65%
17.10%
13.52%
Region
Northeast
17.92%
17.15%
23.35%
Midwest
22.54%
22.79%
20.79%
107.8065 (3)
South
37.30%
37.53%
35.71%
West
22.23%
22.53%
20.16%
Urban/Rural Status
Urban
75.26%
74.94%
77.56%
42.5240 (2)
Rural
18.15%
18.60%
14.99%
[33.9319 (1)
Missing
6.59%
6.47%
7.45%
Race of Reference Person
White Only
72.12%
77.88%
31.60%
Black Only
10.93%
11.56%
6.52%
Asian Only
3.95%
4.24%
1.87%
15,004.0635 (4)
Other Race/
[31.6227 (3)
1.97%
2.10%
1.06%
Two or More Races
Blank
11.02%
4.21%
58.95%
Gender of Reference Person
Male
46.86%
50.26%
22.94%
47,930.0749 (2)
Female
46.60%
49.73%
24.56%
[4.7499 (1)
Blank
6.53%
Z%
52.50%
Hispanic Origin of Reference Person
Hispanic
11.51%
12.24%
6.39%
24,609.0812 (2)
Non-Hispanic
80.92%
87.16%
36.97%
[10.4806 (1)
Blank
7.57%
0.60%
56.64%
Age of Reference Person
15-29
10.43%
10.93%
6.84%
30-39
15.07%
16.02%
8.39%
40-49
15.88%
16.97%
8.16%
19,475.2661 (6)
50-59
18.06%
19.45%
8.21%
[263.9506 (5)
60-69
14.89%
16.33%
4.74%
70+
14.76%
16.41%
3.13%
Blank or Less than 15
10.92%
3.87%
60.52%
Overall
100%
100%
100%
Source: U.S. Census Bureau internal data from December 2014 CPS interviews.
Z: value rounds to zero but is not zero.
* May not sum to totals due to rounding.
[] The values within brackets are the chi-square statistic, df, and p-value when the blanks/missings are
excluded from the chi-square test.

A-5

P-value

< 0.0001
< 0.0001]

< 0.0001

< 0.0001

< 0.0001
< 0.0001]

< 0.0001
< 0.0001]

< 0.0001
0.0293]
< 0.0001
0.0012]

< 0.0001
< 0.0001]

Attachment A
Table A-6: Respondent and Nonrespondent Distributions for 2014 FSS Households
Household
Household
% of Total
Chi-Square
Characteristic
Respondent Nonrespondent
P-value
Sample*
Statistic (df)
%*
%*
Type of Living Quarters
Housing Unit
95.31%
95.22%
95.65%
N/A (2)
N/A
Non-Housing Unit
4.67%
4.76%
4.35%
[2.3893 (1)
0.1222]
Blank
0.02%
0.02%
0.00%
Principal City Status
Principal City within
32.42%
31.85%
34.64%
CBSA/MSA
Not Part of a Principal City
61.3230 (2)
< 0.0001
50.94%
50.51%
52.59%
within CBSA/MSA
Outside of a CBSA/MSA
16.64%
17.64%
12.76%
Region
Northeast
17.92%
17.58%
19.25%
Midwest
22.54%
23.11%
20.37%
21.2941 (3)
< 0.0001
South
37.30%
37.38%
37.01%
West
22.23%
21.94%
23.37%
Urban/Rural Status
Urban
75.20%
74.53%
77.80%
40.9023 (2)
< 0.0001
Rural
18.33%
19.11%
15.32%
[36.8120 (1)
< 0.0001]
Missing
6.47%
6.36%
6.88%
Race of Reference Person
White Only
77.72%
78.96%
72.95%
Black Only
11.64%
11.21%
13.31%
Asian Only
4.29%
4.10%
5.02%
227.2791 (4)
< 0.0001
Other Race/
[64.6297 (3)
< 0.0001]
2.09%
2.14%
1.93%
Two or More Races
Blank
4.25%
3.60%
6.79%
Gender of Reference Person
Male
50.26%
49.91%
51.60%
N/A (2)
N/A
Female
49.74%
50.09%
48.39%
[7.0552 (1)
0.0079]
Blank
Z%
0.00%
0.01%
Hispanic Origin of Reference Person
Hispanic
12.27%
12.04%
13.17%
583.4608 (2)
< 0.0001
Non-Hispanic
87.13%
87.83%
84.44%
[13.5478 (1)
0.0002]
Blank
0.60%
0.14%
2.39%
Age of Reference Person
15-29
10.95%
11.00%
10.78%
30-39
16.06%
16.17%
15.64%
40-49
16.97%
17.12%
16.40%
1,301.6616 (6)
< 0.0001
50-59
19.45%
19.65%
18.67%
[34.5533 (5) < 0.0001]
60-69
16.31%
17.04%
13.52%
70+
16.38%
16.95%
14.18%
Blank or Less than 15
3.88%
2.08%
10.82%
Overall
100%
100%
100%
Source: U.S. Census Bureau internal data from December 2014 CPS interviews.
Note: The chi-square tests for type of living quarters and gender of reference person could not be calculated
due to frequencies of 0 within the blank/nonrespondent cell and blank/respondent cell, respectively.
Z: value rounds to zero but is not zero.
* May not sum to totals due to rounding.
[] The values within brackets are the chi-square statistic, df, and p-value when the blanks/missings are
excluded from the chi-square test.

A-6

Attachment A
Table A-7: Respondent and Nonrespondent Distributions for 2014 FSS Households for Characteristics
Only Available for Responding CPS Households
Household
Household
% of Total
Chi-Square
Characteristic
Respondent Nonrespondent
P-value
Sample*
Statistic (df)
%*
%*
Tenure (Edited)
Owned or Mortgage
65.59%
66.00%
64.02%
17.3884 (2)
0.0002
Rented for Cash
33.12%
32.65%
34.93%
No Cash Rent
1.29%
1.35%
1.04%
Family Income
Less than $10,000
6.00%
6.27%
4.96%
$10,000-$19,999.99
8.82%
9.52%
6.09%
$20,000-$19,999.99
9.03%
9.95%
5.46%
$30,000-$19,999.99
9.00%
9.87%
5.64%
$40,000-$19,999.99
6.46%
6.99%
4.41%
$50,000-$19,999.99
6.05%
6.60%
3.91%
3,774.8240 (11)
< 0.0001
$60,000-$19,999.99
7.44%
8.19%
4.55%
[37.3832 (9)
< 0.0001]
$75,000-$19,999.99
8.36%
9.17%
5.23%
$100,000-$19,999.99
8.51%
9.28%
5.53%
$150,000-$19,999.99
7.19%
7.74%
5.06%
Blank or Don’t Know
3.72%
2.74%
7.54%
Refused
19.42%
13.68%
41.62%
Household Type
Husband/Wife Primary
48.91%
49.09%
48.22%
Family
Unmarried Householder
16.92%
16.67%
17.90%
Primary Family
N/A (4)
N/A
Primary Individual
34.08%
34.14%
33.84%
{10.6672 (3)
0.0137}
Group Quarters with
0.01%
0.01%
0.00%
Family
Group Quarters without
0.08%
0.09%
0.04%
Family
Child(ren) Present (Edited)
No
73.23%
73.24%
73.20%
0.0081 (1)
0.9281
Yes
26.77%
26.76%
26.80%
Measure of Labor Force of Reference Person
Employed
59.38%
59.30%
59.71%
Unemployed
2.94%
3.14%
2.14%
103.3932 (3)
< 0.0001
Not in Labor Force
36.95%
37.04%
36.61%
[20.5905 (2)
< 0.0001]
Blank
0.73%
0.53%
1.54%
Overall
100%
100%
100%
Source: U.S. Census Bureau internal data from December 2014 CPS interviews.
Note: The chi-square test for household type could not be calculated due to a frequency of 0 within the GQ
with family/nonrespondent cell.
* May not sum to totals due to rounding.
[] The values within brackets are the chi-square statistic, df, and p-value when the blanks and refusals are
excluded from the chi-square test.
{} The values within braces are the chi-square statistic, df, and p-value when the GQ groups are combined.

A-7

Attachment B
Key differences between these 2013 tables and the 2015 tables are:
•

For December 2015 CPS, there is significant difference in response rates for
Hispanic origin of reference person, excluding blanks (Table 2). This is not the case
for December 2013 CPS (Table B-2).

•

For 2015 FSS, there is significant difference in response rates for principal city status
(Table 3). This is not the case for 2013 FSS (Table B-3).

•

For 2015 FSS, there is no significant difference in response rates for gender of
reference person, excluding blanks (Table 3). This is not the case for 2013 FSS
(Table B-3).

•

For December 2015 CPS, the respondent and nonrespondent distributions (excluding
blanks/missings) are significantly different at the α=0.10 level for Hispanic origin of
reference person (Table 5). This is not the case for December 2013 CPS (Table B-5).

•

For 2015 FSS, the respondent and nonrespondent distributions are significantly
different at the α=0.10 level for principal city status (Table 6). This is not the case
for 2013 FSS (Table B-6).

•

For 2015 FSS, the respondent and nonrespondent distributions (excluding
blanks/missings) are not significantly different at the α=0.10 level for gender of
reference person (Table 6). This is not the case for 2013 FSS (Table B-6).

•

For 2015 FSS within responding CPS households, the respondent and nonrespondent
distributions (with GQ groups combined) are significantly different at the α=0.10
level for household type (Table 7). This is not the case for 2013 FSS within
responding CPS households (Table B-7).

•

For 2015 FSS within responding CPS households, the respondent and nonrespondent
distributions (excluding blanks/missings) are not significantly different at the α=0.10
level for measure of labor force of reference person (Table 7). This is not the case
for 2013 FSS within responding CPS households (Table B-7).

B-1

Attachment B
Table B-1: 2013 Food Security Unit Response Rates
Response Category

Count

Weighted
Sum 15*
130,027,288
108,047,388
96,922,728
108,047,389
85,019,318

Sampled CPS Households
72,654
Eligible CPS Households
59,759
CPS Household Response
53,410
Food Security Households
53,410
Food Security Household Response
42,147
Overall Food Security Response
Source: U.S. Census Bureau internal data from December 2013 CPS interviews.
* May not sum to totals due to rounding.

15

See footnote 3.

B-2

Response Rates
Unweighted*
Weighted*
89.38%

89.70%

78.91%
70.53%

78.69%
70.59%

Attachment B
Table B-2: Response Rates for December 2013 CPS Households
Unweighted
Weighted
Characteristic
Households
Households*

Weighted
Response
Rate (%)

Standard
Error
(%)

Significance
Grouping×

Type of Living Quarters
Housing Unit
56,948
103,015,202
89.55%
0.147%
B
2,735
4,902,128
A
93.55%
0.550%
Non-Housing Unit 16
76
130,058
C
65.08%
7.469%
Blank 17
Principal City Status
Principal City within
19,499
35,937,244
88.86%
0.243%
C
CBSA/MSA
Not Part of a Principal City
28,021
54,929,762
89.59%
0.197%
B
within CBSA/MSA
Outside of a CBSA/MSA
12,239
17,180,383
91.83%
0.404%
A
Region
Northeast
12,320
19,627,931
85.08%
0.399%
B
Midwest
14,090
24,721,632
90.45%
0.337%
A
South
19,002
39,557,362
90.89%
0.244%
A
West
14,347
24,140,464
90.76%
0.279%
A
Urban/Rural Status
Urban
46,331
86,849,144
89.29%
0.154%
B
Rural
13,428
21,198,244
91.38%
0.353%
A
Race of Reference Person
White Only
44,568
79,321,936
95.30%
0.111%
A
Black Only
5,937
12,064,561
93.95%
0.359%
B
Asian Only
2,224
4,161,550
95.06%
0.563%
A, B
Other Race/Two or More Races
1,475
2,223,255
94.38%
0.752%
A, B
Blank
5,555
10,276,086
38.32%
0.816%
C
Gender of Reference Person
Male
28,354
51,060,524
94.73%
0.148%
A
Female
28,169
51,247,854
94.73%
0.148%
A
Blank
3,236
5,739,010
0.09%
0.066%
B
Hispanic Origin of Reference Person
Hispanic
5,937
12,784,540
94.79%
0.291%
A
Non-Hispanic
49,997
88,543,453
95.15%
0.111%
A
Blank
3,825
6,719,395
8.30%
0.586%
B
Age of Reference Person
15-29
6,271
11,149,853
94.45%
0.305%
C, D
30-39
9,042
16,486,139
93.79%
0.287%
D
40-49
9,892
18,185,607
94.34%
0.241%
D
50-59
11,090
19,888,352
94.95%
0.244%
C
60-69
8,948
15,989,538
96.89%
0.199%
B
70+
8,733
15,769,520
97.81%
0.179%
A
Blank or Less than 15
5,783
10,578,380
37.55%
0.850%
E
Overall
59,759
108,047,388
89.70%
0.140%
Source: U.S. Census Bureau internal data from December 2013 CPS interviews.
* May not sum to totals due to rounding. For weighted percent of total sample, see Table B-5.
×
Within each characteristic, response rates identified with the same letter are not significantly different at the
α=0.10 level. A indicates the highest response rates, B indicates the next highest rates, etc.

16

17

See footnote 4.
See footnote 5.

B-3

Attachment B
Table B-3: Response Rates for 2013 FSS Households
Unweighted
Weighted
Characteristic
Households
Households*

Weighted
Response
Rate (%)

Standard
Error (%)

Significance
Grouping×

Type of Living Quarters
Housing Unit
50,820
102,923,127
78.66%
0.240%
B
2,537
5,028,834
B
79.15%
1.027%
Non-Housing Unit 18
53
95,428
A
87.33%
4.916%
Blank 19
Principal City Status
Principal City within
17,250
35,930,575
78.51%
0.430%
A
CBSA/MSA
Not Part of a Principal City
25,021
54,901,940
78.59%
0.313%
A
within CBSA/MSA
Outside of a CBSA/MSA
11,139
17,214,874
79.37%
0.915%
A
Region
Northeast
10,441
19,627,931
77.03%
0.541%
B
Midwest
12,775
24,721,632
80.30%
0.486%
A
South
17,219
39,557,362
79.28%
0.432%
A
West
12,975
24,140,464
77.41%
0.515%
B
Urban/Rural Status
Urban
41,230
86,711,148
78.25%
0.258%
B
Rural
12,180
21,336,241
80.46%
0.586%
A
Race of Reference Person
White Only
42,370
84,152,559
79.80%
0.253%
A
Black Only
5,557
12,671,887
76.35%
0.686%
B
Asian Only
2,113
4,452,831
75.86%
1.080%
B
Other Race/Two or More Races
1,392
2,331,908
79.78%
1.482%
A
Blank
1,978
4,438,203
66.57%
1.294%
C
Gender of Reference Person
Male
26,785
53,911,069
78.13%
0.293%
B
Female
26,623
54,130,203
79.25%
0.321%
A
Blank
2
6,116
0.00%
0.000%
N+
Hispanic Origin of Reference Person
Hispanic
5,620
13,517,612
77.08%
0.673%
B
Non-Hispanic
47,451
93,904,476
79.36%
0.242%
A
Blank
339
625,301
12.52%
2.071%
C
Age of Reference Person
15-29
5,900
11,752,732
80.02%
0.602%
B
30-39
8,448
17,244,286
80.08%
0.493%
B
40-49
9,304
19,143,468
80.16%
0.516%
B
50-59
10,515
21,058,840
79.91%
0.468%
B
60-69
8,659
17,250,667
81.86%
0.510%
A
70+
8,537
17,177,697
80.43%
0.534%
B
Blank or Less than 15
2,047
4,419,699
38.37%
1.214%
C
Overall
53,410
108,047,389
78.69%
0.236%
Source: U.S. Census Bureau internal data from December 2013 CPS interviews.
* May not sum to totals due to rounding. For weighted percent of total sample, see Table B-6.
N indicates that the estimate is not comparable.
+
Exercise caution: The sample size is extremely small, leading to unreliable estimates.
×
Within each characteristic, response rates identified with the same letter are not significantly different at the
α=0.10 level. A indicates the highest response rates, B indicates the next highest rates, etc.

18

19

See footnote 4.
See footnote 5.

B-4

Attachment B
Table B-4: Response Rates for 2013 FSS Households for Characteristics Only Available for Responding
CPS Households
Unweighted
Weighted
Weighted
Standard Significance
Characteristic
Households
Households*
Response Error (%)
Grouping×
Rate (%)
Tenure (Edited)
Owned or Mortgage
35,856
71,726,224
79.26%
0.286%
A
Rented for Cash
16,860
34,997,315
77.56%
0.417%
B
No Cash Rent
694
1,323,850
77.44%
1.817%
A, B
Family Income
Less than $10,000
3,081
6,403,275
81.54%
0.847%
D
$10,000-$19,999.99
4,754
9,879,454
85.41%
0.602%
C
$20,000-$29,999.99
4,909
9,909,735
85.52%
0.518%
B, C
$30,000-$39,999.99
4,670
9,439,860
86.72%
0.546%
A, C
$40,000-$49,999.99
3,419
6,773,123
87.43%
0.626%
A
$50,000-$59,999.99
3,353
6,612,555
86.83%
0.647%
A, B
$60,000-$74,999.99
4,004
7,870,043
86.91%
0.596%
A
$75,000-$99,999.99
4,530
8,877,888
87.35%
0.614%
A
$100,000-$149,999.99
4,312
8,553,757
87.58%
0.613%
A
$150,000+
3,366
6,887,927
85.16%
0.740%
C
Blank or Don’t Know
2,046
4,273,537
57.76%
1.260%
E
Refused
10,966
22,566,234
55.90%
0.627%
E
Household Type
Husband/Wife Primary Family
26,169
52,691,131
79.16%
0.294%
A
Unmarried Householder
8,824
18,600,115
77.77%
0.557%
B
Primary Family
Primary Individual
18,367
36,644,017
78.50%
0.413%
A, B
Group Quarters with Family
21
57,069
74.18%
8.999%
A, B+
Group Quarters without Family
29
55,057
65.06%
16.544%
A, B+
Child(ren) Present (Edited)
No
39,444
79,122,293
78.35%
0.270%
B
Yes
13,966
28,925,096
79.61%
0.413%
A
Measure of Labor Force of Reference Person
Employed
31,910
63,722,088
79.08%
0.285%
B
Unemployed
1,817
3,837,758
82.76%
0.986%
A
Not in Labor Force
19,266
39,661,117
78.14%
0.358%
C
Blank
417
826,425
55.58%
2.518%
D
Overall
53,410
108,047,389
78.69%
0.236%
Source: U.S. Census Bureau internal data from December 2013 CPS interviews.
* May not sum to totals due to rounding. For weighted percent of total sample, see Table B-7.
+
Exercise caution: The sample size is small, leading to unreliable estimates.
×
Within each characteristic, response rates identified with the same letter are not significantly different at the
α=0.10 level. A indicates the highest response rates, B indicates the next highest rates, etc.

B-5

Attachment B

Table B-5: Respondent and Nonrespondent Distributions for December 2013 CPS Households
Household
Household
% of Total
Chi-Square
Characteristic
Respondent Nonrespondent
P-value
Sample*
Statistic (df)
%*
%*
Type of Living Quarters
Housing Unit
95.34%
95.18%
96.75%
59.0288 (2)
< 0.0001
Non-Housing Unit
4.54%
4.73%
2.84%
[31.7686 (1) < 0.0001]
Blank
0.12%
0.09%
0.41%
Principal City Status
Principal City within
33.26%
32.95%
35.97%
CBSA/MSA
Not Part of a Principal City
38.0232 (2)
< 0.0001
50.84%
50.77%
51.40%
within CBSA/MSA
Outside of a CBSA/MSA
15.90%
16.28%
12.62%
Region
Northeast
18.17%
17.23%
26.32%
Midwest
22.88%
23.07%
21.23%
197.8466 (3)
< 0.0001
South
36.61%
37.10%
32.39%
West
22.34%
22.60%
20.06%
Urban/Rural Status
Urban
80.38%
80.01%
83.58%
25.1508 (1)
< 0.0001
Rural
19.62%
19.99%
16.42%
Race of Reference Person
White Only
73.41%
78.00%
33.49%
Black Only
11.17%
11.69%
6.56%
Asian Only
3.85%
4.08%
1.85%
12,197.8081 (4)
< 0.0001
Other Race/
[16.7158 (3)
0.0008]
2.06%
2.16%
1.12%
Two or More Races
Blank
9.51%
4.06%
56.98%
Gender of Reference Person
Male
47.26%
49.91%
24.17%
21,706.9115 (2)
< 0.0001
Female
47.43%
50.09%
24.29%
[0.0008 (1)
0.9775]
Blank
5.31%
0.01%
51.54%
Hispanic Origin of Reference Person
Hispanic
11.83%
12.50%
5.99%
19,548.5788 (2)
< 0.0001
Non-Hispanic
81.95%
86.92%
38.62%
[1.3533 (1)
0.2447]
Blank
6.22%
0.58%
55.39%
Age of Reference Person
15-29
10.32%
10.86%
5.57%
30-39
15.26%
15.95%
9.21%
40-49
16.83%
17.70%
9.25%
14,714.3966 (6)
< 0.0001
50-59
18.41%
19.48%
9.02%
[213.8015 (5) < 0.0001]
60-69
14.80%
15.98%
4.47%
70+
14.60%
15.91%
3.10%
Blank or Less than 15
9.79%
4.10%
59.38%
Overall
100%
100%
100%
Source: U.S. Census Bureau internal data from December 2013 CPS interviews.
* May not sum to totals due to rounding.
[] The values within brackets are the chi-square statistic, df, and p-value when the blanks are excluded from
the chi-square test.

B-6

Attachment B
Table B-6: Respondent and Nonrespondent Distributions for 2013 FSS Households
Household
Household
% of Total
Chi-Square
Characteristic
Respondent Nonrespondent
P-value
Sample*
Statistic (df)
%*
%*
Type of Living Quarters
Housing Unit
95.26%
95.22%
95.40%
1.9102 (2)
0.3848
Non-Housing Unit
4.65%
4.68%
4.55%
[0.2265 (1)
0.6341]
Blank
0.09%
0.10%
0.05%
Principal City Status
Principal City within
33.25%
33.18%
33.53%
CBSA/MSA
Not Part of a Principal City
0.9431 (2)
0.6240
50.81%
50.75%
51.05%
within CBSA/MSA
Outside of a CBSA/MSA
15.93%
16.07%
15.42%
Region
Northeast
18.17%
17.78%
19.58%
Midwest
22.88%
23.35%
21.15%
26.6025 (3)
< 0.0001
South
36.61%
36.89%
35.60%
West
22.34%
21.98%
23.68%
Urban/Rural Status
Urban
80.25%
79.81%
81.89%
11.2207 (1)
0.0008
Rural
19.75%
20.19%
18.11%
Race of Reference Person
White Only
77.88%
78.98%
73.83%
Black Only
11.73%
11.38%
13.01%
Asian Only
4.12%
3.97%
4.67%
165.6895 (4)
< 0.0001
Other Race/
[36.9013 (3)
< 0.0001]
2.16%
2.19%
2.05%
Two or More Races
Blank
4.11%
3.48%
6.44%
Gender of Reference Person
Male
49.90%
49.54%
51.21%
N/A (2)
N/A
Female
50.10%
50.46%
48.77%
[8.1684 (1)
0.0043]
Blank
0.01%
0.00%
0.03%
Hispanic Origin of Reference Person
Hispanic
12.51%
12.26%
13.45%
592.8209 (2)
< 0.0001
Non-Hispanic
86.91%
87.65%
84.17%
[10.9915 (1)
0.0009]
Blank
0.58%
0.09%
2.38%
Age of Reference Person
15-29
10.88%
11.06%
10.20%
30-39
15.96%
16.24%
14.91%
40-49
17.72%
18.05%
16.49%
1,620.9468 (6)
< 0.0001
50-59
19.49%
19.79%
18.38%
[10.7216 (5)
0.0572]
60-69
15.97%
16.61%
13.59%
70+
15.90%
16.25%
14.60%
Blank or Less than 15
4.09%
1.99%
11.83%
Overall
100%
100%
100%
Source: U.S. Census Bureau internal data from December 2013 CPS interviews.
Note: The chi-square test for gender of reference person could not be calculated due to a frequency of 0 within
the blank/respondent cell.
* May not sum to totals due to rounding.
[] The values within brackets are the chi-square statistic, df, and p-value when the blanks are excluded from
the chi-square test.

B-7

Attachment B
Table B-7: Respondent and Nonrespondent Distributions for 2013 FSS Households for Characteristics
Only Available for Responding CPS Households
Household
Household
% of Total
Chi-Square
Characteristic
Respondent Nonrespondent
P-value
Sample*
Statistic (df)
%*
%*
Tenure (Edited)
Owned or Mortgage
66.38%
66.87%
64.60%
14.0598 (2)
0.0009
Rented for Cash
32.39%
31.93%
34.11%
No Cash Rent
1.23%
1.21%
1.30%
Family Income
Less than $10,000
5.93%
6.14%
5.13%
$10,000-$19,999.99
9.14%
9.92%
6.26%
$20,000-$19,999.99
9.17%
9.97%
6.23%
$30,000-$19,999.99
8.74%
9.63%
5.44%
$40,000-$19,999.99
6.27%
6.96%
3.70%
$50,000-$19,999.99
6.12%
6.75%
3.78%
4,056.4412 (11)
< 0.0001
$60,000-$19,999.99
7.28%
8.05%
4.47%
[65.9528 (9)
< 0.0001]
$75,000-$19,999.99
8.22%
9.12%
4.88%
$100,000-$19,999.99
7.92%
8.81%
4.61%
$150,000-$19,999.99
6.37%
6.90%
4.44%
Blank or Don’t Know
3.96%
2.90%
7.84%
Refused
20.89%
14.84%
43.21%
Household Type
Husband/Wife Primary
48.77%
49.06%
47.69%
Family
Unmarried Householder
17.21%
17.01%
17.96%
Primary Family
6.4872 (4)
0.1656
Primary Individual
33.91%
33.84%
34.21%
{5.3871 (3)
0.1456}
Group Quarters with
0.05%
0.05%
0.06%
Family
Group Quarters without
0.05%
0.04%
0.08%
Family
Child(ren) Present (Edited)
No
73.23%
72.92%
74.39%
6.9965 (1)
0.0082
Yes
26.77%
27.08%
25.61%
Measure of Labor Force of Reference Person
Employed
58.98%
59.27%
57.89%
129.6623 (3)
< 0.0001
Unemployed
3.55%
3.74%
2.87%
[18.6827 (2) < 0.0001]
Not in Labor Force
36.71%
36.45%
37.64%
Blank
0.76%
0.54%
1.59%
Overall
100%
100%
100%
Source: U.S. Census Bureau internal data from December 2013 CPS interviews.
* May not sum to totals due to rounding.
[] The values within brackets are the chi-square statistic, df, and p-value when the blanks and refusals are
excluded from the chi-square test.
{} The values within braces are the chi-square statistic, df, and p-value when the GQ groups are combined.

B-8

Attachment C
Key differences between these 2012 tables and the 2015 tables are:
•

For 2015 FSS, the respondent and nonrespondent distributions (excluding
blanks/missings) are not significantly different at the α=0.10 level for gender of
reference person (Table 6). This is not the case for 2012 FSS (Table C-6).

•

For 2015 FSS within responding CPS households, the respondent and nonrespondent
distributions are significantly different at the α=0.10 level for tenure (Table 7). This
is not the case for 2012 FSS within responding CPS households (Table C-7).

•

For 2015 FSS within responding CPS households, the respondent and nonrespondent
distributions (excluding blanks/missings) are not significantly different at the α=0.10
level for measure of labor force of reference person (Table 7). This is not the case
for 2012 FSS within responding CPS households (Table C-7).

Table C-1: 2012 Food Security Unit Response Rates
Response Category

Count

Weighted
Sum 20*
129,411,414
107,888,749
97,144,887
107,888,749
87,847,406

Sampled CPS Households
72,263
Eligible CPS Households
59,759
CPS Household Response
53,704
Food Security Households
53,704
Food Security Household Response
43,942
Overall Food Security Response
Source: U.S. Census Bureau internal data from December 2012 CPS interviews.
* May not sum to totals due to rounding.

20

See footnote 3.

C-1

Response Rates
Unweighted*
Weighted*
89.87%

90.04%

81.82%
73.53%

81.42%
73.32%

Attachment C
Table C-2: Response Rates for December 2012 CPS Households
Unweighted
Weighted
Characteristic
Households
Households*

Weighted
Response
Rate (%)

Standard
Error
(%)

Significance
Grouping×

Type of Living Quarters
Housing Unit
56,824
102,676,729
89.98%
0.143%
B
2,847
5,059,503
A
92.39%
0.704%
Non-Housing Unit 21
88
152,517
C
55.32%
7.190%
Blank 22
Principal City Status
Principal City within
19,399
35,573,619
88.81%
0.256%
C
CBSA/MSA
Not Part of a Principal City
27,902
54,735,677
90.14%
0.187%
B
within CBSA/MSA
Outside of a CBSA/MSA
12,458
17,579,453
92.22%
0.353%
A
Region
Northeast
12,359
19,624,243
87.27%
0.390%
C
Midwest
14,091
24,715,110
90.96%
0.320%
A
South
18,990
39,526,802
90.76%
0.213%
A, B
West
14,319
24,022,593
90.19%
0.295%
B
Urban/Rural Status
Urban
46,090
86,180,702
89.62%
0.155%
B
Rural
13,669
21,708,046
91.73%
0.291%
A
Race of Reference Person
White Only
44,779
79,537,863
95.35%
0.110%
A
Black Only
5,932
12,069,571
93.38%
0.333%
C
Asian Only
2,179
4,034,039
94.02%
0.539%
B, C
Other Race/Two or More Races
1,487
2,176,711
95.28%
0.702%
A, B
Blank
5,382
10,070,565
41.36%
0.778%
D
Gender of Reference Person
Male
28,632
51,446,267
94.71%
0.145%
A
Female
28,188
51,198,833
94.57%
0.147%
A
Blank
2,939
5,243,648
0.05%
0.051%
B
Hispanic Origin of Reference Person
Hispanic
5,856
12,529,458
94.25%
0.363%
B
Non-Hispanic
50,457
89,246,579
95.06%
0.106%
A
Blank
3,446
6,112,712
8.14%
0.573%
C
Age of Reference Person
15-29
6,625
11,874,827
93.61%
0.324%
D
30-39
9,144
16,501,896
93.72%
0.298%
D
40-49
10,278
18,782,626
94.25%
0.256%
D
50-59
11,165
20,032,892
94.98%
0.200%
C
60-69
8,911
15,882,394
96.60%
0.217%
B
70+
8,587
15,513,739
97.74%
0.181%
A
Blank or Less than 15
5,049
9,300,374
35.78%
0.709%
E
Overall
59,759
107,888,749
90.04%
0.143%
Source: U.S. Census Bureau internal data from December 2012 CPS interviews.
* May not sum to totals due to rounding. For weighted percent of total sample, see Table C-5.
×
Within each characteristic, response rates identified with the same letter are not significantly different at the
α=0.10 level. A indicates the highest response rates, B indicates the next highest rates, etc.

21

22

See footnote 4.
See footnote 5.

C-2

Attachment C
Table C-3: Response Rates for 2012 FSS Households
Unweighted
Weighted
Characteristic
Households
Households*

Weighted
Response
Rate (%)

Standard
Error (%)

Significance
Grouping×

Type of Living Quarters
Housing Unit
51,043
102,684,653
81.36%
0.224%
B
2,607
5,110,086
B
82.47%
0.953%
Non-Housing Unit 23
54
94,010
A
92.25%
4.110%
Blank 24
Principal City Status
Principal City within
17,220
35,572,515
81.30%
0.380%
B
CBSA/MSA
Not Part of a Principal City
25,078
54,705,150
80.95%
0.336%
B
within CBSA/MSA
Outside of a CBSA/MSA
11,406
17,611,084
83.16%
0.656%
A
Region
Northeast
10,734
19,624,243
79.12%
0.560%
C
Midwest
12,870
24,715,110
83.72%
0.440%
A
South
17,160
39,526,802
81.52%
0.422%
B
West
12,940
24,022,593
80.78%
0.469%
B
Urban/Rural Status
Urban
41,228
86,079,702
81.11%
0.240%
B
Rural
12,476
21,809,046
82.66%
0.564%
A
Race of Reference Person
White Only
42,614
84,012,912
82.60%
0.264%
A
Black Only
5,527
12,615,940
77.83%
0.689%
C
Asian Only
2,053
4,272,517
78.93%
1.089%
B, C
Other Race/Two or More Races
1,415
2,295,419
81.73%
1.371%
A, B
Blank
2,095
4,691,962
72.17%
1.110%
D
Gender of Reference Person
Male
27,076
54,076,962
81.09%
0.277%
A
Female
26,627
53,808,880
81.76%
0.329%
A
Blank
1
2,907
0.00%
0.000%
N+
Hispanic Origin of Reference Person
Hispanic
5,515
13,215,624
80.73%
0.596%
B
Non-Hispanic
47,893
94,121,104
81.91%
0.246%
A
Blank
296
552,021
14.37%
2.416%
C
Age of Reference Person
15-29
6,205
12,382,387
82.99%
0.546%
A, C
30-39
8,576
17,204,571
83.54%
0.480%
A
40-49
9,660
19,672,809
81.86%
0.422%
C
50-59
10,589
21,122,044
82.44%
0.427%
B, C
60-69
8,580
17,015,377
83.72%
0.435%
A
70+
8,387
16,799,896
82.99%
0.486%
A, B
Blank or Less than 15
1,707
3,691,665
40.43%
1.422%
D
Overall
53,704
107,888,749
81.42%
0.227%
Source: U.S. Census Bureau internal data from December 2012 CPS interviews.
* May not sum to totals due to rounding. For weighted percent of total sample, see Table C-6.
N indicates that the estimate is not comparable.
+
Exercise caution: The sample size is extremely small, leading to unreliable estimates.
×
Within each characteristic, response rates identified with the same letter are not significantly different at the
α=0.10 level. A indicates the highest response rates, B indicates the next highest rates, etc.

23

24

See footnote 4.
See footnote 5.

C-3

Attachment C
Table C-4: Response Rates for 2012 FSS Households for Characteristics Only Available for Responding
CPS Households
Unweighted
Weighted
Weighted
Standard Significance
Characteristic
Households
Households*
Response Error (%)
Grouping×
Rate (%)
Tenure (Edited)
Owned or Mortgage
36,042
71,585,561
81.73%
0.293%
A
Rented for Cash
16,936
34,861,832
80.82%
0.362%
B
No Cash Rent
726
1,441,356
80.87%
1.785%
A, B
Family Income
Less than $10,000
3,401
7,005,328
85.90%
0.714%
E
$10,000-$19,999.99
5,032
10,285,677
87.57%
0.549%
C, D
$20,000-$29,999.99
5,191
10,519,664
88.12%
0.555%
A, C
$30,000-$39,999.99
4,904
9,811,107
87.89%
0.506%
B, C
$40,000-$49,999.99
3,613
7,222,021
88.67%
0.688%
A, C
$50,000-$59,999.99
3,529
6,891,757
89.47%
0.635%
A
$60,000-$74,999.99
4,233
8,186,706
88.13%
0.609%
A, C
$75,000-$99,999.99
4,678
9,236,390
88.38%
0.553%
A, C
$100,000-$149,999.99
4,339
8,574,543
88.89%
0.574%
A, B
$150,000+
3,258
6,632,657
86.33%
0.673%
D, E
Blank or Don’t Know
1,896
3,886,426
63.39%
1.453%
F
Refused
9,630
19,636,474
56.88%
0.636%
G
Household Type
Husband/Wife Primary Family
26,414
52,827,843
81.97%
0.300%
B
Unmarried Householder
8,940
18,710,955
81.17%
0.490%
B, C
Primary Family
Primary Individual
18,303
36,254,769
80.75%
0.377%
C
Group Quarters with Family
14
27,840
97.16%
2.276%
A+
Group Quarters without Family
33
67,342
83.16%
8.902%
A, B, C
Child(ren) Present (Edited)
No
39,389
78,547,270
81.01%
0.263%
B
Yes
14,315
29,341,478
82.54%
0.370%
A
Measure of Labor Force of Reference Person
Employed
32,304
64,067,991
81.79%
0.246%
B
Unemployed
2,171
4,567,889
85.45%
0.792%
A
Not in Labor Force
18,842
38,505,975
80.76%
0.383%
C
Blank
387
746,894
59.55%
2.734%
D
Overall
53,704
107,888,749
81.42%
0.227%
Source: U.S. Census Bureau internal data from December 2012 CPS interviews.
* May not sum to totals due to rounding. For weighted percent of total sample, see Table C-7.
+
Exercise caution: The sample size is small, leading to unreliable estimates.
×
Within each characteristic, response rates identified with the same letter are not significantly different at the
α=0.10 level. A indicates the highest response rates, B indicates the next highest rates, etc.

C-4

Attachment C

Table C-5: Respondent and Nonrespondent Distributions for December 2012 CPS Households
Household
Household
% of Total
Chi-Square
Characteristic
Respondent Nonrespondent
P-value
Sample*
Statistic (df)
%*
%*
Type of Living Quarters
Housing Unit
95.17%
95.10%
95.78%
72.6346 (2)
< 0.0001
Non-Housing Unit
4.69%
4.81%
3.58%
0.0026]
[9.0544 (1)
Blank
0.14%
0.09%
0.63%
Principal City Status
Principal City within
32.97%
32.52%
37.04%
CBSA/MSA
Not Part of a Principal City
60.2820 (2)
< 0.0001
50.73%
50.79%
50.22%
within CBSA/MSA
Outside of a CBSA/MSA
16.29%
16.69%
12.74%
Region
Northeast
18.19%
17.63%
23.25%
Midwest
22.91%
23.14%
20.80%
82.5535 (3)
< 0.0001
South
36.64%
36.93%
34.01%
West
22.27%
22.30%
21.94%
Urban/Rural Status
Urban
79.88%
79.50%
83.29%
39.5566 (1)
< 0.0001
Rural
20.12%
20.50%
16.71%
Race of Reference Person
White Only
73.72%
78.07%
34.39%
Black Only
11.19%
11.60%
7.44%
Asian Only
3.74%
3.90%
2.25%
13,308.6592 (4)
< 0.0001
Other Race/
[43.8268 (3)
< 0.0001]
2.02%
2.13%
0.96%
Two or More Races
Blank
9.33%
4.29%
54.97%
Gender of Reference Person
Male
47.68%
50.16%
25.33%
22,179.3853 (2)
< 0.0001
Female
47.46%
49.84%
25.89%
[0.4829 (1)
0.4871]
Blank
4.86%
Z%
48.78%
Hispanic Origin of Reference Person
Hispanic
11.61%
12.16%
6.70%
17,909.1271 (2)
< 0.0001
Non-Hispanic
82.72%
87.33%
41.03%
[4.7530 (1)
0.0292]
Blank
5.67%
0.51%
52.26%
Age of Reference Person
15-29
11.01%
11.44%
7.07%
30-39
15.30%
15.92%
9.65%
40-49
17.41%
18.22%
10.05%
15,908.9147 (6)
< 0.0001
50-59
18.57%
19.59%
9.35%
[210.4975 (5) < 0.0001]
60-69
14.72%
15.79%
5.02%
70+
14.38%
15.61%
3.26%
Blank or Less than 15
8.62%
3.43%
55.59%
Overall
100%
100%
100%
Source: U.S. Census Bureau internal data from December 2012 CPS interviews.
Z: value rounds to zero but is not zero.
* May not sum to totals due to rounding.
[] The values within brackets are the chi-square statistic, df, and p-value when the blanks are excluded from
the chi-square test.

C-5

Attachment C
Table C-6: Respondent and Nonrespondent Distributions for 2012 FSS Households
Household
Household
% of Total
Chi-Square
Characteristic
Respondent Nonrespondent
P-value
Sample*
Statistic (df)
%*
%*
Type of Living Quarters
Housing Unit
95.18%
95.10%
95.49%
4.4383 (2)
0.1087
Non-Housing Unit
4.74%
4.80%
4.47%
[1.3521 (1)
0.2449]
Blank
0.09%
0.10%
0.04%
Principal City Status
Principal City within
32.97%
32.92%
33.19%
CBSA/MSA
Not Part of a Principal City
9.5214 (2)
0.0086
50.71%
50.41%
52.01%
within CBSA/MSA
Outside of a CBSA/MSA
16.32%
16.67%
14.80%
Region
Northeast
18.19%
17.67%
20.44%
Midwest
22.91%
23.55%
20.08%
41.6774 (3)
< 0.0001
South
36.64%
36.68%
36.45%
West
22.27%
22.09%
23.03%
Urban/Rural Status
Urban
79.79%
79.48%
81.13%
6.3112 (1)
0.0120
Rural
20.21%
20.52%
18.87%
Race of Reference Person
White Only
77.87%
78.99%
72.94%
Black Only
11.69%
11.18%
13.95%
Asian Only
3.96%
3.84%
4.49%
147.7389 (4)
< 0.0001
Other Race/
[61.0095 (3)
< 0.0001]
2.13%
2.14%
2.09%
Two or More Races
Blank
4.35%
3.85%
6.52%
Gender of Reference Person
Male
50.12%
49.92%
51.02%
N/A (2)
N/A
Female
49.87%
50.08%
48.96%
[2.7654 (1)
0.0963 ]
Blank
Z%
0.00%
0.01%
Hispanic Origin of Reference Person
Hispanic
12.25%
12.14%
12.71%
569.3827 (2)
< 0.0001
Non-Hispanic
87.24%
87.76%
84.93%
[3.4271 (1)
0.0641]
Blank
0.51%
0.09%
2.36%
Age of Reference Person
15-29
11.48%
11.70%
10.51%
30-39
15.95%
16.36%
14.13%
40-49
18.23%
18.33%
17.80%
1,697.2412 (6)
< 0.0001
50-59
19.58%
19.82%
18.50%
[13.2286 (5)
0.0213]
60-69
15.77%
16.22%
13.83%
70+
15.57%
15.87%
14.26%
Blank or Less than 15
3.42%
1.70%
10.97%
Overall
100%
100%
100%
Source: U.S. Census Bureau internal data from December 2012 CPS interviews.
Note: The chi-square test for gender of reference person could not be calculated due to a frequency of 0 within
the blank/respondent cell.
Z: value rounds to zero but is not zero.
* May not sum to totals due to rounding.
[] The values within brackets are the chi-square statistic, df, and p-value when the blanks are excluded from
the chi-square test.

C-6

Attachment C
Table C-7: Respondent and Nonrespondent Distributions for 2012 FSS Households for Characteristics
Only Available for Responding CPS Households
Household
Household
% of Total
Chi-Square
Characteristic
Respondent Nonrespondent
P-value
Sample*
Statistic (df)
%*
%*
Tenure (Edited)
Owned or Mortgage
66.35%
66.60%
65.26%
4.1217 (2)
0.1273
Rented for Cash
32.31%
32.07%
33.36%
No Cash Rent
1.34%
1.33%
1.38%
Family Income
Less than $10,000
6.49%
6.85%
4.93%
$10,000-$19,999.99
9.53%
10.25%
6.38%
$20,000-$19,999.99
9.75%
10.55%
6.24%
$30,000-$19,999.99
9.09%
9.82%
5.93%
$40,000-$19,999.99
6.69%
7.29%
4.08%
$50,000-$19,999.99
6.39%
7.02%
3.62%
3,848.2684 (11)
< 0.0001
$60,000-$19,999.99
7.59%
8.21%
4.85%
[26.2447 (9)
0.0019]
$75,000-$19,999.99
8.56%
9.29%
5.36%
$100,000-$19,999.99
7.95%
8.68%
4.75%
$150,000-$19,999.99
6.15%
6.52%
4.52%
Blank or Don’t Know
3.60%
2.80%
7.10%
Refused
18.20%
12.72%
42.24%
Household Type
Husband/Wife Primary
48.97%
49.29%
47.53%
Family
Unmarried Householder
17.34%
17.29%
17.58%
Primary Family
10.5261 (4)
0.0324
Primary Individual
33.60%
33.33%
34.82%
{7.8583 (3)
0.0490}
Group Quarters with
0.03%
0.03%
Z%
Family
Group Quarters without
0.06%
0.06%
0.06%
Family
Child(ren) Present (Edited)
No
72.80%
72.43%
74.43%
12.4996 (1)
0.0004
Yes
27.20%
27.57%
25.57%
Measure of Labor Force of Reference Person
Employed
59.38%
59.65%
58.22%
Unemployed
4.23%
4.44%
3.32%
126.6214 (3)
< 0.0001
Not in Labor Force
35.69%
35.40%
36.96%
[26.7450
< 0.0001]
Blank
0.69%
0.51%
1.51%
Overall
100%
100%
100%
Source: U.S. Census Bureau internal data from December 2012 CPS interviews.
Z: value rounds to zero but is not zero.
* May not sum to totals due to rounding.
[] The values within brackets are the chi-square statistic, df, and p-value when the blanks and refusals are
excluded from the chi-square test.
{} The values within braces are the chi-square statistic, df, and p-value when the GQ groups are combined.

C-7

Attachment D
Key differences between these 2011 tables and the 2015 tables are:
•

For 2015 FSS, the respondent and nonrespondent distributions are significantly
different at the α=0.10 level for principal city status (Table 6). This is not the case
for 2011 FSS (Table D-6).

•

For 2015 FSS within responding CPS households, the respondent and nonrespondent
distributions (excluding blanks/missings) are not significantly different at the α=0.10
level for measure of labor force of reference person (Table 7). This is not the case
for 2011 FSS within responding CPS households (Table D-7).

Table D-1: 2011 Food Security Unit Response Rates
Response Category

Count

Weighted
Sum 25*
128,975,102
106,970,281
96,515,523
106,970,281
87,308,032

Sampled CPS Households
71,982
Eligible CPS Households
59,207
CPS Household Response
53,446
Food Security Households
53,446
Food Security Household Response
43,770
Overall Food Security Response
Source: U.S. Census Bureau internal data from December 2011 CPS interviews.
* May not sum to totals due to rounding.

25

See footnote 3.

D-1

Response Rates
Unweighted*
Weighted*
90.27%

90.23%

81.90%
73.93%

81.62%
73.64%

Attachment D
Table D-2: Response Rates for December 2011 CPS Households
Unweighted
Weighted
Characteristic
Households
Households*

Weighted
Response
Rate (%)

Standard
Error
(%)

Significance
Grouping×

Type of Living Quarters
Housing Unit
56,307
101,770,940
90.09%
0.136%
B
Non-Housing Unit 26
2,836
5,099,943
93.23%
0.611%
A
Blank 27
64
99,397
74.76%
7.910%
C
Principal City Status
Principal City within
19,184
35,321,616
89.00%
0.246%
C
CBSA/MSA
Not Part of a Principal City
27,620
54,210,509
90.24%
0.195%
B
within CBSA/MSA
Outside of a CBSA/MSA
12,403
17,438,156
92.66%
0.353%
A
Region
Northeast
12,278
19,591,878
88.31%
0.341%
C
Midwest
13,976
24,486,003
91.71%
0.274%
A
South
18,669
38,980,091
90.42%
0.263%
B
West
14,284
23,912,309
89.96%
0.294%
B
Urban/Rural Status
Urban
45,621
85,419,020
89.73%
0.154%
B
Rural
13,586
21,551,261
92.19%
0.272%
A
Race of Reference Person
White Only
44,398
78,753,944
95.36%
0.118%
A
Black Only
5,919
12,040,880
93.58%
0.311%
C
Asian Only
2,149
4,029,535
93.90%
0.602%
B, C
Other Race/Two or More Races
1,434
2,042,774
95.04%
0.750%
A, B
Blank
5,307
10,103,149
43.78%
0.892%
D
Gender of Reference Person
Male
28,386
51,174,793
94.72%
0.162%
A
Female
28,035
50,691,742
94.77%
0.142%
A
Blank
2,786
5,103,746
0.09%
0.057%
B
Hispanic Origin of Reference Person
Hispanic
5,552
11,927,076
94.27%
0.333%
B
Non-Hispanic
50,387
89,101,541
95.16%
0.114%
A
Blank
3,268
5,941,664
8.19%
0.571%
C
Age of Reference Person
15-29
6,562
11,664,835
93.59%
0.358%
E
30-39
9,156
16,632,191
94.50%
0.244%
D
40-49
10,435
19,003,939
93.74%
0.267%
E
50-59
11,113
19,907,729
95.77%
0.217%
C
60-69
8,635
15,419,603
96.38%
0.231%
B
70+
8,530
15,400,566
97.57%
0.209%
A
Blank or Less than 15
4,776
8,941,419
34.82%
0.812%
F
Overall
59,207
106,970,281
90.23%
0.136%
Source: U.S. Census Bureau internal data from December 2011 CPS interviews.
* May not sum to totals due to rounding. For weighted percent of total sample, see Table D-5.
×
Within each characteristic, response rates identified with the same letter are not significantly different at the
α=0.10 level. A indicates the highest response rates, B indicates the next highest rates, etc.

26
27

See footnote 4.
See footnote 5.

D-2

Attachment D
Table D-3: Response Rates for 2011 FSS Households
Unweighted
Weighted
Characteristic
Households
Households*

Weighted
Response
Rate (%)

Standard
Error (%)

Significance
Grouping×

Type of Living Quarters
Housing Unit
50,761
101,679,154
81.57%
0.215%
A
Non-Housing Unit 28
2,635
5,208,328
82.50%
0.930%
A
Blank 29
50
82,799
80.43%
7.408%
A
Principal City Status
Principal City within
17,070
35,318,528
80.82%
0.385%
B
CBSA/MSA
Not Part of a Principal City
24,935
54,184,961
81.89%
0.285%
A
within CBSA/MSA
Outside of a CBSA/MSA
11,441
17,466,792
82.40%
0.810%
A
Region
Northeast
10,858
19,591,878
79.08%
0.546%
C
Midwest
12,871
24,486,003
84.02%
0.479%
A
South
16,854
38,980,091
81.65%
0.400%
B
West
12,863
23,912,309
81.19%
0.475%
B
Urban/Rural Status
Urban
40,955
85,266,834
81.24%
0.212%
B
Rural
12,491
21,703,447
83.11%
0.533%
A
Race of Reference Person
White Only
42,336
82,998,917
82.96%
0.246%
A
Black Only
5,524
12,598,316
77.70%
0.685%
C
Asian Only
2,020
4,253,758
78.54%
0.988%
B, C
Other Race/Two or More Races
1,362
2,158,759
81.28%
1.385%
A, B
Blank
2,204
4,960,530
71.84%
1.156%
D
Gender of Reference Person
Male
26,869
53,670,453
81.65%
0.277%
A
Female
26,574
53,294,690
81.60%
0.282%
A
Blank
3
5,137
0.00%
0.000%
N+
Hispanic Origin of Reference Person
Hispanic
5,235
12,566,477
80.35%
0.768%
B
Non-Hispanic
47,932
93,864,304
82.14%
0.233%
A
Blank
279
539,500
20.85%
2.936%
C
Age of Reference Person
15-29
6,150
12,122,686
84.55%
0.533%
A
30-39
8,645
17,466,937
83.03%
0.435%
B
40-49
9,769
19,743,802
81.89%
0.453%
C
50-59
10,631
21,133,803
82.88%
0.446%
B
60-69
8,323
16,447,312
83.76%
0.450%
A, B
70+
8,331
16,608,451
82.81%
0.483%
B, C
Blank or Less than 15
1,597
3,447,290
38.95%
1.391%
D
Overall
53,446
106,970,281
81.62%
0.214%
Source: U.S. Census Bureau internal data from December 2011 CPS interviews.
* May not sum to totals due to rounding. For weighted percent of total sample, see Table D-6.
N indicates that the estimate is not comparable.
+
Exercise caution: The sample size is extremely small, leading to unreliable estimates.
×
Within each characteristic, response rates identified with the same letter are not significantly different at the
α=0.10 level. A indicates the highest response rates, B indicates the next highest rates, etc.

28
29

See footnote 4.
See footnote 5.

D-3

Attachment D
Table D-4: Response Rates for 2011 FSS Households for Characteristics Only Available for Responding
CPS Households
Unweighted
Weighted
Weighted
Standard Significance
Characteristic
Households
Households*
Response Error (%)
Grouping×
Rate (%)
Tenure (Edited)
Owned or Mortgage
36,182
71,705,134
81.99%
0.250%
A
Rented for Cash
16,557
33,883,724
80.79%
0.354%
B
No Cash Rent
707
1,381,423
82.77%
1.531%
A, B
Family Income
Less than $10,000
3,430
7,056,158
84.37%
0.738%
E
$10,000-$19,999.99
5,264
10,723,830
87.83%
0.506%
B, C, D
$20,000-$29,999.99
5,152
10,472,197
86.82%
0.555%
D
$30,000-$39,999.99
4,917
9,799,131
87.33%
0.532%
D
$40,000-$49,999.99
3,552
6,961,920
89.05%
0.558%
A
$50,000-$59,999.99
3,538
6,997,425
88.90%
0.613%
A, C
$60,000-$74,999.99
4,315
8,397,841
89.36%
0.492%
A
$75,000-$99,999.99
4,485
8,871,528
89.01%
0.563%
A, B
$100,000-$149,999.99
4,111
8,009,206
89.48%
0.547%
A
$150,000+
3,016
6,191,828
87.43%
0.722%
C, D
Blank or Don’t Know
1,899
3,843,176
61.61%
1.271%
F
Refused
9,767
19,646,041
58.62%
0.688%
G
Household Type
Husband/Wife Primary Family
26,447
52,796,322
82.16%
0.270%
A
Unmarried Householder
8,833
18,321,760
79.98%
0.507%
B
Primary Family
Primary Individual
18,130
35,772,699
81.66%
0.336%
A
Group Quarters with Family
13
31,524
75.34%
9.030%
A, B+
Group Quarters without Family
23
47,976
82.78%
9.566%
A, B+
Child(ren) Present (Edited)
No
38,971
77,408,984
81.26%
0.265%
B
Yes
14,475
29,561,297
82.57%
0.354%
A
Measure of Labor Force of Reference Person
Employed
32,159
63,422,068
82.16%
0.263%
B
Unemployed
2,444
5,044,325
83.69%
0.889%
A
Not in Labor Force
18,488
37,857,018
80.79%
0.350%
C
Blank
355
646,870
60.43%
2.931%
D
Overall
53,446
106,970,281
81.62%
0.214%
Source: U.S. Census Bureau internal data from December 2011 CPS interviews.
* May not sum to totals due to rounding. For weighted percent of total sample, see Table D-7.
+
Exercise caution: The sample size is small, leading to unreliable estimates.
×
Within each characteristic, response rates identified with the same letter are not significantly different at the
α=0.10 level. A indicates the highest response rates, B indicates the next highest rates, etc.

D-4

Attachment D

Table D-5: Respondent and Nonrespondent Distributions for December 2011 CPS Households
Household
Household
% of Total
Chi-Square
Characteristic
Respondent Nonrespondent
P-value
Sample*
Statistic (df)
%*
%*
Type of Living Quarters
Housing Unit
95.14%
95.00%
96.46%
26.8536 (2)
< 0.0001
Non-Housing Unit
4.77%
4.93%
3.30%
[18.8691 (1) < 0.0001]
Blank
0.09%
0.08%
0.24%
Principal City Status
Principal City within
33.02%
32.57%
37.16%
CBSA/MSA
Not Part of a Principal City
64.4473 (2)
< 0.0001
50.68%
50.69%
50.59%
within CBSA/MSA
Outside of a CBSA/MSA
16.30%
16.74%
12.25%
Region
Northeast
18.32%
17.93%
21.90%
Midwest
22.89%
23.27%
19.42%
56.2852 (3)
< 0.0001
South
36.44%
36.52%
35.70%
West
22.35%
22.29%
22.97%
Urban/Rural Status
Urban
79.85%
79.42%
83.89%
53.5985 (1)
< 0.0001
Rural
20.15%
20.58%
16.11%
Race of Reference Person
White Only
73.62%
77.81%
34.95%
Black Only
11.26%
11.67%
7.40%
Asian Only
3.77%
3.92%
2.35%
10,491.5877 (4)
< 0.0001
Other Race/
[36.3750 (3)
< 0.0001]
1.91%
2.01%
0.97%
Two or More Races
Blank
9.44%
4.58%
54.33%
Gender of Reference Person
Male
47.84%
50.22%
25.86%
22,548.6033 (2)
< 0.0001
Female
47.39%
49.77%
25.37%
[0.0597 (1)
0.8070]
Blank
4.77%
Z%
48.77%
Hispanic Origin of Reference Person
Hispanic
11.15%
11.65%
6.53%
18,185.3438 (2)
< 0.0001
Non-Hispanic
83.30%
87.85%
41.29%
[7.0650 (1)
0.0079]
Blank
5.55%
0.50%
52.18%
Age of Reference Person
15-29
10.90%
11.31%
7.15%
30-39
15.55%
16.28%
8.75%
40-49
17.77%
18.46%
11.38%
14,075.5021 (6)
< 0.0001
50-59
18.61%
19.75%
8.05%
[180.1130 (5) < 0.0001]
60-69
14.41%
15.40%
5.34%
70+
14.40%
15.57%
3.59%
Blank or Less than 15
8.36%
3.23%
55.74%
Overall
100%
100%
100%
Source: U.S. Census Bureau internal data from December 2011 CPS interviews.
Z: value rounds to zero but is not zero.
* May not sum to totals due to rounding.
[] The values within brackets are the chi-square statistic, df, and p-value when the blanks are excluded from
the chi-square test.

D-5

Attachment D
Table D-6: Respondent and Nonrespondent Distributions for 2011 FSS Households
Household
Household
% of Total
Chi-Square
Characteristic
Respondent Nonrespondent
P-value
Sample*
Statistic (df)
%*
%*
Type of Living Quarters
Housing Unit
95.05%
95.00%
95.28%
1.0072 (2)
0.6043
Non-Housing Unit
4.87%
4.92%
4.64%
[0.9522 (1)
0.3292]
Blank
0.08%
0.08%
0.08%
Principal City Status
Principal City within
33.02%
32.69%
34.45%
CBSA/MSA
Not Part of a Principal City
4.5178 (2)
0.1045
50.65%
50.82%
49.91%
within CBSA/MSA
Outside of a CBSA/MSA
16.33%
16.48%
15.63%
Region
Northeast
18.32%
17.74%
20.85%
Midwest
22.89%
23.56%
19.90%
44.7775 (3)
< 0.0001
South
36.44%
36.46%
36.37%
West
22.35%
22.24%
22.88%
Urban/Rural Status
Urban
79.71%
79.34%
81.35%
11.0281 (1)
0.0009
Rural
20.29%
20.66%
18.65%
Race of Reference Person
White Only
77.59%
78.87%
71.91%
Black Only
11.78%
11.21%
14.29%
Asian Only
3.98%
3.83%
4.64%
188.0986 (4)
< 0.0001
Other Race/
[85.5954 (3)
< 0.0001]
2.02%
2.01%
2.06%
Two or More Races
Blank
4.64%
4.08%
7.10%
Gender of Reference Person
Male
50.17%
50.19%
50.09%
N/A (2)
N/A
Female
49.82%
49.81%
49.88%
[0.0215 (1)
0.8835]
Blank
Z%
0.00%
0.03%
Hispanic Origin of Reference Person
Hispanic
11.75%
11.56%
12.56%
351.7495 (2)
< 0.0001
Non-Hispanic
87.75%
88.31%
85.27%
[4.9394 (1)
0.0263]
Blank
0.50%
0.13%
2.17%
Age of Reference Person
15-29
11.33%
11.74%
9.53%
30-39
16.33%
16.61%
15.07%
40-49
18.46%
18.52%
18.18%
1,703.3365 (6)
< 0.0001
50-59
19.76%
20.06%
18.40%
[17.8574 (5)
0.0031]
60-69
15.38%
15.78%
13.59%
70+
15.53%
15.75%
14.52%
Blank or Less than 15
3.22%
1.54%
10.70%
Overall
100%
100%
100%
Source: U.S. Census Bureau internal data from December 2011 CPS interviews.
Note: The chi-square test for gender of reference person could not be calculated due to a frequency of 0 within
the blank/respondent cell.
Z: value rounds to zero but is not zero.
* May not sum to totals due to rounding.
[] The values within brackets are the chi-square statistic, df, and p-value when the blanks are excluded from
the chi-square test.

D-6

Attachment D
Table D-7: Respondent and Nonrespondent Distributions for 2011 FSS Households for Characteristics
Only Available for Responding CPS Households
Household
Household
% of Total
Chi-Square
Characteristic
Respondent Nonrespondent
P-value
Sample*
Statistic (df)
%*
%*
Tenure (Edited)
Owned or Mortgage
67.03%
67.34%
65.69%
9.6869 (2)
0.0079
Rented for Cash
31.68%
31.36%
33.10%
No Cash Rent
1.29%
1.31%
1.21%
Family Income
Less than $10,000
6.60%
6.82%
5.61%
$10,000-$19,999.99
10.03%
10.79%
6.64%
$20,000-$19,999.99
9.79%
10.41%
7.02%
$30,000-$19,999.99
9.16%
9.80%
6.32%
$40,000-$19,999.99
6.51%
7.10%
3.88%
$50,000-$19,999.99
6.54%
7.12%
3.95%
3,853.2497 (11)
< 0.0001
$60,000-$19,999.99
7.85%
8.60%
4.54%
[61.9517 (9)
< 0.0001]
$75,000-$19,999.99
8.29%
9.04%
4.96%
$100,000-$19,999.99
7.49%
8.21%
4.28%
$150,000-$19,999.99
5.79%
6.20%
3.96%
Blank or Don’t Know
3.59%
2.71%
7.50%
Refused
18.37%
13.19%
41.34%
Household Type
Husband/Wife Primary
49.36%
49.69%
47.90%
Family
Unmarried Householder
17.13%
16.78%
18.65%
Primary Family
20.0066 (4)
0.0005
Primary Individual
33.44%
33.46%
33.37%
{18.1695 (3)
0.0004}
Group Quarters with
0.03%
0.03%
0.04%
Family
Group Quarters without
0.04%
0.05%
0.04%
Family
Child(ren) Present (Edited)
No
72.36%
72.04%
73.79%
8.5649 (1)
0.0034
Yes
27.64%
27.96%
26.21%
Measure of Labor Force of Reference Person
Employed
59.29%
59.68%
57.53%
Unemployed
4.72%
4.84%
4.18%
89.9370 (3)
< 0.0001
Not in Labor Force
35.39%
35.03%
36.98%
[15.4464 (2)
0.0004]
Blank
0.60%
0.45%
1.30%
Overall
100%
100%
100%
Source: U.S. Census Bureau internal data from December 2011 CPS interviews.
* May not sum to totals due to rounding.
[] The values within brackets are the chi-square statistic, df, and p-value when the blanks and refusals are
excluded from the chi-square test.
{} The values within braces are the chi-square statistic, df, and p-value when the GQ groups are combined.

D-7

Attachment E
In the following tables, the key differences between all 5 years of tables are summarized.
Note: within each of Tables E-1 and E-2, a reference to tables within the report refers to both
that report table as well as the corresponding tables within the attachments. For example,
Table 2 references Tables 2, A-2, B-2, C-2, and D-2.
In Table E-1, an X indicates that, excluding blanks/missings, there were no significant
differences in response rates for the identified characteristic for that year. This table includes
all characteristics that had no significant differences in response rates (excluding
blanks/missings) for any of the years, not just those that were different between the years.
Table E-1: Summary of Key Differences Between 2011-2015 Response Rate Results
Characteristic
2015 2014 2013 2012 2011
December CPS (Table 2)
Gender of Reference Person
X
X
X
X
Hispanic Origin of Reference Person
X
FSS (Table 3)
Type of Housing Unit
X
X
X
X
X
Principal City Status
X
Gender of Reference Person
X
X
X
FSS within Responding CPS Households (Table 4)
Child(ren) Present (Edited)
X
Source: U.S. Census Bureau internal data from December of 2011, 2012, 2013, 2014, and 2015 CPS interviews.

In Table E-2, an X indicates that, excluding blanks/missings, the respondent and
nonrespondent distributions were not significantly different for the identified characteristic
for that year. This table includes all characteristics that had no significant differences in
distributions (excluding blanks/missings) for any of the years, not just those that were
different between the years.
Table E-2: Summary of Key Differences Between 2011-2015 Respondent and Nonrespondent
Distribution Results
Characteristic
2015
2014
2013
2012 2011
December CPS (Table 5)
Gender of Reference Person
X
X
X
X
Hispanic Origin of Reference Person
X
FSS (Table 6)
Type of Housing Unit
X
X
X
X
X
Principal City Status
X
X
Gender of Reference Person
X
X
FSS within Responding CPS Households (Table 7)
Tenure (Edited)
X
Household Type
X
Child(ren) Present (Edited)
X
Measure of Labor Force of Reference Person
X
Source: U.S. Census Bureau internal data from December of 2011, 2012, 2013, 2014, and 2015 CPS interviews.

E-1