Skip to main content
ABC News
No Sitting President Was As Far Behind As Trump Going Into The Conventions

Welcome to the FiveThirtyEight Election Update for Monday, August 24! As of 5 p.m. on Sunday, Joe Biden had a 73 in 100 chance of winning the 2020 presidential election, while President Trump had a 27 in 100 chance. Biden also has a hefty lead in our forecast of the national popular vote: 53 percent to 46 percent.

This isn’t all that different from when we launched the forecast earlier in August, as Biden’s lead has been steady for months. But with the Republican National Convention starting today — and the Democratic National Convention having ended last Thursday — we have entered a traditionally volatile period for presidential campaigns. So we wanted to pause and take stock of where previous races stood ahead of the conventions.

We don’t have retroactive data on what our forecast — with all its bells and whistles — would have said in past elections, but we have calculated what our polling averages would have said just before the first party convention. (You can download this data on our GitHub page if you want to play around with it yourself.) And based only on the polls,1 Biden led nationally by an average of 8.0 percentage points last Sunday (the day before the Democratic convention began). That was the second-biggest lead for any Democratic candidate heading into the convention period since at least 1968, and the largest for any candidate since 1996, when Bill Clinton led by about 15 points.

Biden’s pre-convention lead was the biggest in two decades

What the FiveThirtyEight nationapolling average would have said the day before the first national convention vs. the final election margin, for presidential elections since 1968

Election 1st convention Poll Margin November Margin Difference
1968 R R+4.1 R+0.7 D+3.4
1972 D R+19.6 R+23.2 R+3.6
1976 D D+7.0 D+2.1 R+4.9
1980 R R+4.3 R+9.7 R+5.4
1984 D R+12.2 R+18.2 R+6.1
1988 D D+1.7 R+7.7 R+9.4
1992 D R+2.7 D+5.6 D+8.2
1996 R D+15.3 D+8.5 R+6.8
2000 R R+7.5 D+0.5 D+8.0
2004 D D+1.3 R+2.5 R+3.8
2008 D D+1.5 D+7.3 D+5.7
2012 R D+0.0 D+3.9 D+3.9
2016 R D+2.3 D+2.1 R+0.2
2020 D D+8.0

Sources: Polls, Dave Leip’s Atlas of U.S. Presidential Elections

Undoubtedly, it’s a bullish sign for Biden to be this far ahead of Trump. In fact, since 1968, no incumbent president has trailed by as much as Trump heading into the first convention. However, before we get too carried away, the size of Biden’s pre-convention lead is unlikely to hold. After all, two other presidential contenders led by margins similar to Biden’s: Jimmy Carter in 1976 and George W. Bush in 2000, but by November their leads had all but evaporated. Carter only narrowly beat Gerald Ford by about 2 points nationally, while Bush won the most highly contested election in modern times, which took the U.S. Supreme Court to sort out.

Since 1968, the final national margin has differed from the pre-convention polling margin by an average of about 5 points. So if we were to apply that to the 2020 electoral environment, we’d be talking about only a 3-point national edge for Biden. That, of course, would increase the possibility that Trump could win the Electoral College even while losing the national popular vote.

One thing we’re watching closely as the GOP convention begins is just how much of a bounce Biden and Trump receive from the conventions. Given the unprecedented virtual convention format, it’s hard to know how much of a bounce Trump or Biden will receive, especially considering it’s gotten smaller in recent years and the conventions are scheduled back-to-back. And we’ve seen little change in Biden’s polling average — he’s averaged about 51 percent since last Sunday — and his lead has barely ticked up, going from 8 points to 9 points. For his part, Trump’s standing in the polls also hasn’t moved very much — he’s slid ever so slightly, from 43 percent to 42 percent — but of course, we haven’t gotten that many polls yet either.

Bottom line: Ahead of the Democratic convention, Biden had the largest pre-convention lead in two decades, but past cycles show that the electoral environment can shift quite a bit from the start of the convention period to November. In other words, this still gives Trump about two and a half months to narrow the gap, although because of the uncertainty surrounding the polls during the convention period, it might be a little bit until we have a firmer grasp of where the race stands.



  1. Our forecast takes into account myriad other factors, such as the election’s “fundamentals” — economic conditions, demographic data, etc.

Geoffrey Skelley is a senior elections analyst at FiveThirtyEight.

Nathaniel Rakich is a senior editor and senior elections analyst at FiveThirtyEight.