Now that President Biden has announced his intention to run for a second term, economists and politicos are assessing whether his candidacy will be helped or hurt by the performance of the economy. If there’s a recession, will it be over and mostly forgotten by Election Day?
Oxford Economics did an initial run of its election forecasting model, which takes economic factors into account, and found that Biden is in line to get around 55 percent of the popular vote, without any assumption about his opponent, according to a research briefing on Wednesday. Paul Krugman, my Opinion colleague, wrote Thursday that “the idea that the economy is going to pose a huge problem for Democrats next year isn’t backed by the available data.”
The truth is, though, that we really don’t know who will win the 2024 election, or even what role the economy will play in it. As somebody who writes about economics, I’d love to say that the state of the economy leading up to Nov. 5, 2024, will matter a lot. But that does not seem to be the case, according to people I spoke with this week. One possible reason is that voters have become more polarized and set in their preferences, and thus less swayed by the ups and downs of the economy.
For example, let’s say former President Donald Trump captures the Republican nomination. Most Biden supporters wouldn’t vote for him no matter how bad the economy got in 2024 — just as most Trump supporters won’t vote for Biden no matter how good the economy gets under the incumbent. James Carville’s admonition in 1992 that it’s “the economy, stupid” doesn’t hold up in this era of hyperpartisanship.
In 1978 a pioneering article by the Yale economist Ray Fair, “The Effect of Economic Events on Votes for President,” made the case that “economic events as measured by the change in real economic activity in the year of the election do appear to have an important effect on votes for president.” In January, the model predicted, using his economic projections as inputs, that Biden, if renominated, would get 50.07 percent of the two-party presidential vote next year, with no assumption about the Republican opponent.
Fair’s model, which he has tweaked over the years, correctly predicted the winner of the popular vote in every election from 1980 through 2008 except for 1992, when it incorrectly predicted Bill Clinton would get only 43 percent of the two-party popular vote against George H.W. Bush. It hasn’t done well lately, though: It predicted the Democrat would get less than half the popular vote in the two-candidate matchup in 2012, when President Barack Obama won a second term; in 2016, when Hillary Clinton won the popular vote but lost to Trump in the Electoral College; and in 2020, when Biden beat Trump.
I asked Fair if it was fair to say that polarization of voters has weakened the predictive value of economic indicators. He acknowledged that Trump, an especially polarizing figure, had underperformed in both 2016 and 2020 given economic conditions that the Fair model considered favorable to him. (And he said it could happen again if Trump is the G.O.P. candidate next year.) But he said he’s not convinced that his economic model has lost its predictive power. “Personally, I think there’s still a pretty big middle group” of voters who are influenced by economic factors, he said.
A good example of why it’s dangerous to over-rely on economic models is what happened in the spring of 2020, when Covid hit. Based on historical patterns, several of the best-known models (though not Fair’s) put a heavy weight on how the economy performs in the second quarter of an election year, namely April through June. Because of the Covid shutdown, the gross domestic product fell at an annual rate of 29.9 percent in the second quarter of 2020. Going by that one data point, the election should have been a disaster for the incumbent, Trump. But voters understood that Trump couldn’t be blamed for Covid. What’s more, the economy grew at an annual rate of 35.3 percent in the following quarter as it rebounded from the shutdown. In the end, Trump did worse than Fair’s model had predicted, but way better than predicted by models that heavily weighted second-quarter economic growth.
Another problem with some economics-based forecasting models such as Fair’s is that they predict the popular vote, rather than the one that really matters, the Electoral College vote, which depends on state-by-state results. Fair is sticking to forecasting the popular vote because he thinks it’s of academic interest. Some other forecasters have switched to predicting the Electoral College result, but it’s much trickier. The outcome of the election comes down to a handful of swing states, and within those few states, to the behavior of a small minority of voters whose minds aren’t made up. “The Electoral College throws a monkey wrench into the business,” Alan Abramowitz, an emeritus professor of political science at Emory University, told me.
One thing that puzzles me is why it’s even worthwhile to plug economic factors into an election forecast. If the relevance of the economy is that it affects voters’ feelings about the candidates, why not just cut to the chase and focus on the voters’ feelings? (Nate Cohn, my colleague on the news side, pointed out this week in another subscriber-only newsletter that the polls are showing a tight contest, with Biden slightly ahead of Trump and slightly behind Gov. Ron DeSantis of Florida in hypothetical matchups.)
I asked Charles Tien, a political science professor at the City University of New York’s Graduate Center and Hunter College, why he and others put economic indicators into their models. “When you add in the economy, it improves the results,” he said. But he acknowledged that it’s not obvious why that’s the case. When I asked Fair the same question by email, he wrote, “My empirical results are quite strong that the economy has mattered over time.” I wonder if it’s because economic indicators signal something about voters’ situations that they don’t fully express in surveys, which in any case have become less reliable as response rates have declined.
Peter Enns, a political scientist at Cornell who ran the Cornell-based Roper Center for Public Opinion Research until last year, told me he thinks it’s too soon for predictions about the 2024 race. First, because there are too many unknowns, such as the field of candidates and the business cycle. Second, because at this stage voters should be focusing on who should win, not who will win. OK, that’s fair. No more horse-race prognosticating from me. For now.
The Readers Write
What your newsletter about innovation misses is the input by experimentation. Our five senses are permanently providing us with personal experience that facilitates new creative thoughts. Only when chatbots are equipped with sensors can they become independent thinkers.
Heinrich Muller
Rancho Palos Verdes, Calif.
As for fertility, maybe all you men should think a little. Maybe women do not want to be saddled all through their adult years with raising your kids! Or maybe if men did a little more of the work, women wouldn’t mind so much.
Marilynn Miller
Chicago area
To test peer effects on fertility, why not hire actors to wheel baby carriages around one area and not around a demographically matched area nearby? Joking, of course.
David Auerbach
Durham, N.C.
With the large run-up in housing prices that you mentioned, selling a home and buying a new one may involve a huge increase in property tax payments. This may be a significant disincentive to selling in addition to higher mortgage rates.
Randy K. Vogel
Laguna Hills, Calif.
Quote of the Day
“I used to say to our audiences: ‘It is difficult to get a man to understand something, when his salary depends upon his not understanding it!’”
— Upton Sinclair, “I, Candidate for Governor: And How I Got Licked” (1935)
Source: Elections - nytimes.com
