fb-pixelMoody’s Analytics tool shows Biden winning in a ‘nail-biter’ Skip to main content

Moody’s Analytics election prediction model shows Biden winning in a ‘nail-biter’

In testing using historical data, the Moody’s Analytics tool correctly picked the winner in 10 of the 11 presidential matchups since 1980

Dark horse.Carolyn Kaster/Associated Press

This column is from Trendlines, my business newsletter that covers the forces shaping the economy in Boston and beyond. If you’d like to receive it via email on Mondays and Fridays, sign up here.

In the showdown that few people really want, Donald Trump holds a slight edge over Joe Biden in the polls. The former president is also the odds-on favorite in online betting markets.

But Democrats can take a little solace in this early indicator: An election prediction model designed by Moody’s Analytics shows Biden winning the White House in a “nail-biter.”

The economic research firm’s model has the president taking 308 electoral votes, 38 more votes than the 270 needed to win and two more than in 2020.


So what? It’s just one prediction, but in testing using historical data, the Moody’s model correctly picked the winner in 10 of the 11 presidential matchups since 1980, including Trump’s upset over Hillary Clinton in 2016. The one misfire: 2020, when fallout from Trump’s handling of COVID and extraordinary turnout by Democrats put Biden over the top.

The model factors in political variables, including the advantage held by incumbents, approval ratings, and the share of the popular vote won in previous elections by third-party candidates. It also incorporates a variety of economic data and assumptions. It spits out results for each state, using “a pooled ordinary least-squares regression with state fixed effects” — whatever the heck that is.

“On the margin, political factors favor Trump’s candidacy, while economic factors favor Biden’s,” the firm said in a summary.

Wait just one minute: I know what some of you are thinking. On Friday I wrote that “It’s the economy, stupid” — James Carville’s famed declaration about the Clinton vs. Bush I election — may not matter in this election.


I cited Nate Cohn, chief political analyst at The New York Times, who argues that extreme partisanship has left the country divided on a host of ideologically charged issues that could motivate voters.

But Moody’s model doesn’t assess ideological issues. It’s all about data and historical trends. And it gives Biden a small advantage, for now.

How it works: I asked Mark Zandi, the firm’s chief economist, if economic variables carry more weight than the political ones in the model.

“Both the political and economic factors are critical. Neither by themselves are sufficient to explain presidential election outcomes,” he said in an email.

The most statistically important factor is political: the popular vote that the incumbent party received in each state in the previous presidential election. The math shows that the incumbent party usually wins almost 90 percent of the votes it got four years earlier.

The most important economic variable is whether there’s been a recession in the election year or the year before. Since 1948, an incumbent president has lost just three times, each amid a downturn: 1980 (Reagan over Carter), 1992 (Clinton over G.H.W. Bush), and 2020 (Biden over Trump).

Details: Here’s a closer look at Moody’s analysis of key economic factors.

Gasoline prices have fallen substantially since peaking at $5 a gallon in June 2022. Prices are closer to $3 now but are forecast to climb toward $3.50 by Election Day, limiting the bump for Biden.


The president should get a substantial boost from rising incomes. Wages after inflation are rising and by November will be well above pre-pandemic levels in most states, supported by continued low unemployment.

Declining mortgage rates are a modest plus for Biden. Moody’s sees the rate on a 30-year fixed loan dropping to near 6.5 percent by the time voters go to the polls, down from 8 percent last year. The president may also benefit if the Federal Reserve, as expected, begins cutting interest rates in May or June, which will make it cheaper for consumers and businesses to borrow.

The Conference Board’s consumer confidence index is close to its decades-long average and moving higher, favoring Biden, according to the firm.

Caveats: Moody’s cautions that small changes in its assumptions could flip the margin of victory to Trump.

For example, a 2-percentage-point increase in Republican turnout over 2020 would turn the election in his favor. The same would happen if third-party candidates took 7.5 percent of the vote, compared with 1.8 percent four years ago, which was low by historical standards.

Final thought: A lot can happen before Election Day. Trump could be convicted. Biden could suffer a health setback. The economy could tank. Like the polls, election models can tell us only so much.

The conventional wisdom is a few swing states will determine the next president. Moody’s has Biden winning four battleground states — Georgia, North Carolina, Nevada, and Pennsylvania — by less than 1 percentage point, and losing one, Arizona, by a similar margin.


“The outcome may not be determined on election night as several states face the prospect of automatic recounts and court challenges — a situation likely to raise anxieties given the aftermath of the 2020 election,” Moody’s said.

Two unpopular candidates in a tight race that might end in turmoil. No wonder so many of us dread the next nine months.

Larry Edelman can be reached at larry.edelman@globe.com. Follow him @GlobeNewsEd.