SUPPORTING STATEMENT - SECTION B v2

5

SUPPORTING STATEMENT

U.S. Department of Commerce

U.S. Census Bureau

Public Employment & Payroll Forms

(Forms E-1, E-2, E-3, E-4, E-5, E-6, E-7, E-9)

OMB Control No. 0607-0452

Part B. Collections of Information Employing Statistical Methods

1. Universe and Respondent Selection

Data in 2012 are collected for all 50 state governments and all local governments in the United States. There is no sampling for 2012. We expect to mail survey forms to 99,935 government entities with a preliminary universe of 90,515 governments.

Data in 2013 are collected for all 50 state governments and for a sample of local governments, including general-purpose governments, school districts, and special districts.

The sample for the 2013 annual survey was selected based on the 2007 Census of Governments in 2009 with births added and deaths removed since the 2007 Census. Sample counts at the parent government level by type of local governments for the 2013 survey are similar to the counts for the 2010 and 2011 survey cycles.

Type of Local Government	Estimate of Sample	2007 Universe
Counties	1,456	3,033
Municipalities	3,044	19,492
Townships	623	16,519
School Districts	2,148	13,051
Special Districts	3,194	37,381
Total	10,465	89,476

Historically, the unit response rate for the Public Employment & Payroll program was 75-80 percent. For 2008 and 2009, the unit response rate was in the 88-91 percent range. The unit response rate for the 2007 Census of Governments – Employment was 88.5 percent.

2. Procedures for Collecting Information

The 2012 Census of Governments: Employment is a census of all government in the United States. There is no estimation.

The 2013 Annual Survey of Public Employment & Payroll is a sample. The sample is designed and selected to include governments that comprise the bulk of employment and payroll activity. Initial criteria for certainty units are as follows: 1) all counties having a population of 100,000 or greater in 2007, 2) all municipalities having a population of 75,000 or greater in 2007, 3) all townships having a population of 50,000 or greater in 2007, 4) all independent school districts having an enrollment of 10,000 or more in 2007, and 5) all special districts having a full-time equivalent of 1,000 or more. Noncertainty units are selected with probability proportional to the size (PPS) of the unit. For this sample and for future births in the annual survey years, the Census Bureau will sample all new general-purpose governments and schools and 1 in 25 of all newly established special district governments. For the 2009 sample, the Census Bureau selected a sample of 1 in 25 of the governmental units that had no activity (no employees or payroll) in 2007. Units in the PPS portion of the sample were sub-sampled below a cutoff for townships and special districts.

In Mid to late 2009, the Maine school systems began to major reorganization of their systems resulting in a number of unit deaths and births (mergers). In the 2011 survey sample, all Maine independent school systems were all sampled. This was done to provide quality estimates for the 2011 sample year given the reorganiztion. The addition of the units and removal of units to the sample will be carried forward to the 2013 sample.

The Census Bureau will prepare estimates by state for local government employment and payrolls (full-time employment, full-time payroll, part-time employment, and part-time payrolls) and by government function in sample years. For national level totals, the Census Bureau will prepare estimates of total local government employment and payrolls by government function. Model-assisted methods will be used to determine state-by-type of governmental totals. Small-area estimation methods will be used to determine functional estimates.

3. Methods to Maximize Response

The Governments Division consults with, and presents to, various public and professional groups that use the data. These groups endorse survey efforts and encourage participation by their members. Survey publicity appears in state organization publications distributed to member governments.

Survey procedures include original request questionnaires supplemented by one reminder letter (Attachment 5) and one follow-up mailing. Staff conduct telephone follow-up to determine if appropriate individuals have received mail correspondence, to encourage response, and to obtain additional information about reported data.

Central response arrangements with 45 state governments and for school systems in four states ensure 100 percent response rates for units included in those responses.

Alternative ways for reporting data are available for respondents. Data can be reported by paper form, electronic files, over the Internet, or by telephone.

The Census Bureau conducted Quality Improvement Program (QuIP) trips to make contact with government agencies or associations who could provide the most relevant information to determine is we are appropriately covering that states local governments. The QuIP trips gather data to reduce coverage gaps or over-coverage of the frame, and by verifying and updating contact information for certain governments that are determined to be ‘problematic’ to the Governments Division surveys. The definition of ‘problematic’ is defined as government units that have not responded in several years to both the Survey of Public Employment & Payroll and the Survey of Government Finances. The QuIP trips also provided exposure to the Governments Division work, educate government offices and organizations on what the Governments Division’s surveys, and promote the upcoming Census of Governments. These QuIP trips are expected to continue through the next few years.

4. Tests of Procedures or Methods

The Census Bureau used these questionnaires successfully in the last Census of Governments, and the employment and pay portion successfully in the Annual Survey of Public Employment & Payroll. The Census Bureau has used the basic definitions and reporting instructions since the 1957 Census of Governments. Ongoing contact with survey respondents indicates that present methods of data collection are effective.

The Census Bureau conducted cognitive testing in early 2011 on the 2012 form changes to ensure government representatives understood the questions being asked on the forms. The cognitive testing also gave survey analysts a better understanding of the data available and terminology used by government representatives.

5. Contacts for Statistical Aspects and Data Collection

Statistical procedures:

Carma Hogue

Assistant Division Chief, Statistical Research & Methodology

Governments Division

U.S. Census Bureau

(301) 763-4882

Data collection and analysis:

Erika H Becker Medina

Chief, Employment & Benefit Statistics Branch

Governments Division

U.S. Census Bureau

(301) 763-1494

List of Attachments

1. 2012 Forms E-1 through E-9

2. Letter of Support from BEA

3. 2012 Letter and file layout for Central Collection

4. 2012 Letter from the Director

5. 2012 Letter of Reminder

6. 2012 Letter of Endorsement

7. 2012 Postcard