Skip to main content
What Absentee Voting Looked Like In All 50 States

We may have seen it coming, but now we know for sure: The coronavirus pandemic made the 2020 election look different from any other election in recent memory. Due to the massive expansion of mail voting, a staggering number of Americans cast their ballots before Election Day. And due to then-President Donald Trump’s false claims that mail voting would lead to election fraud, a huge partisan gap emerged between ballots cast by mail and ballots cast on Election Day.

First, the share of voters casting mail ballots far exceeded that of any other recent national election, and the share of voters who reported going to a polling place on Election Day dropped to its lowest point in at least 30 years. According to preliminary findings from the 2020 Survey on the Performance of American Elections, a poll of 18,200 registered voters run by MIT political scientist Charles Stewart III, 46 percent of 2020 voters voted by mail or absentee — up from 21 percent in 2016, which at the time was considered high. Only 28 percent of people reported voting on Election Day — less than half of the 60 percent who did so in 2016. In-person early voting also reached a modern high (26 percent), although the change from 2016 (when it was 19 percent) was less dramatic.

The shifts took place all across the country, too. According to the SPAE, 47 states and the District of Columbia saw their rates of mail voting rise from 2016 to 2020. The only exceptions were the three states that have held predominantly mail elections for years: Colorado, Oregon and Washington. And, perhaps unsurprisingly, the biggest spikes in mail voting occurred in places that went the furthest to encourage mail voting (i.e., those that automatically sent every registered voter a ballot), especially those with little history of mail voting prior to 2020. These include New Jersey (where only 7 percent of voters voted by mail in 2016, but 86 percent did so in 2020), the District of Columbia (12 percent in 2016 versus 70 percent in 2020) and Vermont (17 percent in 2016 versus 72 percent in 2020).

By contrast, the five states that clung to the requirement that voters provide a non-pandemic-related excuse in order to vote by mail (Indiana, Louisiana, Mississippi, Tennessee and Texas) saw some of the smallest increases. For example, Texas’s rate of mail voting in 2020 was only 11 percent (barely changed from 7 percent in 2016), while Mississippi’s was only 10 percent (just a tad higher than the 4 percent in 2016).

The week of the election already gave us a vivid illustration of how blue absentee votes were and how red Election Day votes were. (You’ll recall that initial results from states that counted absentee votes first, such as North Carolina, were overly rosy for President Biden, and states that counted in-person votes first, such as Pennsylvania, were misleadingly favorable for Trump.) But over the past month, FiveThirtyEight has collected data on the partisanship of absentee and Election Day votes from state election officials — and the numbers are striking.

We have data for only 15 of the 50 states,1 but it tells a consistent story: Biden won the absentee vote in 14 out of the 15 states (all but Texas), and Trump won the Election Day vote in 14 out of the 15 as well (all but Connecticut).2

Indeed, Trump won the in-person vote even in deep-blue states like Hawaii (by 71 percent to 27 percent).3 He even won the Election Day vote in Biden’s home state of Delaware, though it was extremely close there (49.25 percent for Trump versus 49.19 percent for Biden). Conversely, Biden won the absentee vote even in reliably red states like Arkansas (61 percent to 37 percent) and South Carolina (60 percent to 39 percent). If we had data for all 50 states, we would likely see Trump winning the Election Day vote in almost all of them and Biden winning the absentee vote in almost all of them.

Absentee votes broke blue, Election Day votes red

How absentee and Election Day votes in the 2020 presidential election broke down by candidate in the 15 states tracking results by voting method

Absentee Election Day
State Biden Trump Margin Biden Trump Margin Gap
Pennsylvania 76% 23% D+54 34% 65% R+32 85pt
Maryland 81 17 D+65 39 57 R+18 83
Hawaii† 66 32 D+33 27 71 R+44 77
North Carolina 70 28 D+42 33 65 R+32 75
Rhode Island 79 19 D+60 44 54 R+10 70
Arkansas 61 37 D+24 26 70 R+43 67
Oklahoma 58 40 D+18 26 72 R+46 65
Delaware 79 20 D+59 49 49 EVEN 59
Iowa 57 41 D+16 27 70 R+43 59
South Carolina 60 39 D+21 31 67 R+35 57
Connecticut 77 22 D+56 49 49 EVEN 55
Alaska 58 39 D+19 30 66 R+36 54
Georgia 65 34 D+30 38 60 R+23 53
Arizona* 52 47 D+5 32 66 R+34 38
Texas* 48 51 R+3 39 59 R+20 17

*Arizona and Texas do not distinguish between mail votes and in-person early votes.
†Hawaii does not distinguish between Election Day votes and in-person early votes.

Source: State election officials

At the very least, the magnitude of this divide would have shocked anyone looking at the same data for 2016. Of these 15 states, 114 also broke down the results of the 2016 presidential election by voting method. And although absentee votes in 2016 were consistently more Democratic than Election Day votes (just as in 2020), the average gap between them was much smaller than in 2020 — just 14 points in 2016 compared with 65 points in 2020.

In 2016, there wasn’t much of a gap in how people voted

How absentee and Election Day votes in the 2016 presidential election broke down by candidate in the 11 states tracking results by voting method

Absentee Election Day
State Clinton Trump Margin Clinton Trump Margin Gap
Iowa 52% 42% D+10 35% 57% R+22 32pt
Maryland 69 25 D+44 54 39 D+15 30
South Carolina 48 49 EVEN 38 57 R+19 18
North Carolina 46 49 R+3 39 55 R+16 13
Rhode Island 60 33 D+27 54 39 D+14 12
Hawaii 66 27 D+38 59 33 D+26 12
Oklahoma 34 60 R+26 28 66 R+38 12
Arkansas 39 56 R+17 32 61 R+29 12
Delaware 56 40 D+16 53 42 D+11 5
Georgia 47 49 R+2 45 51 R+6 4
Alaska 37 52 R+15 35 52 R+17 2

Source: State election officials

Put another way: In 2016, several states had negligible differences between absentee and Election Day votes, but in 2020, even the smallest differences were gaping chasms. For example, in Alaska (where in 2016 Trump won absentees by 15 points and Election Day votes by 17), absentee votes in 2020 were Biden+19 and Election Day votes were Trump+36. And in Georgia (where in 2016 Trump won absentees by 2 points and Election Day votes by 6), absentees in 2020 were Biden+30 and Election Day votes were Trump+23.

It’s not hard to see why Trump, then, in his desperation to hold onto power, claimed that Democrats used mail ballots to steal the election from him. Biden indeed would not have won without mail votes, but there is no evidence that a significant number of these votes were cast fraudulently. Rather, the increase in their use was a response to the pandemic — one that was even encouraged by most election officials — and the fact that these votes were so Democratic is very likely due to Trump himself. By casting doubt on the security of mail ballots, he all but ensured that most of his voters would cast their votes using traditional methods, leaving the pool of absentee ballots strikingly — but not surprisingly — blue. (Paired with Republicans’ legal efforts to throw out entire batches of absentee ballots, this may even have been a deliberate strategy to improve Trump’s chances by disenfranchising Democratic voters.)

As a result, it will be interesting to see whether these sudden changes surrounding mail voting represent a new normal or prove to be just a blip in history. Some states are thinking about making their expansions of vote-by-mail permanent, while other states have shown little interest — still others are even considering bills to restrict absentee voting. But given that mail voting can make campaign operatives’ lives easier, we might expect more Republicans to embrace it now that Trump is no longer president; then again, that may depend on how much influence he wields in the GOP going forward. According to the SPAE, 65 percent of Democrats and 49 percent of Republicans who voted by mail in 2020 said they were “very likely” to vote by mail again (though, of course, it’s not fully up to them). So, perhaps mail voting will maintain some popularity among members of both parties, but with an even wider division between them.


  1. Alaska, Arizona, Arkansas, Connecticut, Delaware, Georgia, Hawaii, Iowa, Maryland, North Carolina, Oklahoma, Pennsylvania, Rhode Island, South Carolina and Texas. The remaining 35 states plus Washington, D.C., either do not break down statewide election results by voting method or did not respond to our data requests.

  2. Frankly, it could easily have been 15 out of 15 for both. Trump’s victory among “absentee” votes in Texas is misleading, since Texas doesn’t distinguish mail votes from in-person early votes; it’s very likely that Trump won this combined early/absentee-vote category only because so few of the ballots in that category were actually cast by mail (about one-thirteenth, according to the SPAE). Meanwhile, Biden won Election Day voters in Connecticut by a very small margin: 49.23 percent to 49.13 percent.

  3. We say “in-person vote” as opposed to “Election Day vote” because Hawaii’s 2020 data uniquely lumps together in-person early votes and Election Day votes.

  4. Alaska, Arkansas, Delaware, Georgia, Hawaii, Iowa, Maryland, North Carolina, Oklahoma, Rhode Island and South Carolina.

Nathaniel Rakich is an elections analyst at FiveThirtyEight.

Jasmine Mithani is a visual journalist for FiveThirtyEight.