2020 Election: Live Results And Coverage

One justification Trump is giving for his confidence going into today are his campaign rallies. Recent research suggests that these rallies produce a short-lived increase in support for Trump. At the same time, research also suggests that these same rallies have led to more COVID-19 cases (which could depress turnout for Trump). We should take caution in over-interpreting these findings since they have not yet undergone peer review yet.

You will see a lot of election maps in the next few days, and cartograms are becoming more and more popular to show differences in scale. As CityLab explained today, rural counties represent a lot of land, but urban counties are more densely populated, which is why choropleth maps might be misleading. Land doesn’t vote! The Wall Street Journal is tracking results using both types of maps, including the cartogram shown here, although Bloomberg’s bodies of water serve as better visual cues.

In the past four years, political observers have often talked about three states — Michigan, Pennsylvania and Wisconsin — in a single breath. They were the three so-called Blue Wall states that had traditionally backed Democratic presidential candidates but put Trump over the top in 2016. But now, polls — and FiveThirtyEight’s final forecast — suggest that Trump’s chances may be better in Pennsylvania than in Wisconsin or Michigan.

Why is that?

It’s actually not all that clear. Take election evening 2018. Pennsylvania reelected a Democratic senator and governor by double digits, while Wisconsin’s Republican Gov. Scott Walker nearly won a third term. Pennsylvania also has, in many ways, an electorate that would seem to be friendlier to Democrats, given that it has more college-educated voters than Wisconsin. So before 2020, it seemed that if Wisconsin — not Pennsylvania — was a little more likely to lean GOP this year.

But is Pennsylvania the more competitive state tonight? That’s something I’ll be watching closely. One possibility to keep in mind is that Pennsylvania is a state with party registration, while Wisconsin (and Michigan) are not. That might make those other states harder to poll — and so might make polling in Pennsylvania more accurate.

Pinellas County in Florida, home to Tampa, looks promising for Biden so far. He leads by about 37,000 votes with 75 percent reported. That number so far includes very little Election Day vote, which will help Republicans. But Republicans led the Election Day vote by around 21,000. So if the vote matches party registration — which it may not! — Biden would win a county that Clinton lost by 1 point in 2016.

To Nate’s point about ABC’s being conservative about projections tonight: There’s a long history here, one that’s rooted in election night 2000. The TV networks all competed that night to get the scoop, to call the race early, and they all later had to retract their calls. As CBS’s Dan Rather said that night on air, “If you’re disgusted with us, frankly, I don’t blame you.” So, yeah, networks have a responsibility to be cautious! They learned their lesson the hard way.

In response to Lee’s earlier question about whether the Democrats might be able to persuasively claim an election mandate for reforms to the system if they win big tonight, the answer is, well, it’s complicated. Let’s start with the obvious — it’s hard to expect a party coming off of a big victory to invest its energy in, or rally its supporters around, changes to the rules they just won under.

On the other hand, the reforms in H.R. 1 (that Democratic-sponsored House bill to expand voting rights) are likely to be quite popular — even beyond Democrats. Americans want to see campaign spending reined in, and their growing sense of dissatisfaction with the political system might make the public more amenable to structural improvements in our democracy.

That said, it’s worth noting that when we have seen major expansions of enfranchisement and democracy, as with the 19th amendment or the Voting Rights Act, these changes came about because of decades of work by activists, with elected officials coming in toward the end of the process. I say that because Democratic officeholders might want to be careful in claiming a mandate for such changes and taking credit for them.

ABC News projects Republican wins in KY-2, KY-5 and VA-9.

Note that ABC News is being more conservative with its projections than some other news organizations, which I think is a good move. You just want to be doubly sure you’re getting everything right when people are voting by all these different means.

Well, I know there was a lot of doubt about this race, but Indiana Republican Gov. Eric Holcomb is projected to win. (There wasn’t any doubt.)

ABC News projects a Republican win in IN-3.

One warning … don’t make too many inferences about which states get called when polls close and which do not. A lot of it has to do with which states have exit polls, because the networks will be more cautious in states that don’t have them.

It’s now 7 p.m. Eastern, which means the last polls have closed in Georgia, Indiana, Kentucky, South Carolina, Vermont and Virginia. Here’s what our forecast says about the races in those states.

With polls closing, networks are starting to talk about making projections. I found this piece about the changes The Associated Press is making this year really fascinating — they are going to be more transparent about how they make these decisions and why they might hold back on some calls. These changes are partly in response to the complicated environment in 2020, but they also seem to reflect longer-term shifts — the evolving media environment has made people expect more transparency, not just a few authoritative networks making calls, and polarization has increased attention to these processes as well.

Hey! It’s me, Fivey! Your friendly neighborhood data fox!

Tonight, whenever ABC News’s Decision Desk projects a winner in a presidential, Senate or House race, I’m going to be sharing those projections on our live blog. It’s going to keep me busy: I’ll be checking for new presidential and Senate projections every minute, and for new House projections every five minutes. (The Decision Desk “projects” a winner in a particular contest when enough of the vote has been counted in enough of the state or district to be as certain as possible about who will win — for more about how that process works, listen to this episode of the FiveThirtyEight Politics Podcast.)

But there’s something else you should know about how I report projections. I’ll let you in on a little secret: I’m a bit of an emotional fox — at least as it concerns probability. That’s why I use FiveThirtyEight’s final forecast to contextualize the results — is this what we anticipated or were we surprised? And if we were surprised, based on our final forecast, I might react as such. For example:

• If our final forecast listed a candidate as “slightly favored,” meaning that the forecast gave them somewhere between a 55 and 70 percent chance of winning, but one of their opponents wins, I will be surprised (😮) when I share the projection.
• If our final forecast listed a candidate as “favored,” giving them between a 70 and 90 percent chance of winning, but one of their opponents is projected as winning, I will be shocked (😲).
• And finally, if our final forecast listed a candidate as either “clearly favored” or “very likely” to win, meaning they had over a 90 percent chance of winning, but they still are projected to lose, my little fox mind will be completely blown (🤯).

Now if the election goes as my human data-journalist counterparts expected, based on the forecast, then you may not see many emojis from me. But if the night takes a completely different turn, you can bet your bushy tail I’m going to have a hard time containing my reactions.

To Amelia’s point about young people, the last time we saw really notable youth turnout was in 2008 when young people turned out in droves for Obama. I think one of the most fascinating things about the past four years is how much politics has permeated pop culture. Gen Z (the TikTok generation) seems particularly tuned in (says an Elder Millennial jealously).

Well, one thing we know, Sarah, is that lots of young voters seem to be turning out, especially in battleground states like Texas. And that’s potentially a big deal, since turnout is usually quite low among young people.

Over 100 million Americans voted early this year, which is record-breaking turnout, as this map from The Washington Post helpfully illustrates. What do we know at this point, if anything, about the vote that has already been cast?

Even without full results, we can look at the differential in how the top of the ticket is doing versus down-ballot races. So far in Kentucky, it does look like McGrath, who is challenging McConnell for his Senate seat, is running ahead of Biden. That is not surprising. We expect other Senate candidates like Cornyn in Texas and Kelly in Arizona to run ahead of the top of the ticket. There are still ticket splitters!

Wow, Julia, that is really clever!

Election night data viz will always hold a special spot in my heart, particularly the cartograms. Maps that size states by their electoral votes are notoriously difficult to produce, and you almost always have to make some … strange geographic choices to pull it off (yes, we know that South Dakota is above North Dakota on the one on our House forecast. But I am especially in love with Bloomberg’s cartogram tonight. They got the Mississippi River in there!

Another thing about democracy being on the ballot (can you tell I’m a bit obsessed?) is that under Trump, the Republican Party has moved to a very illiberal place, at least as judged by global democracy experts, who place the GOP at about the same place as the AfD in Germany, and approaching Viktor Orban’s Fidesz party in Hungary. So given Trump’s long list of attacks on the principles of democracy, the argument that democracy really is on the ballot kind of rings true.

I’m watching CNN at the moment and I am seeing a lot of coverage of partial results. For example, they just highlighted that Biden is leading in Kentucky with 8 percent of the vote in. They are also running through a slew of counties reporting 20 percent of the vote or less. They have heavily caveated these numbers, but this kind of coverage is probably just not necessary. While we’ve talked a lot about the risks of partial results at FiveThirtyEight, most viewers are probably not familiar with all the differences in what types of votes get reported when. These very-partial-vote analyses are ripe for manipulation by partisans as well. We need to hold off on drawing any real conclusions until full counties are reporting. We can indeed look at small counties in Kentucky to give us a sense of the shift from 2016, but in order to do that, we need all the vote in.

The initial results from Kentucky, while unrepresentative, are consistent with a nationalized presidential election. According to the Decision Desk, the minimal early vote results from a few counties show the same partisan vote shares for the presidential and Senate race, suggesting few split-ticket voters. All those Democratic donors trying to defeat McConnell yielded a lot of advertising, but might not have persuaded many Trump voters to support McGrath.

Last week on the FiveThirtyEight Politics podcast, Galen Druke interviewed Dan Merkle, the executive director of elections at ABC News. They talked about how the ABC News Decision Desk will be projecting the winner of races this year, including the presidential election in individual states and the Electoral College.

It looks as though Republicans will wind up with around a net +2 percent on party registration in Florida. That’s not terrible for them, and they reversed a slight Democratic disadvantage based on early and absentee voting. But it’s also not great. Based on the partisan splits in recent polls of Florida, I estimated that — because Biden is doing better among independents in most polls of the state and gets somewhat more crossover voters — the breakeven point for a Trump win was about R +3.5 in Florida. That is, if Republicans led in turnout by 3.5 points or more, Trump would be the favorite to win; otherwise Biden would be. We’ll probably end up just to the Biden side of the line. With that said, there’s a lot of uncertainty. Some polls in Florida showed Biden with big leaders among independents in Florida, while in others he was running evenly with Trump.

While we’re waiting for results … on today’s episode of Hot Takedown — FiveThirtyEight’s sports podcast — we discussed a great collaboration we did with ESPN on political donations by sports team owners. I was surprised by how some owners spend their political money.

Matt, one thing political scientists do know is that when elections are competitive and high-stakes, more people vote because they feel like it matters. So driving us to existential angst is certainly one way to increase turnout. But even if we get to 65 percent turnout this year (conceivable), which would be the highest since 1908, we’d still be in the middle of the pack among advanced democracies, many of which regularly have turnout in the 70 or even 80 percent range.

Our politics editor, Sarah Frostenson, helps us set the scene for tonight as we get closer to results.

As I mentioned earlier today on the live blog, I’m anticipating a big gender gap in the vote this year. But it’s not simply Trump’s candidacy that is driving diverging party support between men and women (although it probably explains some of the gap).

A recently published paper by several political scientists suggests that the growing divide between men and women in their vote choice is due to factors that extend beyond each election cycle. They argue that differences in policy preferences have existed for decades (e.g., support varies between men and women on issues like social welfare and criminal justice), but only more recently has issue-based polarization at the elite level (like among members of Congress) more clearly crystalized.

So voters are simply better sorting into the parties that represent their policy interest than they used to be. And according to this paper, the gender gap is widest among more educated Americans, who are more aware of their elite status.

We just got a big dump of votes from the Lexington, Kentucky, area. However, they are probably early and absentee votes, which as we know skew Democratic. So I’m not sure this tells us much.

Political scientists and popular analysts used to talk a lot about cultural explanations for low voter turnout in the 1990s and early 2000s. But it’s now looking like polarization and the high stakes of today’s campaigns are driving high turnout. We have not revived league bowling or increased social trust, but we are still voting at very high levels.

Even though we’ve already told you to basically ignore exit polls … I’d say the results I’ve seen so far would make me even less confident that they’re going to tell us much of predictive value tonight. Based on what’s been publicly reported, there’s some stuff that would seem pretty good for Biden and some that would seem pretty good for Trump, but not in a way that tells an especially coherent story. That’s not meant as a criticism — the people who run the exit polls are smart, but exit polling is really hard this year. I just think we need to wait for actual votes.

Elliott County, in the coal fields of eastern Kentucky, was notable for voting Democratic in every presidential election from 1872 to 2012. But illustrating how white, working-class areas like Appalachia have moved away from the party, Trump carried Elliott by 44 points in 2016. Now, with 47 percent of the expected vote reporting, Trump is leading by 33 points in Elliott, according to The New York Times.

Pew asked in August whether significant changes in the fundamental design and structure of American government were needed. Overall, 62 percent agreed. But that goes up to 79 percent among Democrats, so I think there’s going to be tremendous pressure for Democrats to put forward a pretty big democracy reform agenda if they gain the trifecta. Pew asked the same question in 2018, and then 68 percent of Democrats supported significant changes. Republicans, however, are more divided.

In response to Sarah’s question, I’m going to be looking for signs that North Carolina and Georgia are going one way and Florida is going another. That might give us a sense of what kinds of places are driving the vote totals for each candidate.

In several states that report turnout by party registration, the share of registered voters has been getting bluer over the course of the day. In Broward County, Florida, for instance, Democrats led by 4 points in votes cast between 7 and 1030 a.m., by 8 points in votes cast between 10:30 a.m. and 12:30 p.m., by 11 points in votes cast between 12:30 p.m. and 3 p.m., by 15 points in votes cast between 3 and 4, and by 21 points in votes cast between 4 and 6. Among other things, this may make exit polls even less reliable than usual for predicting the vote, because if you’re only surveying people who have voted by a certain point in the afternoon, you’re not getting a representative sample of Election Day turnout.

In one very specific respect, though, the Trump era seems to have bolstered American democracy: We’re talking about record turnout, two years after a high-turnout midterm election.

We are starting to get the very first results in Indiana. Because they’re very sporadic, though, they don’t tell us much yet.

How could Biden get to 270 without Pennsylvania or Florida? If he were to win North Carolina plus Wisconsin, Michigan and the Clinton states, he’d be there. He could also do it by winning Arizona plus one Electoral College vote from Nebraska. That would be razor-thin, though: 270-268.

Well, it’s longer term than just the election, but I’m going to be watching final in-person voting numbers in the states that are being hardest hit by this third wave of COVID-19. Research suggests that human density at the polls has an impact on the spread of disease, so I’d expect to see an uptick in cases a couple of weeks from now in places that had the most dense voter turnout — especially if you have dense voter turnout and already high rates of COVID-19 spread.

I just want to reemphasize for people that initial results could be misleading in many states. For example, one of the bluest counties in Florida will dump a bunch of (probably very Democratic-leaning) mail ballots right after polls close there at 7 p.m. Eastern. That will make the results in Florida look really good for Biden, but the state will probably undergo a “red shift” as Election Day votes are counted.

OK, team. The first polls don’t close until 7 p.m., meaning that’s the earliest we might get a projection (projection is lingo for “call,” reader, and what we’ll be using to describe races that have been Biden or Trump have won). Let’s walk through a little what we can expect then — and as the night progresses, as in many states, the vote count will shift, depending on what ballots are counted.

We’ve taken a crack at outlining when to expect election results in every state, and expect 17 states to have nearly all of their expected vote counted tonight, but what should we be watching tonight as the vote trickles in?

Lee and Julia just mentioned the argument that “democracy is on the ballot.” In my panel, I’ve been tracking how many Americans agreed that “at present I feel very critical of our political system.” In October 2012, 54 percent of respondents agreed. But last month, that number was up to 70 percent.

We know that mail ballots are far more likely to be cast by Democrats than Republicans, who are more likely to vote on Election Day. This means that in some states — though certainly not all — more Republican-leaning Election Day votes may be counted ahead of many Democratic-leaning mail ballots, and as the latter are reported, they could swing the vote toward Biden, creating what’s known as a “blue shift.”

Take Pennsylvania, where election officials couldn’t begin processing mail ballots until 7 a.m. today. As a result, Trump could very well lead there tonight but lose that edge as the count proceeds. To get some idea about what such a shift might look like, I analyzed Pennsylvania’s June 2 presidential primary, which saw a huge surge in mail balloting that slowed counting. By 3 a.m. on election night (June 3), only 46 percent of the final Democratic vote had been reported, compared to 71 percent of the GOP total, based on data from ABC News.

If we apply the pattern of returns from the primary to data from FiveThirtyEight’s forecast, we can get an idea of what a blue shift could look like. On Oct. 28 — that’s when I originally did this analysis, but not much has changed — our forecast estimated that about 6.8 million votes will be cast in Pennsylvania, on average, and the average popular vote result gives Biden about a 5-point edge over Trump, 52 percent to 47 percent. If that panned out, Biden would win about 3.6 million votes to Trump’s 3.2 million (with a few additional votes for other candidates). But if the count were to mimic the primary’s trend, Trump would have about a 16-point lead at 3 a.m., 58 percent to 42 percent, based on approximately 60 percent of the total expected vote. But over the course of the next few days, Biden would win two-thirds of the remaining votes, which would precipitate a 21-point shift in the overall margin, as the chart below shows.

Now, this analysis has its limits, as it’s impossible to say just how the count will go and how much the vote may shift in Pennsylvania. However, the primary results there suggest that we should be ready for a sizable swing.

And one fear about a blue shift is that Trump could try to declare victory based on initial returns, as he’s claimed the result should be known on election night. A premature declaration of victory by Trump would have no legal bearing, but he could then point to the shift as evidence of fraud to cast doubt on Pennsylvania’s vote, as he’s made false claims about voter fraud numerous times already. But in reality, such a swing would simply be due to a prolonged count brought on by the increase in mail voting — nothing sinister at all. So we shouldn’t be surprised if Trump leads tonight but loses Pennsylvania once a lengthy count is completed after tonight.

Lee and Julia, in thinking about whether “democracy is on the ballot,” it seems important to highlight that maybe it’s really about laboratories of democracy on the ballot. For instance (as Lee pointed out), Massachusetts has ranked-choice voting on the ballot this year. My bet is that we’ll see many more of these sorts of institutional reforms at the state and local level before we see it at the federal level.

Also speaking of democracy being on the ballot this year: In several states, it literally is! Massachusetts and Alaska are deciding whether they want to use ranked-choice voting. Florida and Alaska are deciding whether they want to have open nonpartisan primaries. California is deciding on voting rights for felons on parole. Virginia is voting on independent redistricting. The Fulcrum has a helpful guide to these and more democracy issues on the ballot this year.

If he wins tonight, Biden’s plan for COVID-19 include a nationwide test and trace system, which many experts have said the U.S. should have had up and running months ago. The problem: It may now be too late for such a system to offer meaningful benefits. Test and trace — interviewing people who have tested positive for COVID-19 and tracking down their recent contacts so those people can also be tested or quarantined — works best when there are clear points of contact and people can make educated guesses about where they might have contracted the illness. Thanks to the recent surge, there are many parts of the country where this is no longer possible. Running for president this year was always going to be about how each candidate will deal with the pandemic — but at this point, it’s going to affect what options any potential Biden administration has available to get the disease under control. If he wins tonight, he has some big challenges ahead.

One county in Georgia is keeping polls open for an extra two hours after having technical problems this morning:

It’s a fine line between a landslide and a nail-biter.

As I wrote in our final forecast overview, there wasn’t a lot of change over the past 24 or 48 hours, as most of the late polling either came in close to our previous polling averages, or came from — frankly — fairly random pollsters that don’t get a lot of weight in our forecast, which means the forecast didn’t shift that much.

Our final forecast had Biden with an 89 percent chance of winning the Electoral College, as compared to a 10 percent chance for Trump. (The remaining 1 percent reflects rounding error, plus the chance of an Electoral College tie.)

But the reason I say it’s a fine line between a landslide and a nail-biter, is that it wouldn’t take that big of a polling error in Trump’s favor to make the election interesting.

Importantly, interesting isn’t the same thing as a likely Trump win; instead, the probable result of a 2016-style polling error would be a Biden victory but one that took some time to resolve and which could imperil Democrats’ chances of taking over the Senate. On the flip side, it wouldn’t take much of a polling error in Biden’s favor to turn 2020 into a historic landslide against Trump.

Biden’s standing is considerably stronger than Clinton’s at the end of the 2016 race, but a 10 percent chance of something happening isn’t a zero percent chance, and as I wrote earlier this week, Trump still has a path to the White House.