Skip to main content
ABC News
Trump May Have Gotten A Convention Bounce. But It’s Very Slight And May Already Be Fading.

We’ve now seen an array of new national polls conducted in the aftermath of the Republican National Convention, and collectively they show nothing major has shifted in the race between President Trump and Joe Biden. Trump may have enjoyed a very slight convention bounce, but one that represents only a small boost compared to where the race stood before the party conventions — and even that small bounce is showing some signs of fading.

FiveThirtyEight’s general election forecast tells the tale. On Aug. 16, the Sunday before the Democratic convention, Trump had a 28 in 100 shot at winning. A little over a week later, as the GOP convention was taking place, Trump’s chances ticked up to 32 in 100, as there wasn’t much evidence of Biden getting any post-DNC bounce. And as of the afternoon of Wednesday, Sept. 2, Trump has a 30 in 100 chance. All in all, there’s been only a very small shift in Trump’s direction.

Those forecasted odds are derived mostly from the polling data being released. So let’s take a closer look at that data; we’ll separate the polls into batches according to their methodologies — telephone surveys, online polls and tracking polls.

First, four live-telephone pollsters1 released surveys on Wednesday that offered only modestly positive news for the president. Quinnipiac University found Biden ahead by 10 points, 52 percent to 42 percent, in a poll conducted after both conventions, which marked an improvement over Trump’s 15-point deficit in the pollster’s mid-July survey (the new poll is among likely voters, while the earlier one was of registered voters — the former tend to be a tad more Republican-leaning than the latter). Meanwhile, USA Today/Suffolk University gave Biden a 7-point edge, 50 percent to 43 percent, smaller than the 12-point lead Biden had in late June. Grinnell College/Selzer & Co. also released a new survey, finding Biden ahead by 8 points, 49 percent to 41 percent, although it hadn’t tested the race since late March when it gave Biden just a 4-point lead over Trump. Lastly, CNN/SSRS’s post-convention survey actually showed a marginal improvement for Biden — he led by 8 points, 51 percent to 43 percent, up from his 4-point lead in mid-August.

However, pollsters that conduct surveys online haven’t shown much movement in their national numbers. And the online pollsters have tended to poll more often than live-phone outfits, giving us more recent benchmarks to judge the conventions’ effects (mostly pre-convention polls conducted just prior to the Democratic event). Ipsos showed a marginal improvement for Trump, who went from trailing by 11 points in mid-August to 7 points after the conventions. But others found no real change. Canadian pollster Léger found Biden’s lead going from 8 points in early August to 7 points after the GOP convention, while Morning Consult found no change in voter sentiment in three-day polls before and after the convention period, giving Biden an 8-point edge in both surveys. Meanwhile, three pollsters showed a very slight increase in Biden’s margin: YouGov/The Economist, IBD/TIPP and Redfield & Wilton all found Biden’s margin actually increased by 1 point. Lastly, Emerson College partially conducts its surveys online, and its latest poll found Trump down by only 3 points after he trailed by 6.5 points at the end of July.2

Finally, tracking polls — which conduct daily interviews over a rolling period of time — have been a mixed bag. In its first 7-day tracking poll ending on Aug. 17, leading into the Democratic convention, USC Dornsife found Biden up by 10 points based on responses to its probabilistic voting questions, and Biden’s lead expanded to 14 points following the Democratic convention. But as of Sept. 1, with the GOP convention days mostly out of this weekly tracker, Biden led by 9 points. Morning Consult also conducted a handful of 1-day polls in and around the convention period, and the pollster found that Trump received a small bounce the day after his convention; though, as we mentioned earlier, the pollster didn’t find any improvement for Trump in its lengthier polls.

Of course, these national surveys tell only part of the story, as the Electoral College will be decided by the vote in each state. New state-level surveys are coming in, as well, and we plan to have more to say about those tomorrow. We’ll be looking for whether Trump’s slight bounce is enduring or fading (or was even there to begin with). But for now, it looks like the conventions didn’t fundamentally alter the trajectory of the campaign.


  1. That is, pollsters that use actual people to call respondents’ on cell phones and landlines.

  2. Note that Emerson College released two results for the Trump-Biden race. The poll included an initial vote choice question and a follow-up one that forced respondents who said they were undecided to choose between Trump and Biden. We’re using that second one here. The initial question, with undecided voters, found Biden leading by 2 points.

Geoffrey Skelley is a senior elections analyst at FiveThirtyEight.