How to Read the Census Voting and Registration Report — Turnout Bases, Self-Reports and Voting Methods

As of 10 November 2026, 01:13 KST (retrieved by this page at that time); latest listed on the agency page at retrieval: Voting and Registration in the Election of November 2024 (Current Population Reports P20-590), U.S. Census Bureau, report cover and publication page dated August 2026 (page last revised August 12, 2026). The detailed P20 tables for the same election were dated April 2025. Data: the Current Population Survey (CPS) November 2024 Voting and Registration Supplement, about the election held on November 5, 2024. Primary source: https://www2.census.gov/library/publications/2026/demo/p20-590.pdf^[1][2][5][6]

Comparison basis: counts are in millions, rounded by this page from Table 1 (thousands), and shares are percentages rounded to one decimal as printed. "For the remainder of this report, the voter turnout refers to the percentage of the CVAP who voted unless otherwise noted." "All comparisons presented in this report have taken sampling error into account and are significant at the 90 percent confidence level unless otherwise noted." The report adds: "All comparisons within this text are simple pair-wise comparisons; no multiple comparison adjustments were made to the tests." "We acknowledge that some of the conclusions made in this text may not remain statistically significant if the multiple comparison adjustments were included." The 2022 midterm survey is the previous wave, but the report compares presidential elections only.^[1][3]

In three lines

1. 65.3% of citizens 18 and older reported voting in November 2024 (Figure 1), down from 66.8% in 2020; 154.3 million voters, not significantly different from 2020 (Table 1).
2. Mail voting fell to 29.1% of voters from 43.2%, while early in-person voting rose to 31.0% (Figure 9).
3. Voters 65 and older were 28.1% of voters, the largest share since 1980 (Figure 2).^[1]

Self-reports from a household survey. The CPS samples approximately 60,000 occupied housing units in the 50 states and the District of Columbia. Eligible respondents are citizens 18 and older in the civilian noninstitutionalized population, and proxy answers are allowed. "Those who indicated that they voted are assumed to be registered and are asked further questions regarding their method of voting and registration." "Therefore, the supplement does not capture voting for citizens living in institutions in the United States or voting for citizens, whether civilian or military, residing outside of the United States who cast absentee ballots."^[1]

Four bases for a voting rate. "Voters made up 59.3 percent of the VAP, 65.3 percent of the CVAP, and 88.8 percent of the registered population in 2024." The voting-age population of 260.4 million includes 24.2 million noncitizens. "Nonrespondents of the main voting question were an estimated 14.1 percent of the CVAP in 2024." Nonrespondents stay in the citizen total, so the 65.3% counts them as not having reported voting.^[1]

Survey counts and official tallies. "In general, sample surveys like the CPS tend to yield higher voting rates than official results." Both editions list possible reasons on both sides, including ballots left out of official counts and respondents overreporting voting; neither report estimates the size of each.^[1][3]

Representation measures. "A group is overrepresented when its share of the voting population is significantly larger than its share of the CVAP; underrepresentation is the opposite, with a smaller share of people voting than their presence in the CVAP would suggest." The report also cautions about voter shares over time: "In other words, an increase in voter share for a particular demographic group could be due to an increase in their overall population, an increase in their relative representation among voters, or a combination of both."^[1]

+2.0% or +1.9% for the citizen population. The report says: "In 2024, the CVAP increased by 2.0 percent, the number of registered people increased by 3.3 percent, and the number of voters did not significantly change, remaining at approximately 154.3 million." From the rounded Table 1 figures, 236.1 / 231.6 gives +1.9%. Rounding of the published figures to the nearest 100,000 may explain the gap; the cause is not established.^[1]

Which population the 65.3% describes. The topic-page teaser of April 30, 2025 says: "In the 2024 presidential election, 73.6% of the voting-age population was registered to vote and 65.3% voted according to new voting and registration tables." The release it links to, of the same date, explicitly uses the citizen voting-age population: "In the 2024 presidential election, 73.6% (or 174 million people) of the citizen voting-age population was registered to vote and 65.3% (or 154 million people) voted according to new voting and registration tables released today by the U.S. Census Bureau." P20-590 also gives 65.3% for the citizen voting-age population; the share of the voting-age population is 59.3%. Dividing P20-590's rounded Table 1 figures (173.9 million registered by 236.1 million citizens) gives 73.7%, which differs from the release's 73.6%; rounding of the published figures may explain the gap, but the cause is not established.^[1][5][7]

48.5 million nonvoters, 18.2 + 28.0 million in Table 3. Registered nonvoters (18.2 million) plus the nonregistered (28.0 million) give 46.2 million, below the 48.5 million nonvoters in Figure 1. Footnote 22 limits the 28.0 million: "The estimate is limited to nonregistered Americans in the CVAP who said they did not vote and answered the reason for not voting question." The two totals cover different groups of respondents.^[1]

Income groups changed. The 2020 report's Table 2 used Under $20,000, $20,000 to $49,999, $50,000 to $99,999 and $100,000 and over, limited to people in families; the 2024 report uses Under $50,000 through $150,000 and over, and "Estimates are limited to persons in families who report income or refuse to answer the question." The income shares of voters in the two reports are not directly comparable.^[1][3]

Two dates for the 2020 report. P20-585's cover reads January 2022, while its publication page and the Census topic page list it as February 2022 (page last revised February 16, 2022). The difference between the cover and page dates is not established.^[3][4]

Methods. "Voting by mail reached its peak during the 2020 election (43.2 percent) and decreased by 14.1 percentage points in 2024." Early in-person voting rose to 31.0% of voters from 26.2%, and traditional Election Day voting was 39.9%. The 2020 report had said: "For the first time on record, a majority (69.4 percent) of voters cast ballots by a nontraditional method in the 2020 presidential election." Figure 9 notes: "The estimates presented in this figure are only for individuals with valid responses to the method and timing questions." The April 30, 2025 release, based on the detailed tables, gave 29.0% by mail, 30.7% in person before Election Day and 39.6% on Election Day; the differences from the report's 29.1%, 31.0% and 39.9% are not explained in either document and are not established.^[1][3][7]

Reasons. "With both nonparticipation questions, respondents can pick only one main reason." Of 18.2 million registered nonvoters in 2024, 19.7% said they were not interested or felt their vote would not make a difference, 17.8% were too busy and 14.7% disliked the candidates or campaign issues (Table 3). Of the 28.0 million not registered, 42.3% were not interested in the election or not involved in politics.^[1]

Age and representation. Voters 18 to 29 were 15.3% of voters and 19.6% of the citizen voting-age population in 2024, an underrepresentation of 4.4 points; voters 65 and older were overrepresented by 3.5 points (Table 2). In 2020 the figures were 3.9 and 2.7 points (P20-585, Table 2). Group differences are self-reported survey estimates with sampling error.^[1][3]

Dates kept apart. Election: November 5, 2024. Detailed P20 tables: April 2025 (news item April 30, 2025; tables page last revised August 10, 2026). P20-590 report: August 2026. Previous presidential report P20-585: cover January 2022, page February 2022. The topic page shows Page Last Revised June 20, 2025 although it lists the August 2026 report.^{[1][2][3][4][5][6]}

Next release. The topic page says: "The P20 detailed tables are released every two years following national level elections." It gives no date for the products on the November 2026 election, so the next release date is not established.^[5]