How to Read the Census Annual Survey of Public Employment & Payroll — Headcount, March Payroll and State vs Local

As of 6 November 2026, 00:56 KST (retrieved by this page at that time); most recently published report: Annual Survey of Public Employment & Payroll Summary Report: 2025 (G25-ASPEP) from the U.S. Census Bureau, released April 16, 2026, with data for March 2025. The Census release schedule page states: "The 2026 Annual Survey of Public Employment & Payroll will be released by the end of May 2027." Primary source: https://www.census.gov/content/dam/Census/library/publications/2025/econ/ASPEP%20Summary%20Report%202025.pdf^[1][2][4]

Comparison basis: all figures are as printed in the report's text, rounded to one decimal in millions of people or billions of dollars; the percent changes are the report's own, computed by Census, and this page does not recompute them from the rounded levels. Where the report gives a change but no previous level, the previous level is shown as not printed.^[1]

In three lines

1. This annual Census survey counts civilian employees of U.S. state and local governments in March and the payroll they were paid for that month.
2. In March 2025 these governments employed 20.3 million people, 1.4 percent more than in March 2024.
3. Their March 2025 payroll was $115.1 billion for the month, up 5.9 percent.^[1][3]

One month, civilian only. "The Survey of Public Employment & Payroll measures the number of state and local civilian government employees and their gross monthly payroll for March of the survey year." The payroll total is therefore a gross payroll for the single month of March, not an annual wage bill, and the employee figure is a headcount of full- and part-time workers, not full-time equivalents.^[3]

Sample or census. "A full census is conducted every five years (in year ending in ‘2’ and ‘7’)." "In the intervening years, data are collected from a sample of state and local governments." "A new sample is selected following each census." The 2025 figures therefore come from a sample year. The cited pages report national and state totals for state and local governments and do not state whether territories are included; they print no sampling errors, and this page does not test any change for significance.^[3]

Revisions. "In addition to the new 2026 data, revised tables for 2024 and 2025 will be released." The 2025 figures here are as first published in April 2026.^[4]

Functional categories. The report's figure of the top functions carries this note: ""All other and unallocable" is not a summation of the remaining functions shown separately in the source data; it refers only to employees not applicable to other employment functions."^[1]

State vs local. Local governments employed 14.7 million people (72.2 percent) and state governments 5.7 million (27.8 percent). The two rounded figures add to 20.4 million against the published 20.3 million (this page's sum); with three figures each rounded to 0.1 million, rounding alone can open a gap of up to 0.15 million, so the gap of 0.1 is within that range.^[1]

Full-time vs part-time. Of 20.3 million, 15.8 million were full-time and 4.5 million part-time. Full-time payroll was 92.7 percent of the total; part-time employees received $8.4 billion. The four full- and part-time payroll figures add to $115.2 billion against the published $115.1 billion (this page's sum), within the up-to-$0.25 billion range that rounding five figures to 0.1 can produce.^[1]

Headcount vs payroll growth. Employment rose 1.4 percent while March payroll rose 5.9 percent. The report gives both changes but does not establish why payroll grew faster; it does not separate pay rates, hours or the mix of full- and part-time jobs.^[1]

Three functions vs the rest. Education (11.4 million), hospitals (1.2 million) and police protection (1.0 million) add to 13.6 million in this page's sum, against the report's 13.7 million for these functions, a rounding-sized difference; over 6.6 million employees worked in other functions.^[1]