A "State of the Horserace" Post
I haven't done one in a while; let's throw a bone to the junkies
The Races at Longchamp by Edouard Manet
I haven’t done a horserace post in a while, and I have a hook for one: Nate Silver just released his new model showing Trump a 2:1 favorite to win the election in November. Silver attracts a lot of ill-informed criticism of his core competency, but in this case I know he’s done a good job because his conclusion accords pretty much perfectly with my own gut feeling about the race. I’ll use this space to hitch the wagon of my gut to the star of Silver’s model, and hope it makes the former look a lot more impressive than it would be on its own.
Here’s how I’ve been thinking about the election for quite some time.
Biden is an unpopular president, and has been for most of his tenure. Some of the reasons why he is unpopular strike me as unfair. (He gets little credit for his legislative agenda, for example.) In other areas, I think Biden has made generally good decisions but the results have left people understandably frustrated and unsatisfied. (I think both his economic management and his foreign policy record fall into this category.) And in other areas, I think Biden has earned disillusionment. (His record on asylum and migration, for example.) Biden has also spent political capital appeasing the left wing of his party in ways that delivered neither important policy wins nor popular talking points, a pattern which has left him looking like a weak leader unable to discipline factions that should be among his strongest supporters. This perception of being a weak party leader in turn has reinforced legitimate concerns about his age and whether he is up to the job (or will remain up to the job for much if any of a second term). Add in residual anger over the pandemic, translated into nostalgia for the Trump years, and the base level of extreme polarization and disillusion with government that make it hard for any president to be broadly popular, and I think Biden’s poor ratings are eminently understandable.
In sum: popular dissatisfaction with Biden is such that, were he facing a challenger who could position himself as a popular agent of change, he’d be expected to lose both the popular vote and the electoral college by a significant margin. We’d be looking at a replay of 1980 or 1992.
But Biden isn’t facing a challenger who is a popular agent of change. He’s facing a challenger who is a former incumbent, and therefore well-known, one who is personally unpopular, was widely considered a failure at the end of his term (which is why he lost his bid for reelection), and who after losing engaged in outlandish efforts to overturn the result, ending with a mob violently invading the capitol. Since then, that challenger has become a convicted felon. This candidate, surely, shouldn’t be able to beat Biden, no matter how badly people feel about Biden’s tenure.
Except that the voters who will decide the election clearly aren’t thinking about Trump in such black-and-white terms. If they were, Trump wouldn’t have remotely the level of support that he has. His core supporters are fanatical in their devotion, but what has kept him competitive in his bid to return to office are those people who don’t like these aspects of Trump’s personality or his record, but who don’t view them as disqualifying. Maybe they don’t appreciate their gravity, or they have such a low opinion of politicians generally that Trump doesn’t stand out, or they see Biden and the Democrats as being equally disqualified for real or imagined abuses. In many cases, I suspect voters are engaging in motivated reasoning, choosing not to see Trump’s character and behavior as disqualifying because they want the opportunity to vote against Biden, who is the sitting president, and Trump is the only realistic alternative they’ve got.
If voters do wind up seeing this as a change election, a referendum about how they feel about Biden, then Trump will win handily. Biden needs to make this a choice election—and to make Trump into the less-palatable choice for both policy reasons and personal reasons (he has ample material to work with on both fronts) while making himself and his party look more palatable (a much bigger challenge at this late date). But Biden has been repeatedly tempted to try to make this a no choice election, to suggest that Trump is simply unacceptable and therefore voters have to reelect the president—to preserve democracy, to preserve America’s soul, etc. I didn’t think that was going to work; indeed, I think it ran the risk of backfiring, leading a certain number of voters to basically take the dare rather than being scared into line. Biden needed to make the case for his second term (or step aside in favor of someone who could), I thought; he couldn’t just coast on incumbency or try to rule his opponent out of bounds.
That’s how I’ve thought about the election for at least the past year. For most of that time, Biden’s position has deteriorated slightly but steadily, to the point where Trump was clearly winning in both the battleground states and, for the first time, in polls of the national popular vote. Biden has rebounded slightly since Trump’s felony conviction, but he’s still running slightly behind. More important than this deterioration or the rebound, though, has been the race’s stability, which is a key contrast to 1980 or 1992, making the race look more like 2004 or 2012, but with the incumbent as the underdog. And running slightly behind in a stable race is not a good place for an incumbent to be.
Biden still has a real chance to win, of course, and he has some things potentially working for him. High-propensity voters now favor Biden by a meaningful margin, which should give him an advantage in a low-turnout election (which 2024 might well be given how disliked both candidates are). RFK Jr.’s vote is also highly uncertain, which could play to either candidate’s advantage. The base-case assumption should be that he pulls from both parties, as Perot did in 1992, and therefore doesn’t meaningfully alter the outcome. If low-propensity voters can be convinced to vote for RFK Jr. once they learn reasons to dislike Trump, though, he might help Biden win. On the other hand, Trump might consolidate those undecideds who dislike him but want change while RFK Jr. consolidates those who will not vote for Trump but who want to vote against Biden, which would be a replay of John Anderson’s 1980 bid.
There also appear to be some interesting developments in the Electoral College skew, which mildly favored Obama in 2012, but favored Trump in 2016 and, even more strongly, 2020. If you believe the polls are accurate, then it looks like Trump’s Electoral College advantage may have shrunk meaningfully or even vanished. In 2020, Biden won the popular vote by 4.5% but won the tipping point state—Wisconsin—by less than 1%. In current polls, Biden is behind by about 1% on average in the national polls, and in the tipping point states—Pennsylvania, Wisconsin and Michigan—he’s also polling behind by 1-2%. What happened? Well, part of what happened is Dobbs, which has made abortion a meaningful electoral issue in these states, one that cuts strongly against Republicans. Biden has also lost ground across the Sun Belt, which hurts both his national numbers and his Electoral College options—but not necessarily the Electoral College skew. He’s now polling meaningfully behind in states that he won in 2020—Nevada, Arizona and Georgia—as well as North Carolina, which he lost. In 2020 these states were all close to the tipping point, which means they skewed toward the GOP—this year, they continue to skew toward the GOP, but appear much less-likely to tip the election. That means the Electoral College is less skewed even as the playing field as a whole is tougher for Biden. Finally, I suspect that another part of what’s happening is that Biden has lost ground in the Northeast. Biden is still going to win New York by a large margin, but it will be a smaller margin than in 2020, and New York will move more in the GOP direction than the nation as a whole (or that’s what I expect, anyway). In Electoral College terms, those are wasted gains for Trump, just as Hillary Clinton’s gains in Texas were wasted since she didn’t come close to winning the state.
If all of the foregoing is correct, then Biden really does need to go all-in on a “rebuild the great big beautiful Blue Wall” strategy and make sure he wins Wisconsin, Michigan and Pennsylvania. He not only has no backup strategy for winning the Electoral College, but if the polls are right then these states are states he has a very real shot at winning even in a tied race. That “if the polls are right” is a key caveat, though; in 2020, the biggest state polling miss was Wisconsin—on average, Biden was polling ahead by over 8% on the eve of the election, and wound up winning the state by less than 1%. Nonetheless, Biden doesn’t really have anything else to guide him but the polls, and I can’t think of an argument that he has a better shot at making up ground with Hispanic voters and voters upset about immigration, which is what he’d need to regain competitiveness in the Sun Belt.
Actually, by this point, I think it is probably too late for Biden to change voter perceptions of his own administration in any meaningful way. The economy isn’t going to bail him out even if it improves, and any new initiatives he announces won’t have any time to take effect, if they even have a chance of being enacted. Biden has finally paid attention to the asylum question, for example, but it’s too late to neutralize the issue. Surprises from “events” are far more likely to redound to Biden’s detriment rather than his favor; it’s not like he can hope for an “October surprise” like the release of American hostages that the Reagan campaign feared. Trump lost some support from his conviction, but right now it looks unlikely that more shoes will drop on that front before the election. Biden can still blunt voter perceptions of his personal incapacity due to age by performing well in debates and such, but the best he can do is stanch further bleeding on that front. He needs to actually regain ground.
His best hope, then, is to get into a sharp policy argument with Trump on matters that the public can clearly understand—and win it. Trump will cut taxes on the rich. He’ll raise taxes (in the form of tariffs) on the middle class, further worsening a typical family’s finances. He’ll enact federal restrictions on abortion, and he’ll cut funding for healthcare. His biggest risk is that voters see Trump simultaneously as an agent of desired change and someone who doesn’t really want to do anything radical but will just let things “naturally” ease back to the good old days of 2019. He’s got to try to turn that on its head, convince people that Trump is an ordinary right-wing Republican, just more corrupt, authoritarian and incompetent—and he has to do so in the face of popular nostalgia. That’s a tall order, but it’s probably easier than convincing people that a second Biden term will be notably different in ways that they will approve, rather than in ways they will hate even more than they did the first.
Nate Silver thinks that Biden’s chances are about 1 in 3. That’s also what he estimates are the odds of Biden holding onto Pennsylvania, Michigan and Wisconsin—because that’s really his only plausible path to victory. Biden probably has about an even-odds chance of winning each of them—but he has to win all three. I think those are the right states for the above negative pitch about Trump and the agenda he and the GOP are running on. We’ll see soon enough if he and his party can run that play, and win with it.