The 2026 Early Vote Is Running Ahead of 2022. Who Does That Help?

Figures as of October 28, 2026, six days before Election Day. Each state count runs through the date shown beside it; the state comparisons cover selected states and voting methods, so dates and coverage vary.

01The Ballot Race

North Carolina publishes the cleanest day-by-day count in the country, with the same report for 2022, so it is the best place to watch the surge happen. Through October 26 the state had accepted 1,773,889 ballots, against 1,298,679 at the same point four years ago. That is 37% more, and the gap in ballots has widened every day since early voting opened.

The national picture is harder to pin down. By the morning of October 28, state election offices had reported about 34 million early and mail ballots in the 46 states that publish counts, according to two independent trackers that compile the state files (On Point Politics had 33,937,923 and Call the Map 34,244,936, each state counted through its own latest file, between October 19 and October 28). The University of Florida's Election Lab, run by Michael McDonald, had already passed 30 million on October 26, with about half of them cast by mail.

How far ahead of 2022 that is depends on how you count, and there is no single official national comparison. Election Day is November 3 this year and was November 8 in 2022, so a calendar-date comparison flatters 2026 by five days of voting. Matched instead to the same number of days before Election Day, state by state, On Point Politics put 2026 about 7% ahead of 2022 across the states it tracks when I retrieved it at about 3 p.m. Eastern on October 28 (the tracker updates through the day). A rougher check gives a bigger number: McDonald's national counts in 2022 were about 21 million eight days out and 32 million five days out, which suggests something like 28 million six days out and would put 2026 about 20% ahead. That figure is my estimate, drawn between two snapshots taken at different hours, and early voting does not pile up at a steady rate, so treat it as a rough figure, probably on the high side, rather than a measurement. Two weeks before Election Day, an Associated Press analysis counted more than 13 million early ballots, against 9 million at the same point in 2022.

What does hold up under every method is the state-level picture: early voting is ahead of 2022 in most of the states where a like-for-like comparison exists.

02Pick Your State

Early voting is not up everywhere. Kansas, Virginia and Texas are running far ahead of 2022. Florida and Pennsylvania's mail ballots are flat or slightly behind. Pick a state to see its comparison, what it covers, the date it was measured and where the number comes from.

Ballots cast or returned before Election Day, not votes counted. What is counted: all early means in-person early voting plus mail and absentee ballots, statewide; Michigan counts mail ballots only because its early in-person voting began in 2024; Nevada counts Clark County mail ballots only. How 2022 is matched: the same day of the early-voting period, the same number of days before Election Day, or the fixed point AP and the state used. * Where the source publishes only a percentage change, the 2022 figure is implied from it and rounded. Associated Press figures are rounded by AP. The trackers compile state and county files and do not always agree with each other; Georgia, Colorado and other states without a consistent 2022 comparison are left out.

For the races these ballots will decide, the 2026 Midterm Forecast and the Midterm Polls and Prediction Markets Tracker show where each Senate and House contest stands, and the 2026 Midterm Money Tracker shows what has been spent on them.

03Who Is Voting Early

About half the states record party registration, and in those states the early electorate leans Democratic. Across the two dozen-plus states that report it, McDonald counted roughly 45% registered Democrats, 30% Republicans and 25% everyone else as of October 26.

Share of early and mail voters by registered party, from Call the Map's compilation of state files, October 26 to 28, 2026. * Pennsylvania counts mail ballots only; Nevada counts Clark County mail ballots only.

The change since 2022 does not run one way. In Pennsylvania's mail ballots, Clark County in Nevada and Iowa, the Democratic edge is smaller than it was at this point in 2022, which fits Republicans moving more of their own voters to early and mail ballots. In North Carolina, Colorado and Florida it went the other way: the gap moved toward Democrats, and in Florida a six-point Republican lead has shrunk to about one point. Maine is about where it was. Different states, different rules and different campaigns are behind those moves, and the registration numbers alone cannot say which explanation is right in each case.

Remember what these numbers are. Registration is not a vote. Unaffiliated voters, about a third of North Carolina's early ballots, decide close races, and so do registered Democrats and Republicans who cross over. Texas and Georgia, two of the biggest early-voting states, do not record party at all.

04The Vote Is Moving, Not Just Growing

Before reading too much into any early-vote record, it helps to see how fast Americans have changed the way they vote. In 2014, less than a third of midterm voters cast their ballot early or by mail. By 2022 it was half.

Share of midterm voters who voted early, by mail, or both. Source: US Census Bureau, Current Population Survey voting supplements (2014, 2018, 2022). In presidential years the share was 69% in 2020 and about 60% in 2024.

Across the three midterms shown, a bigger slice of the electorate voted before Election Day each time, so each new early-vote total tends to look like a record. In Georgia, nearly 244,000 people who voted on Election Day in 2022 have already voted this year, according to the Atlanta Journal-Constitution's analysis of state files. Those are not new voters. They are the same voters, earlier.

05Does Higher Turnout Favor the Party in Power?

This is the question everyone asks when the early numbers come in, so I went back through every midterm since 1946. Each dot is one election: the final share of eligible Americans who voted, and how many House seats the president's party won or lost. One limit up front: this measures final turnout, not the early vote and not the change from one election to the next, so it cannot say who benefits from this year's early-voting pace. It can say whether high-turnout midterms have tended to go one way.

Turnout of the voting-eligible population: US Elections Project, University of Florida Election Lab (version 1.4). House seats: The American Presidency Project, UC Santa Barbara. Blue dots are Democratic presidents, red dots Republican.

There is no clean rule. Across these 20 midterms the correlation between turnout and the president's party's net House-seat change (losses counted as negative numbers) is -0.16 by my calculation, close to zero. Split the 20 years in half by turnout and the ten higher-turnout midterms cost the president's party 28.7 House seats on average, against 22.6 in the ten lower ones. Split them at the average turnout instead and the gap all but disappears, 25.9 seats against 25.5. These 20 elections do not show a reliable simple relationship between turnout and the president's seat change.

The exceptions are the whole story:

• 2018 had the highest midterm turnout since 1914, 50%, and the president's party lost 41 House seats.
• 2022 had the second-highest midterm turnout since 1970, 46%, and the president's party lost only 9.
• 2010 had ordinary turnout, 41%, and produced the biggest loss on the chart, 63 seats.
• 1962 had 48% turnout and the president's party lost just 4.

Political scientists disagree about turnout itself. Thomas Hansford and Brad Gomez, using Election Day rainfall to measure what turnout alone does, found that higher turnout "has a significant anti-incumbency effect", though their study covered presidential elections and the effect was small: a four-point swing in turnout moved the Democratic share of the vote by just under one point. Daron Shaw and John Petrocik, in The Turnout Myth, went through decades of national elections and concluded there is "no consistent partisan effect associated with voter turnout." Back in 1980, James DeNardo found that heavy turnout tended to hurt whichever party held the majority, not the Democrats as such.

What lines up far better with seat losses is the president's standing. In a 2018 analysis of the midterms from 1946 to 2014, Gallup found that presidents below 50% approval lost 37 House seats on average. The only times the president's party gained seats on this chart, 1998 and 2002, the president was above 60%.

So here is the short answer to whether a big early vote is good news for the party in power or for the challengers: the turnout record alone does not tell you. The clearer patterns are in approval and enthusiasm. In 2006 and 2010 the president was below 50% approval in Gallup's polling (38% and 45%) and the out-party was clearly the more enthusiastic side; the challengers gained 30 and 63 House seats. In 1998 and 2002 the president was above 60% and his party gained seats. This year the president's approval stood at 31% in the Reuters/Ipsos poll published October 19, the lowest it has recorded in either of his terms. By the historical record, that is a bigger warning sign for the party in power than anything in the early-vote count. The most unpopular presidents heading into a midterm and the full history of midterm elections show how those years played out.

06The Enthusiasm Ledger

Turnout is measured after the fact. Enthusiasm can be measured before it. Since 1994, Gallup has asked voters in each midterm year whether they are more enthusiastic than usual about voting. Here is the final reading before each election and what happened in the House.

Share saying they are "more enthusiastic than usual" about voting, final Gallup reading before each election. House results: The American Presidency Project.

Excluding the 1998 tie and the one-point gap in 2018, the party with the higher enthusiasm reading won the House in five of six elections. That is a small sample, and some of the gaps are within the poll's margin of error: in 2022, when Democrats ended slightly more enthusiastic and still lost the House narrowly, Gallup's margins for each party group were about five to six points. In 2006, 2010 and 2014 the more enthusiastic side was the out-party, and it gained seats each time.

Gallup has not published a 2026 reading yet. The closest measure this year comes from Reuters/Ipsos, which found in early October that 48% of 2024 Harris voters were very enthusiastic about voting in the midterms, against 30% of Trump voters. That is a different question asked of different groups (very enthusiastic, by 2024 presidential vote, rather than more enthusiastic than usual, by party), so it is separate context, not the next row of the Gallup table.

07Five Reasons a Record Early Vote Is Not a Record Turnout

1. People are moving their vote earlierNearly 244,000 Georgians who voted on Election Day in 2022 have already voted this year. Early-vote growth partly comes out of the Election Day count.
2. More states make it easyIn 2026, 37 states and DC offer both early in-person voting and mail voting to everyone, according to the Center for Election Innovation and Research. Michigan's early in-person voting did not exist in 2022.
3. The calendar movedElection Day is November 3 this year and was November 8 in 2022. Comparing by calendar date gives 2026 five extra days of voting.
4. There are more people to voteFast-growing states have added registered voters since 2022, so more ballots does not automatically mean a higher share of people voting.
5. Party is not a voteRepublicans lean toward voting on Election Day, unaffiliated voters break late, and registration does not show who anyone voted for. Analysts at Split Ticket put it plainly in 2024: persuasion can never be learned from the early vote.

None of that means the surge is fake. It means the early vote tells you who has voted so far, and how they chose to vote, but not who they voted for or how many people will vote in total. The final turnout number comes after November 3.

08The Countdown Log

For the money behind these races, see the most expensive House and Senate races of 2026. For how often the polls have been wrong in midterms, see the biggest polling misses in midterm history, and for the economy voters are judging, the midterm economy. Markets have their own pattern around these elections: stocks after midterm elections.

↻Updates

§Sources

• North Carolina State Board of Elections: Daily Absentee Stats Report, 2026 General (through October 26, 2026, with the matching 2022 figures)
• Michigan Department of State: More than 1.39 million Michigan voters have cast their ballot (October 27, 2026)
• University of Florida Election Lab: 2026 Early Voting and US VEP Turnout Rates, version 1.4
• The National Desk: Early voting outpacing last midterm (October 26, 2026)
• Associated Press analysis, via EconoTimes (October 22, 2026)
• Atlanta Journal-Constitution: Georgia early voting nears 2 million; FOX 5 Atlanta: Georgia enters final week of early voting
• IBTimes: US midterms early voting (October 24, 2026), citing the Virginia Public Access Project
• The Hill: Early voting tops 30 million (November 3, 2022); KPAX: More than 21 million Americans have voted early (October 31, 2022)
• Associated Press, via KPRC: Many states deciding control of the Senate are seeing a big boost in early voting (October 21, 2026)
• Early-vote trackers: On Point Politics (October 28, 2026; pace compares each state with its own 2022 count at the same number of days before Election Day) and Call the Map (October 28, 2026, 10:14 a.m. ET; each state matched by day of voting, working day or days before Election Day, as noted per state)
• US Census Bureau: High Registration and Early Voting in 2022 Midterm Elections
• Center for Election Innovation and Research: Expansion of Voting Before Election Day, 2000-2026
• The American Presidency Project: Seats in Congress Gained/Lost by the President's Party in Mid-Term Elections
• Gallup: Midterm Seat Loss Averages 37 for Unpopular Presidents (2018, including approval at each midterm) and voting enthusiasm trend (2022)
• Reuters/Ipsos, via The Star (October 8, 2026) and UNI (October 20, 2026), reporting the Reuters/Ipsos poll of October 19
• Thomas G. Hansford and Brad T. Gomez, "Estimating the Electoral Effects of Voter Turnout", American Political Science Review (2010); Daron R. Shaw and John R. Petrocik, The Turnout Myth (Oxford University Press, 2020); James DeNardo, "Turnout and the Vote: The Joke's on the Democrats", American Political Science Review (1980)
• Split Ticket: Understanding the 2024 Early Vote (November 1, 2024)