U.S. flag

An official website of the United States government

Skip Header

2016 Annual Survey of Public Employment & Payroll Methodology

Population of Interest: 2016

The population of interest for this survey includes all agencies of the 50 state governments, and 90,056 local governments (i.e., counties, municipalities, townships, special districts, and school districts) including the District of Columbia.

Data Collection: 2016

Confidentiality: Data collected for the Annual Survey of Public Employment & Payroll are public record and are not confidential, as authorized by Title 13, U.S. Code, Section 9(b).

Dates of Collection: The following are important dates in the data collection process.

2016 Annual Survey of Public Employment & Payroll


Initial mail-out


Reminder letter mail-out


Follow-up mail-out


Preliminary Release to the Census Bureau Internet

Methods: These data tabulations are based on information obtained in the Annual Survey of Public Employment & Payroll.  44 of the state governments provided data from central payroll records for all or most of their agencies/institutions. Data for agencies and institutions for the remaining state governments were obtained by mail canvass questionnaires. Local governments were also canvassed using a mail questionnaire. However, elementary and secondary school system data in Delaware, Florida, and North Dakota were supplied by special arrangements with the state government in each of these states. All respondents had the option of completing the survey using a web-based survey instrument developed for reporting the data. The online survey instrument was completed by 33.4 percent of the state-level responding units and 95.3 percent of the local government respondents.

Data Processing: 2016

Editing: Editing is a process that tries to ensure the accuracy, completeness, and consistency of survey data. Efforts are made at all phases of collection, processing, and tabulation to minimize reporting, keying, and processing errors.

Although some edits are built into the Internet data collection instrument and the data entry programs, the majority of the edits are performed post collection. Edits consist primarily of two types: (1) consistency edits and (2) ratio edits.

The consistency edits check the logical relationships of data items reported on the form. For example, if a value exists for employees for a function then a value must exist for payroll also. If part-time employees and payroll are reported then part-time hours must be reported and vice versa.

For each function reported for the employees, the ratio edits compare data for the number of employees and the average salary between reporting years. If data fall outside of acceptable tolerance levels, the item is flagged for review.

For consistency edits and ratio edits, the edit results are reviewed by analysts and adjusted as needed. When the analyst is unable to resolve or accept the edit failure, contact is made with the respondent to verify or correct the reported data.

Imputation: Not all respondents answer every item on the questionnaire.  There are also questionnaires that are not returned despite efforts to gain a response.  Imputation is the process of filling in missing or invalid data with reasonable values in order to have a complete data set for estimating state and national totals.

For nonresponding general purpose governments, dependent and independent school districts, and for special district governments, the imputations were based on recent historical data from the prior year annual survey.  These data were adjusted by a growth rate that was determined by the growth of responding units that were similar (in size, geography, and type of government) to the nonrespondent.  If there were no recent historical data available, the imputations were based on the data from a randomly selected responding donor that was similar (based on the same criteria) to the nonrespondent.  For general purpose governments, and for dependent and independent school districts, the selected donor’s data were adjusted by dividing each data item by the population (or enrollment) of the donor and multiplying the result by the nonrespondent’s population (or enrollment).

Tabulation: After the 2016 Annual Survey of Public Employment & Payroll data were edited and imputed, they were aggregated to produce the downloadable tables that are available on the website. For employment and pay statistics, full-time employees, full-time pay, part-time employees, part-time pay, full-time equivalent employment, and total March pay are published.

Sampling Variability: The data that are provided come from a sample rather than a census of all possible units.  The particular sample that was selected is one of a large number of possible samples of the same size and sample design that could have been selected.  A different sample would have yielded different estimates.  The estimated coefficient of variation, which is provided for each estimate, is an estimate of the sampling variability.  In this tabulation, the coefficients of variation are expressed as percentages.  The coefficient of variation (CV) is the ratio of the standard error to the expectation of the estimate.  We used a Taylor series method, or the estimate of the posterior distributions, to estimate the standard errors.

State government employment and payroll data are not subject to sampling error.  Consequently, state and local government estimates for individual states are more reliable statistically than the local government only estimates.


Data Quality: 2016

Nonsampling Errors: Although every effort is made in all phases of collection, processing, and tabulation to minimize errors, the data are subject to nonsampling errors such as inability to obtain data for every variable from all units in the population of interest, inaccuracies in classification, response errors, misinterpretation of questions, mistakes in keying and coding, and coverage errors. The Data Processing section describes our efforts to mitigate errors due to nonresponse, keying, reporting errors, etc.

Modal Distribution: Each respondent that received a questionnaire had the option of reporting data using a website developed for reporting data electronically, or working directly with staff members to report over the phone, fax or email. In addition, some governments have developed alternative reporting arrangements, known as central collection. The following table shows the response rate by mode for state and local governments that reported to the 2016 Annual Survey of Public Employment & Payroll.

State Governments

Local Governments




Central Collection






Overall Unit Response Rate: The overall unit response rate to the 2016 Annual Survey of Public Employment & Payroll was 79.2 percent. All unit response rates are well above the 60 percent Census Bureau’s quality standard.  The key variables for the survey are total employment and total payroll. The unit response rate was calculated for each state as well as for the total U.S., and gives the percentage of the units in the eligible universe that actually responded to the survey.

For 2016, weighted response rates are published for each item. This rate is calculated by dividing the weighted value of the item as reported by respondents by the weighted value of the item reported for respondents and imputations for nonrespondents.

Total Quantity Response Rate: The Total Quantity Response Rate (TQRR) was also calculated for the key variables for each state. The key variables for the survey are total employment and total payroll.  This response rate is computed separately for each key variable by summing the data provided by the respondents for the key variable and dividing this sum by the sum of the respondent data and the imputed data for the key variable. The result is multiplied by 100. Files of the unit response rates and TQRR’s for all states are available in the Response Rate Tabulations section below.

The Census Bureau's quality standard on releasing data products requires a 70 percent TQRR for the key variable estimates. However, we found noncompliance of the standard at all three levels of estimates produced:

For the combined state and local governments estimates, eight states (Louisiana, Minnesota, Mississippi, Nebraska, New Jersey, Tennessee, Vermont, and Washington) failed to meet the 70 percent TQRR standard for at least one of the combined estimates for the key variables (total employment and total payroll). 

For the state governments, there are nine states (Maine, Maryland, Minnesota, Nebraska, Nevada, New Jersey, North Carolina, Oregon, and Tennessee) that are noncompliant with the TQRR standard for at least one of the state government’s key variables. 

For the local estimates, there are 12 states (Colorado, Connecticut, Louisiana, Massachusetts, Michigan, Minnesota, Mississippi, New Hampshire, Rhode Island, Vermont, Washington, and Wisconsin) that are noncompliant with the TQRR standard for at least one of the local government’s key variables.

Page Last Revised - October 8, 2021
Is this page helpful?
Thumbs Up Image Yes Thumbs Down Image No
255 characters maximum 255 characters maximum reached
Thank you for your feedback.
Comments or suggestions?


Back to Header