County Election Administration by the Numbers

This analysis draws primarily on the U.S. Election Assistance Commission's 2024 Election Administration and Voting Survey (EAVS) and county-level voter registration data published by state election authorities and compiled by NACo in 2026. The EAVS provides information on election administration, including voter registration activity, polling places, election workers and voting methods. Additional historical election data and research on voting before Election Day provide context for changes over time. County-level election administration indicators are available through NACo's County Explorer.

Notes:

• Election Administration and Voting Survey (EAVS). The EAVS is a biennial survey conducted by the U.S. Election Assistance Commission that collects information from state and local election officials on the administration of federal general elections. This analysis primarily uses the 2024 EAVS Public Release Version 2. Unless otherwise noted, EAVS findings refer to the November 2024 general election.
• County-level voter registration data. NACo compiled the latest available county-level voter registration figures from state election authorities in 2026. Reporting dates vary by state, and the figures represent a snapshot of registered voters rather than a uniform national reporting date. The analysis includes states for which usable county-level registration data were available. Registration totals and classifications may differ across states because of variations in voter registration systems and reporting practices.
• Geographic coverage and county-level analysis. County-level analyses focus on jurisdictions relevant to county election administration. U.S. territories, Connecticut and Rhode Island are excluded because they do not have active county governments. Other jurisdictions are excluded or treated separately where election administration occurs primarily at the municipal, township or state level. New York City's five boroughs are consolidated into a single county-equivalent geography for NACo's county-level analysis. Geographic coverage may vary by indicator depending on the reporting structure and availability of usable data.
• Data quality and reporting limitations. EAVS reporting varies across jurisdictions, and some jurisdictions do not provide complete or internally consistent responses for every measure. NACo applies data cleaning, standardization and validation procedures before preparing county-level indicators for analysis and publication. As a result, the number of counties included varies by indicator depending on data availability and quality. For example, the analysis of Election Day poll workers per polling place includes 2,291 counties with usable data for both measures. Missing or suppressed values are excluded from the relevant calculations and should not be interpreted as zero.
• Election workforce and polling places. Poll worker and polling place measures distinguish between Election Day and early voting operations. Where applicable, the analysis uses Election Day poll workers per Election Day polling place to describe staffing patterns across counties. Ratios are calculated only for jurisdictions with usable data for both measures. These figures describe reported staffing levels and do not establish a recommended staffing standard.
• Voting methods and historical comparisons. National voting-method estimates for 2020, 2022 and 2024 are drawn from the EAC's 2024 EAVS Comprehensive Report. These percentages use state-level casewise deletion and may not sum to 100 percent. Historical findings for 2008–2018 are based on NACo's earlier analysis of EAVS data for jurisdictions relevant to county election administration. Because the historical county-level analysis and the more recent national estimates differ in geographic coverage and methodology, they should not be interpreted as a directly comparable time series
• 2026 election context. Information on the availability of early in-person and no-excuse mail voting is drawn from research published by the Center for Election Innovation & Research. Any preliminary 2026 voting totals are drawn from the University of Florida Election Lab's early voting tracker. These figures reflect reported voting activity as of the date indicated and may not include all jurisdictions. They should not be interpreted as final election results.

County Explorer is NACo's interactive data platform, providing more than 800 demographic, economic, fiscal and county administration indicators to help county leaders better understand the communities they serve. Visit ce.naco.org to explore voter registration, election administration, polling places, election workers and other indicators across America's counties.

1. States have more oversight over policymaking and certain types of elections, while local governments, particularly counties, oversee the implementation of elections at the local level. See Karen L. Shanton, “The State and Local Role in Election Administration: Duties and Structures,” Congressional Research Service, available at: https://www.everycrsreport.com/files/20190304_R45549_24cb977993f4f4b2a1a7fbd7a3e2ba0e32aafbf6.pdf (March 4, 2019). ↩
2. States have significant discretion in how they conduct elections. Most specifics regarding the voting process are governed by state law rather than federal law, and each state has its own unique set of regulations for administering federal elections. This creates a diverse array of local election systems, both between and within states, with counties primarily responsible for administering elections. U.S. Election Assistance Commission, “2024 Election Administration and Voting Survey Comprehensive Report,” available at https://www.eac.gov/sites/default/files/2025-07/2024_EAVS_Report_508.pdf (June 2025). ↩
3. Election administration varies by state and local governance structure, and specific arrangements may also vary within states. See NACo Research and Landscape Project, 2022. ↩
4. State classifications reflect the predominant structure of local election administration and may include jurisdiction-specific exceptions. NACo Research and Landscape Project, 2022. ↩
5. The six states where counties and municipalities share election administration responsibilities are Illinois, Michigan, Minnesota, Mississippi, Virginia and Wisconsin. NACo Research and Landscape Project, 2022. ↩
6. Election administration structures vary across states. Missouri is classified as primarily county-administered, but the Kansas City Board of Election Commissioners independently administers elections for the portion of Kansas City located in Jackson County. The Jackson County Election Board administers elections in the remainder of Jackson County. Counties and municipalities share election administration responsibilities in six states: Illinois, Michigan, Minnesota, Mississippi, Virginia and Wisconsin. Municipalities administer elections in four states: Maine, Massachusetts, New Hampshire and Vermont. Alaska and Delaware have state-managed oversight, with collaboration between election officials and an election board. Source: NACo Research and Landscape Project. ↩
7. Based on NACo's analysis of the latest available county-level voter registration data from state election authorities. County-level voter registration data were not available for Alaska, Hawaii, Louisiana, Missouri, North Dakota and Vermont for this release. North Dakota does not require voter registration and is therefore not represented in the registration total. Instead, voter eligibility is verified at the polls. The state maintains a centralized voter file containing information on individuals who have voted in previous elections, with county auditors responsible for entering and maintaining voter information. The finding that more than four in five registered voters live in states where counties have primary responsibility for administering elections is calculated using NACo's 2026 county-level voter registration data and the state governance classifications from the NACo Research and Governance Project (2022). The percentage represents the share of registered voters in the available dataset who live in primarily county-administered states, rather than the share of all U.S. registered voters whose records are directly maintained by counties (Of the 196,228,437 registered voters represented in the available data, 161,988,031 lived in states classified as primarily county-administered, equivalent to 82.6%.) Sources: NACo analysis of voter registration data from state election authorities; NACo Research and Governance Project (2022). ↩
8. NACo Analysis of voter registration data from state election authorities. ↩
9. Voter registration transactions. Based on NACo's analysis of the 2024 Election Administration and Voting Survey (EAVS), revised dataset (Version 2.0). The analysis includes 2,649 counties across 40 states with usable registration transaction data. Counties with unavailable or excluded data are not included in the total. Registration transactions cover the period between the close of registration for the 2022 and 2024 general elections and may include multiple transactions involving the same voter. ↩
10. Voter registration data and reporting limitations. Voter registration figures reflect the latest county-level data available from individual state election authorities as of September 2026. Reference dates range from January to September 2026, and county-level registration data are not publicly available for every state. State election systems and reporting practices vary. In some states, registration statistics include both active and inactive voters, while in others, only active voters are reported. These differences should be considered when comparing registered voter totals across counties and states. Source: NACo analysis of voter registration data from state election authorities. ↩
11. Federal law establishes requirements for state voter registration systems. The National Voter Registration Act of 1993 (NVRA) expanded voter registration opportunities and requires states to conduct a uniform and nondiscriminatory program to remove individuals no longer eligible from voter lists. The Help America Vote Act of 2002 (HAVA) requires states to maintain a computerized statewide voter registration list. Within these federal requirements, states retain substantial authority over voter registration policies and procedures. See U.S. Election Assistance Commission, “2024 Election Administration and Voting Survey Comprehensive Report,” available at https://www.eac.gov/sites/default/files/2025-07/2024_EAVS_Report_508.pdf (June 2025). ↩
12. Responsibility for maintaining voter registration records varies across states. According to the 2024 Election Administration Policy Survey, local officials alone were responsible for modifying or removing registration records in 58% of states, while state and local officials shared responsibility in 25%. State officials alone held this responsibility in 16 % of states. State voter registration database structures also differ. The U.S. Election Assistance Commission classifies systems as top-down, bottom-up or hybrid. Most states reported using top-down systems in 2024, while 14 states reported using bottom-up or hybrid systems. Maintaining registration records also involves coordination across government data systems. In 2024, 91% of states reported linking voter registration data to motor vehicle agencies, 78% to entities maintaining death records, and 64% to entities maintaining felony or prison records. See U.S. Election Assistance Commission, “2024 Election Administration and Voting Survey Comprehensive Report,” available at https://www.eac.gov/sites/default/files/2025-07/2024_EAVS_Report_508.pdf (June 2025). ↩
13. Same-day registration policies vary when registration and voting may occur, including during early in-person voting, on Election Day or both. In 2024, county election officials processed more than 1.9 million same-day registrations. See Center for Election Innovation & Research, “2026 Voter Registration Deadlines,” available at https://electioninnovation.org/research/voter-registration-deadlines-2026/ (June 2026) and NACo Analysis of the 2024 Election Administration and Voting Survey (EAVS). ↩
14. For the Election Administration and Voting Survey (EAVS), a voter registration transaction is defined as a unit of work performed on a voter registration record. Transactions may include adding a new registration, updating an existing record or removing a record when an individual is no longer eligible. Multiple transactions may be associated with the same voter during the reporting period, including when a voter moves between jurisdictions. See U.S. Election Assistance Commission, “2024 Election Administration and Voting Survey Comprehensive Report,” available at https://www.eac.gov/sites/default/files/2025-07/2024_EAVS_Report_508.pdf (June 2025). Other reported registration transaction categories include duplicate registrations, invalid or rejected registrations, pre-registrations, other transactions and uncategorized transactions. The EAC calculates transaction percentages separately, and the reported percentages do not sum to 100%. ↩
15. Voter registration methods have changed substantially over time. The number of states allowing online voter registration increased from two in 2008 to 43 in 2020. Meanwhile, registrations received in person or by mail, email and fax declined as a share of registrations, accounting for 15.2 % in 2022. In the 2024 EAVS, mail, fax and email accounted for 8.5 % of registration transactions, while in-person registration accounted for 6.4 %. Historical comparisons should be interpreted with caution. Beginning with the 2024 EAVS, the U.S. Election Assistance Commission shifted Section A from measuring voter registration forms to voter registration transactions to better reflect increasingly electronic registration processes. A registration transaction is a unit of work performed on a voter registration record and may include adding a new record, updating an existing record or removing a record when an individual is no longer eligible in that jurisdiction. Because of this methodological change, 2024 registration transaction figures should not be treated as directly comparable with earlier counts of registration forms. See U.S. Election Assistance Commission, “The History, Evolution, and Future Directions of the Election Administration and Voting Survey,” available at https://www.eac.gov/research-and-data/eavs-retrospective (September 2025). ↩
16. NACo Analysis of the 2024 Election Administration and Voting Survey (EAVS), revised dataset (Version 2.0). Analysis is limited to jurisdictions relevant to county election administration and excludes U.S. territories, Connecticut and Rhode Island, which do not have active county governments, as well as municipalities and other local jurisdictions that administer elections independently of their corresponding county. New York City boroughs are consolidated. Election Day and early voting activity are reported separately and should not be added together. A poll worker who served during both periods may be counted in both Election Day and early voting figures, whereas the overall measure counts each poll worker only once. Similarly, a physical polling place used during both periods is counted once in the overall measure. Consistent with the EAC's treatment of the 2024 data, Arkansas is excluded from Election Day and early voting polling-place measures because of inconsistencies in the state's reported polling-place figures. For purposes of the EAVS, poll workers include individuals who perform poll-worker functions during early voting or on Election Day. Observers stationed at polling places, regular office staff who do not perform poll-worker functions and temporary election staff not hired specifically to serve voters during early voting or on Election Day are not counted as poll workers. Regular office staff who perform poll-worker duties during early voting or on Election Day are included. ↩
17. The age distribution of poll workers has shifted since 2020, when workers ages 41 to 60 represented the largest age group. U.S. Election Assistance Commission, “2024 Election Administration and Voting Survey Comprehensive Report,” available at https://www.eac.gov/sites/default/files/2025-07/2024_EAVS_Report_508.pdf (June 2025). ↩
18. NACo Analysis of the 2024 Election Administration and Voting Survey (EAVS), revised dataset (Version 2.0). Analysis is limited to jurisdictions relevant to county election administration and excludes U.S. territories, Connecticut and Rhode Island, which do not have active county governments, as well as municipalities and other local jurisdictions that administer elections independently of their corresponding county. New York City boroughs are consolidated. Election Day and early voting activity are reported separately and should not be added together. A poll worker who served during both periods may be counted in both Election Day and early voting figures, whereas the overall measure counts each poll worker only once. Similarly, a physical polling place used during both periods is counted once in the overall measure. Consistent with the EAC's treatment of the 2024 data, Arkansas is excluded from Election Day and early voting polling-place measures because of inconsistencies in the state's reported polling-place figures. For purposes of the EAVS, poll workers include individuals who perform poll-worker functions during early voting or on Election Day. Observers stationed at polling places, regular office staff who do not perform poll-worker functions and temporary election staff not hired specifically to serve voters during early voting or on Election Day are not counted as poll workers. Regular office staff who perform poll-worker duties during early voting or on Election Day are included. ↩
19. Recruitment challenges varied by election jurisdiction size. Jurisdictions serving 10,000 or fewer registered voters generally reported less difficulty recruiting poll workers than larger jurisdictions. This finding reflects election jurisdictions nationwide and is not limited to county governments. See U.S. Election Assistance Commission, “2024 Election Administration and Voting Survey Comprehensive Report,” available at https://www.eac.gov/sites/default/files/2025-07/2024_EAVS_Report_508.pdf (June 2025). ↩
20. National Conference of State Legislatures, “Election Poll Workers,” available at https://www.ncsl.org/elections-and-campaigns/election-poll-workers (updated September 2, 2026). ↩
21. The Election Administration and Voting Survey (EAVS) distinguishes between precincts, defined as geographic areas to which voters are assigned based on their residence, and physical polling places. A physical polling place may serve voters from one or multiple precincts. Vote centers operate differently by allowing voters to cast ballots at any vote center within their jurisdiction rather than requiring them to use a specifically assigned polling place. See U.S. Election Assistance Commission, “2024 Election Administration and Voting Survey Comprehensive Report,” available at https://www.eac.gov/sites/default/files/2025-07/2024_EAVS_Report_508.pdf (June 2025). ↩
22. NACo Analysis of the 2024 Election Administration and Voting Survey (EAVS), revised dataset (Version 2.0). The analysis includes 2,291 counties with usable values for both Election Day poll workers and Election Day polling places after applying NACo's county-government universe and data-quality exclusions. The number of poll workers per polling place is calculated by dividing the number of reported Election Day poll workers by the number of reported Election Day polling places for each county. The resulting measure describes how reported Election Day poll workers are distributed across polling places and should not be interpreted as a measure of staffing adequacy, efficiency or workload. Polling-place structures vary across jurisdictions, including whether locations serve one or multiple precincts or operate as vote centers. ↩
23. NACo calculated the number of Election Day poll workers per polling place by dividing the number of Election Day poll workers by the number of Election Day polling places for each county with usable data on both measures. The analysis includes 2,291 counties. For mapping purposes, county ratios were grouped into five ranges: Under 5, 5 to less than 7, 7 to less than 9, 9 to less than 12 and 12 or more poll workers per polling place. Because county ratios may include decimal values, these categories correspond to values below 5.0, from 5.0 to less than 7.0, from 7.0 to less than 9.0, from 9.0 to less than 12.0 and 12.0 or higher, respectively. Counties without usable data are shown in gray. ↩
24. The report provides the latest complete national general-election data available for this analysis. Voting-method percentages are calculated using state-level casewise deletion, excluding states with missing or unusable data for the relevant measure. Consequently, the reported percentages do not necessarily sum to 100 percent. See U.S. Election Assistance Commission, “2024 Election Administration and Voting Survey Comprehensive Report,” available at https://www.eac.gov/sites/default/files/2025-07/2024_EAVS_Report_508.pdf (June 2025). ↩
25. NACo analysis of the U.S. Election Assistance Commission's Election Administration and Voting Survey (EAVS), 2008–2018. The analysis is limited to jurisdictions relevant to county election administration and excludes U.S. territories, Alaska, Connecticut, Maine, Massachusetts, New Hampshire, Rhode Island, Vermont and Wisconsin, as well as municipalities and other local jurisdictions reporting independently of their corresponding counties. Some jurisdictions did not report participation by voting method. ↩
26. U.S. Election Assistance Commission, “2024 Election Administration and Voting Survey Comprehensive Report,” available at https://www.eac.gov/sites/default/files/2025-07/2024_EAVS_Report_508.pdf (June 2025). ↩
27. Center for Election Innovation & Research, The Expansion of Voting Before Election Day, 2000–2026 and Dates for 2026 Early In-Person and Mail Voting (2026). Available at https://electioninnovation.org/research/expansion-voting-before-election-day-2000-2026/ and https://electioninnovation.org/research/dates-for-2026-early-in-person-and-mail-voting/. ↩