By Stephen Fisher, 14^th November 2024

In last week’s US presidential election there was a modest swing to the Republicans across the country including in the most marginal states. That swing was broadly consistent across different counties (see hereand here), suggesting national-level factors were the most important. 

Post-covid inflation (which apparently made the median voter poorer) led to a drop in approval for the Biden administration from which it never recovered. As Nate Cohn has pointed out, “no party has ever retained the White House when the president’s approval rating was as low as it is today and when so many Americans thought the country was on the wrong track.”

Voters told opinion pollsters that the economy was their main concern and that they think Trump would manage the economy better than Harris. Those who expressed the most concern about the economy swung more heavily to Trump. Similarly, the most economically insecure socio-demographic groups were the ones that apparently swung most to Trump.

Nearly all the forecasting models based on economic factors from the academic forecasting symposiumsuggested Trump would win the share of the vote as he has done. Most strikingly, Ray Fair’s model, which uses only retrospective economic data, prior shares of the vote and incumbency (which party is in the White House), predicted Trump would win 50.5% of the two-party vote. That is extremely close to the 51.0% he appears to have secured. 

The ousting of the Democrats also fits a broader international pattern whereby incumbent governments have suffered substantial losses in 2024. By such international comparisons, both the US economy and the Democrat vote held up fairly well. Just not well enough for Kamala Harris. 

There are still puzzles still to resolve in the role of economic voting. Why did Biden’s approval ratings fail to recover as the economy improved? Also, how come, as Paul Krugman pointed out, the Democrats lost this year when inflation and unemployment are lower than what they were when Reagan sailed to victory proclaiming it is “Morning again in America”? 

While those and other puzzles should be acknowledged and resolved, they do not substantially undermine the idea that the Democrats lost because of the economy.

Enough said? 

Yes, in the sense that we do not need to invoke any more explanatory factors to account for the election outcome. The Democrats lost by as much as they should have done given the economic circumstances and the way that voters in the US typically react to economic circumstances. On that basis, Trump did not win so much as the Democrats lost because they are in the White House presiding over what appears to most Americans as a bad economy. Maybe Harris, or some other Democrat, could have performed better, but she did no worse than could have reasonably been expected given the economic circumstances.

But economic circumstances were not the only circumstances, and it is not enough to say the election outcome was just down to dissatisfaction with economic performance.

First, a caveat: just because Americans appear to have responded to economic circumstances as they have typically done, does not mean they have behaved rationally. Chris Achen and Larry Bartels have devastatingly argued that the historic record is one of irrational myopia (attention to the past few months instead of a president’s full term) and a tendency to punish presidents for things that are not their fault. Voters might have turned against the Democrats because of the economy, but they were not necessarily right or sensible to do so.

As well as accounting for change, there were important developments that might have affected the outcome but did not.

Perhaps the most profound question about the 2024 election is whether Trump should have been on the ballot at all. He attempted to overturn the 2020 election results, and on January 6^th 2021 he urged his supporters to march to the Capitol, where they subsequently forced entry and disrupted the count of electoral college votes. As a result, he was impeached by the House. The Senate could then have convicted him and banned him from running again. But not enough Republican senators were willing to vote to convict. The legal and political arguments and developments after that point were convoluted, including disagreements among Supreme Court judges.  

Winning the election does not vindicate Trump. After the Civil War the US Constitution was, with good reason, amended to prohibit those who have “engaged in insurrection or rebellion” from holding office. Including the presidency in that clause would have been irrelevant if winning an election were some kind of vindication. The fact that the Insurrection Clause was never successfully used to block Trump’s election does not imply that it could not have been justly applied. Anyone who believes Trump was guilty of insurrection can reasonably take the view that the Insurrection Clause ought to have been applied. On that basis, those who believe there is sufficient evidence that Trump engaged in insurrection are entitled to view Trump’s candidacy as illegitimate; not illegal, but illegitimate in a normative sense, because of the failure of elected politicians and/or the legal system to apply the provisions of the US constitution. 

In opinion polls between March 2023 and January 2024, most Americans agreed that, “Trump should withdraw [his] candidacy due to January 6 charges or not serve or be elected President if charged or convicted of a serious crime.” But various polls showed slightly fewer than half of Americans believing Trump was guilty of conspiring to overturn or attempting to obstruct the certification of the 2020 election results. For example, 45% held that view and a further 20% were unsure in this poll from January 2024. In the same poll some 72% of Republicans and Independents who lean Republican wanted to see Trump as the Republican nominee for president. Ultimately, most Republicans do not think Trump is guilty of trying to subvert the 2020 election (see for example polls here and here).

Partisan disagreements over who is guilty of what are nothing new. In 2016, those intending to vote for Hilary Clinton thought she did nothing wrong in setting up a personal email address and server for her work as Secretary of State, but Trump supporters thought she was a criminal. Conversely, despite widespread publication of a recording of Trump bragging about violating women, New York Times/CBS polls found 41% of all Americans saying that “the allegations that Donald Trump made unwanted sexual advances against women” were “mostly not true.” That 41% probably included a majority of Trump voters. Just before the 2016 election, 73% of Trump supporters said they thought he “respects women”.

By the start of the 2024 campaign, Trump faced 91 felony charges across four criminal cases, of which he has so far been convicted of 34 that were unrelated to the 2020 election. Many of his former allies have criticised him and cautioned against voting for Trump. Various commentators set out before the election reasonable and serious concerns about a second Trump presidency for the future of democracy in America (e.g. see Martin Wolf’s piece here).

Rather than the cumulative effect of all these concerns leading Americans to think less well of Trump than they did when he was first elected, Trump’s reputation has apparently improved. Economist/YouGov polls of registered voters from 2016 and 2024 show, using the same question wording, that 39% had a favourable opinion of Trump eight years ago but as many as 49% do now. Similarly, by comparison with 2016 and 2020, Gallup surveys show Americans are at least, or more, likely to view Trump as a “strong and decisive leader”, able to “manage the government effectively”, “care about the needs of people like you” and “honest and trustworthy”. 

So, Trump won this year not just because the economic circumstances were more favourable to the Republicans, but also because he himself has greater personal support than before. That is despite, or maybe because of, the legal wranglings. 

Republicans argued that Trump was a victim of lawfare from Joe Biden’s Department of Justice (e.g see hereand here). After Trump was shot in the ear, that argument was used to blame Democrats for the assassination attempt. YouGov found that Trump supporters thought Joe Biden and the Democrats (78%) and left-leaning media (77%) were “somewhat” or “very” responsible for the July attempted assassination. But that event on its own seems to have increased Trump’s favourability rating by just a percentage point.

Other policy areas seemed to work on average to Trump’s advantage. AP VoteCast found that more than half of voters said Trump had the “right policy ideas.” As well as the economy, immigration and abortion were each considered by more than 10% of voters to be the most important issue facing the country. Both issues wereconsiderably more important to voters in 2024 than 2020. 

Demand for a reduction in immigration has risen from 28% in 2020 to 55%, and 47% of Americans supportTrump’s policy of deporting all illegal immigrants. 

Back in 2016 Trump promised to appoint anti-abortion Supreme Court justices. He delivered on that promise, and as a results the 1973 Roe v Wade ruling was overturned. As a result, millions of American women lost the right to abortion. Although two-thirds of voters want abortion to be legal, the exit poll found some 29% of them still voted for Trump. Meanwhile, 91% of voters who think abortion should be illegal voted for Trump. White evangelicals (or born-again) Christians have long been strong Republican voters. They voted 79% for George W Bush in 2004, 79% for Romney in 2012, and 80% for Trump in 2016. His ability to deliver on his promise seems to have helped maintain their loyalty with 82% support this year. White evangelicals are the Republican rock on which the temple of Trump was built.

Ultimately it seems as though it is both too easy and impossible to account for Trump’s victory. Easy because the economic circumstances were bad for the Democrats, immigration was an important concern and Trump’s stance was popular, and he himself was more popular than he was before. Meanwhile many of the things that might have been expected to put people off Trump did not dissuade most voters. There is more research to be done, but ultimately it can become impossible to fully and adequately appreciate why something one person thinks is intolerable is not a problem or not even believed to be the case by another.

My explanation of how and why Trump was elected in 2016 is here. It ended by saying:

“To those who feel that voting for Trump was reprehensible, I should say that my attempt to explain the choice is not an attempt to justify it. His voters are responsible for his election, and so somewhat also for the consequences of his presidency. Elections can have profound effects, which makes voting a serious responsibility for citizens. This also means it is worthwhile trying to understand voting behaviour.”

That’s true again now.