For a while now, the conventional wisdom has been that a Democrat's path to 270 electoral votes in a presidential election would run through the Sun Belt, where the nonwhite part of the electorate is growing rapidly. And this view wasn't just theoretical. Here are the results of the last three presidential elections in Georgia:
2012 Republican +7.8
2016 Republican +5.2
2020 Democrat +0.2
And Arizona:
2012 Republican +9.1
2016 Republican +3.6
2020 Democrat +0.3
And Texas:
2012 Republican +15.8
2016 Republican +9.0
2020 Republican +5.6
You can see how, in a plausible interpretation, Arizona and Georgia are on their way to becoming a fairly dark shade of blue, as has in fact happened with Virginia and Colorado, and that when Texas inevitably joins them it will be game-set-match for the Dems.
But as the political journalist Ron Brownstein argues, here, that isn't what's happening—at least not so far in 2024. Forty-five states either voted for or against Trump in both 2016 and 2020. The other five, which all flipped from Trump in 2016 to Biden in 2020, are Pennsylvania, Michigan, Wisconsin, Georgia, and Arizona. Of these five, the heaviest lifts for Biden appear now to be Georgia and Arizona, while his position is strongest in Pennsylvania and Wisconsin. According to Brownstein and the political strategists he's interviewed, the reason for this is that Biden is slipping precipitously with nonwhites, upon whose votes Dems rely to compete in the Sun Belt. Meanwhile, however, he is doing better than expected with white voters, including older white voters, who are a much bigger share of the electorate in the Rust Belt.
If you're interested in this subject of political demographics and voting patterns in America, the whole article is worth reading. The big takeaway, though, is that there's been something like a trade: Hispanics and African Americans (especially men) have drifted away from the Dems, while support from whites (especially more educated and affluent women) has ticked up. There has been evidence of this in recent election results too. I've always thought there was a racial dog whistle in Trump's kvetching, and worse, about election fraud in 2020 in places like Philadelphia and Detroit. Actually, he did about the same in these heavily minority cities in 2020, when he lost both Pennsylvania and Michigan, as he did in 2016, when he won both states. The erosion in his support occurred in suburban counties and those centered on mid-sized cities, like Grand Rapids, Michigan. You can look these things up. Here are the results of the last two elections in Michigan's Wayne County (Detroit), Oakland County (affluent suburbs northwest of Detroit), and Kent County (Grand Rapids):
Wayne
2016 Trump -37.1
2020 Trump -38.0
Oakland
2016 Trump -8.0
2020 Trump -14.0
Kent
2016 Trump +3.0
2020 Trump -6.1
Sorry about your feelings, MAGA-ites, but the evidence does not bear out the claim that "massive fraud" in Detroit swung Michigan to Biden in 2020. Better keep your eye on shifty white election officials this time around!
Brownstein dances a little lightly over the details of the electoral arithmetic, but I think it works pretty much like this:
A candidate needs 270 electoral votes to win. The 20 states (plus D.C.) never carried by Trump have 232 electoral votes. Winning Pennsylvania (19) and Wisconsin (10) would put Biden at 261. The 25 states Trump has never lost plus Georgia and Arizona would give him 262. In this scenario, the outcome in Michigan would be decisive, and it's significant that Michigan's 15 electoral votes would land Biden at 276—because, of the states that have never voted for Trump, Nevada appears alarmingly wobbly, but with Michigan, Pennsylvania, and Wisconsin Biden wouldn't need its 6 electoral votes to hit 270. I'm not mentioning (till now!) the wild card relating to the fact that, unlike the other 48 states, Maine and Nebraska do not necessarily award all their electoral votes to the statewide winner. Instead, the statewide winner gets two electoral votes (for its senators), while the rest are divvied up according to the outcome within congressional districts. Somewhat by chance, this complication probably results in a wash. Biden is almost certain to win Maine, but almost certain to lose the more inland and rural of its two congressional districts. The 1-vote deficit is likely then recaptured in Nebraska, where, though he's certain to lose statewide, Biden is a strong favorite in the congressional district that's home to Omaha—an especially strong favorite if he's doing well enough to win Wisconsin and Michigan. Biden won the district by 7 points in 2020.
All this is enough to make you jittery if, like me, you think another Trump term would be a catastrophe for our country. Note, for example, that Arizona's 11 electoral votes would not make up for losing Michigan: in this scenario, Biden also needs Nevada to win. It will always seem weird to me to talk this way, as if the opinions of people in more than 40 states mean nothing. Thanks electoral college!
Maybe events will change the cliff-hanger outlook. Like what? Law-and-order Republicans turn on Trump after he's convicted of a felony? Evangelicals head for the hills when he absurdly tries to "rebrand" himself on abortion? Dream on.