Biden Is Projected To Be The President-Elect. Here’s How It All Went Down.

Exit polls are usually a key part of election night reporting, especially before actual results come in. They can provide an early sense of who is leading and how different demographics are turning out and voting. But while it’s always a little dangerous to rely too heavily on exit polls on election night, particularly the early waves, which are released while polls are still open and generally are not representative of the electorate (skew toward older voters as many young votes vote late). But pandemic-related changes to the election mean that exit polls are even less reliable than usual this year.

That’s because the pandemic has undermined the major advantage exit polls have over other kinds of polls: Their ability to know they’re sampling actual voters. Because of how many people voted early this year, exit polls will include a phone poll component to reach these voters, and in some states there will also be exit polls of early in-person voters, but this has pretty severe limitations for two reasons. First, the phone poll will have to guess whether respondents who said they voted actually did, and second, Edison Research — the polling firm that produces the exit polls used by ABC News, CBS News, CNN and NBC News — will have to estimate how heavily to weight the traditional exit poll respondents vs. the phone respondents and early voters. That means there’s more uncertainty than usual in the election night exit polls. And that extra uncertainty means that it’s hard to make a case for relying on them tonight — though they’ll still prove useful in telling us about how key demographics voted once they’re weighted after the election to match actual results.

One barrier to interpreting early results in many states is not knowing just how many counted ballots are from early voting, but one thing we should be able to tell is whether any Senate candidates are performing substantially better than their party’s presidential candidate among the same voters. There has been almost no gap in the polls for Senate and President in many states, as those candidates have had trouble separating themselves from their party’s national reputation. Republicans would likely need a candidate-specific Senate vote in Maine and Michigan; Democrats would likely need one in Montana or Kansas. In tight races like North Carolina, even a small advantage for one party’s Senate candidate over their Presidential nominee could be enough for victory.

For the first time in U.S. history, a voter is more likely to have voted before Election Day than on the day itself. That’s the takeaway from the massive early voting numbers we’re seeing around the country, as around 100 million people have already voted, according to the U.S. Elections Project. But let’s try to put that figure in perspective. In 2016, close to 40 percent of 137 million voters cast early ballots in some way, whether by mail or in person — the highest share ever, as the chart below shows.

Yet the FiveThirtyEight presidential forecast estimates that anywhere from 147 million to 168 million people may vote in the 2020 election. So with 100 million or so votes already cast, that means that an overwhelming majority of voters will have voted before Election Day this year.

And the surge in early voting has been driven largely by Democrats, as somewhere around three-fourths of them planned to or already have voted before Election Day, whether by mail or in person, based on recent national polls. By comparison, only a little more than half of Republicans have said the same. In addition, polls suggest that early voting is up across many racial and ethnic groups, but most especially among Black voters, who lean heavily Democratic.