In Michigan, the most notable result comes from long-time Michigan pollster EPIC/MRA, which found Biden up by 12 points in a poll conducted between late May and early June, 53 percent to 41 percent. For context, in 2016 the pollster never found Hillary Clinton with more than 47 percent in a single general election poll. However, in Wisconsin, Biden’s lead might not be that secure. The Marquette Law School Poll — often seen as the gold standard in the Badger State — had Biden up only 3 points in early May, 46 percent to 43 percent. Biden and Trump also ran about even in Pennsylvania, another pivotal state in 2016, which might seem at odds with the data from Michigan and Wisconsin. But unlike those states, Pennsylvania didn’t have any surveys from highly rated pollsters, and what they found was a bit noisy — two put Trump up by 4 to 5 points, while another put Biden ahead by 9 points. We wouldn’t read too much into this yet without higher quality polling, but the inconsistent results here could be evidence that Pennsylvania will remain competitive.
As for the polling picture in the Sun Belt states — Arizona, Georgia and Texas — they all seemed more or less in line with what you would expect, once you account for Biden’s lead in the national polls and how these states voted in 2016. But they do signal potential trouble for Trump. For instance, the fact that Trump carried Arizona by 3.5 points in 2016 seems to have been erased by Biden’s polling lead. On average, Biden led by 3 points, including a high-quality early June survey from Fox News that showed him up 4 points. In Georgia and Texas, on the other hand, Trump was still in the lead, by 1 and 2 points, respectively. Yet this is not as cushy of a margin as one would expect for Trump, considering he carried Georgia by 5 points and Texas by 9 points in 2016. If this trio of states are all in play — and Arizona is possibly even leaning Democratic — that would give Biden many additional paths to 270 electoral votes.