Welcome to FiveThirtyEight’s Election Update for Wednesday, Sept. 9. And at this point, it’s no longer too early to look at the polls. We’re moving away from the convention period of the race and into the thick of the election. However, the overall race hasn’t changed all that much. Joe Biden has a 74 in 100 chance of winning compared to Trump who has a 26 in 100 shot, according to our forecast, and as you can see in the chart below, those odds have been pretty consistent.
But even though the topline numbers haven’t changed all that much, that doesn’t mean there hasn’t been some movement at the state level. It’s not super easy to find patterns where Biden is gaining or losing ground — still a lot of noise at this point — but it does seem as if, on the whole, there’s been mostly good news for Biden. Except for Florida, which, as you can see in the chart below, is where Trump has closed the gap the most.
Polling shifts in battleground states are pretty noisy
Change in Joe Biden’s margin vs. Donald Trump’s margin in battleground states, according to FiveThirtyEight polling averages from Aug. 17 to Sept. 9, as of 2:30 p.m. on Sept. 9
|Biden margin vs. Trump|
|State||Aug. 17||Sept. 9||Change|
And at this point, though, it’s really only Florida that’s moved all that much. In most other states which have been polled about as much as Florida — there’ve been 12 polls there since the start of the Democratic convention — they’ve seen far less of a change or barely any movement at all. Four other states have at least 10 polls in the same timeframe — Pennsylvania (17 polls), Wisconsin (12), Michigan (10) and North Carolina (10) — yet their numbers have budged only a little.1 What’s more, it’s mostly good news for Biden at this point.
It’s true that Biden is polling better in a state like Wisconsin than in his home state of Pennsylvania, which has caused some confusion given the states’ demographic similarities, but it’s important that we not read too much into it at this point.
Case in point, despite losing ground in Pennsylvania, Biden got a great poll there today from NBC News/Marist College, which found him with a 9-point lead among likely voters, whereas a poll from Marquette Law School poll of Wisconsin was more middling, putting Biden ahead by just 4 points, and actually bring down his average just slightly there.
It’s a little surprising that these states haven’t necessarily moved more in tandem as the results in states with similar demographics are often correlated. But it’s an important reminder that sometimes individual polls look like they’re pointing in opposite directions, but with enough data, many of the so-called movements even out in the end, and at this point, it’s still mostly good news for Biden aside from Florida.
As for what’s happening in Florida, we’re still a little unsure of what to make of it, as we’d expect the numbers in fellow Sun Belt states Arizona and Texas to have moved in similar ways — but they haven’t. Some of this might have to do with Florida’s Hispanic population, however. Whereas Mexican Americans comprise more than 80 percent of the Hispanic population in Arizona and Texas, Cuban Americans form a plurality — 29 percent — of Florida Hispanics, and on the whole they are much more conservative.
Recent polling suggests this could be a critical difference, too. For example, Democratic pollster GQR’s new Florida survey found Biden ahead 51 percent to 46 percent overall, but Trump led by 18 points among Cuban Americans while trailing by 17 points among non-Cuban Hispanics. And a Miami Herald/Bendixen & Amandi survey of just Miami-Dade County, home to two-thirds of the Florida’s Cuban population, found them breaking for Trump 68 percent to 30 percent. It’s possible that in the wake of a GOP convention that painted Democrats as socialists and even featured a Cuban American speaker, Cuban Americans responded by becoming more supportive of the president. It also helps that Florida has been trending slightly to the right in presidential elections for a while now.
But as with most questions involving polls, we need more to get a better read on just how much is real, lasting change. We’re still in Trump’s post-convention bounce period, too, so some of Trump’s gains could wear off in Florida and elsewhere in the coming weeks. Moreover, the lack of movement in FiveThirtyEight’s forecast is one reason to not make too much hay of the change in Florida — at least, not yet.