Welcome to FiveThirtyEight’s Election Update for Thursday, September 3! We’ve now gotten several state polls conducted entirely after the end of the Republican National Convention, and they are consistent with the tale told by post-convention national polls: The RNC hasn’t boosted President Trump enough to meaningfully improve his odds of winning the election. In fact, the state polls we’ve gotten this week have been pretty darn good for Joe Biden — with one big exception that we’ll get to in a minute.
Probably the most eye-opening state poll was Fox News’s survey1 of Arizona, which gave Biden a healthy 9-percentage-point lead. The poll was taken in the immediate wake of the RNC, and although candidates generally poll higher right after their conventions, Fox found Trump in a worse position compared with its previous poll.2 Morning Consult appeared to confirm Biden’s upward trajectory in Arizona, releasing an Aug. 21-30 poll showing Biden up by 10. (That was a big shift from Trump’s 2-point lead there in Morning Consult’s Aug. 7-16 poll.)
Previously, Arizona polling had been a little disappointing for Biden, given that the state appeared primed to move into the spotlight as one of the top 2020 swing states. But these two good polls for Biden helped the Democrat jump out to a clearer lead in our Arizona forecast (although it’s still pretty close to a toss-up):
In addition, Biden now leads Trump by 4.6 points in our weighted average of polls of Arizona. That’s significant because, if you were to line up all 50 states from those where Biden has the biggest polling lead to those where Trump’s lead is biggest, right now Arizona would be the Electoral College “tipping point” — the state that puts one candidate over the top. (That said, our forecast, which blends the polls with election “fundamentals” like the economy to predict where the race will stand in November, maintains that Pennsylvania is the likeliest state to wind up as the tipping point on Election Day.)
Fox News also had good news for Biden in Wisconsin, putting the Democrat ahead by 8 points. And two other recent polls (from Morning Consult and Opinium) staked Biden to similarly large leads — 10 and 13 points, respectively. Notably, all three of these polls were conducted mostly or entirely after police shot Jacob Blake, a Black man, in the Wisconsin city of Kenosha, kicking off protests where two demonstrators were killed (a militia member has been charged with their murder). But contrary to what many commentators have insisted, those tragic events — and the Trump campaign’s focus on the violent fallout and a “law and order” message — don’t appear to be boosting the president’s standing in the state. Biden’s chances of winning it are up to 73 in 100, per our forecast.
The one state where Trump does seem to have emerged from the conventions in better shape is Pennsylvania; Monmouth University now has Trump and Biden in a virtual tie among likely voters there. Biden is 3 percentage points ahead in Monmouth’s high-turnout scenario and 1 point ahead in a low-turnout scenario. In Monmouth’s July 9-13 poll of the Keystone State, Biden led by 10 points in the former scenario and 7 points in the latter. As a result, our forecast shows Trump with a better chance to win Pennsylvania than he’s had in months.
It’s a bit surprising to see Biden polling relatively poorly in Pennsylvania, the state where he was born and one that he references frequently on the campaign trail. According to the polling averages with Monmouth factored in, he currently leads Trump by just 3.4 points there — a smaller lead than in demographically similar Michigan (6.5 points) and Wisconsin (7.1 points).
That said, although Monmouth is an excellent pollster, its post-convention survey may be an outlier. Monmouth’s sample sizes are on the small side, and the company isn’t afraid to publish results that diverge from the consensus (that’s a good thing), so its polling can sometimes be subject to large swings.
And the fact remains that Biden is still more likely than Trump to win Pennsylvania. Indeed, as you’ve seen, he is more likely than Trump to win all three of these crucial states. That’s a big reason why he still has a 69 in 100 chance of winning the Electoral College overall — almost exactly the same position he was in before the conventions.
Make sure to check out FiveThirtyEight’s full presidential forecast; you can also see all the 2020 polls we’ve collected, including national pollsFlorida pollsMichigan pollsMinnesota pollsNorth Carolina pollsTexas polls and … well … all the states, really.