Despite Donald Trump’s decisive victory in the presidential election, a political scientist who developed a model that correctly predicted his set of battleground states warns that voters have not necessarily given the president-elect a mandate to make radical changes.
In an article published with little fanfare three weeks before the election, Peter Enns, a professor of government at Cornell University, and his co-authors accurately predict that Trump would win all seven swing states, based on a model they built and that uses state-level presidential approval ratings. and indicators of economic health.
In an interview with the Guardian, Enns said his model’s conclusions suggest voters chose Trump not because they wanted to see his divisive policies implemented, but rather because they were frustrated with the state of the economy during Trump’s presidency. Joe Biden, an obstacle that Kamala Harris was not. popular enough to overcome.
“If this election can be explained by what voters thought of Biden and Harris and economic conditions, that really goes against Trump’s idea of a mandate for big change,” Enns said.
“If Trump wanted to maximize support, being cautious about changes that are huge changes would be the optimal strategy, according to the model.”
During his campaign, Trump promised norm-shattering measures to achieve his goals, ranging from deploying the military to carry out mass deportations of undocumented immigrants to imposing trade tariffs on allies who do not cooperate with his administration.
On November 5, voters responded by giving Trump an overwhelming victory in the Electoral College, and also by making him the first Republican in two decades to win the popular vote.
Both outcomes were predicted in the paper published on October 15 by Enns, Jonathan Colner of New York University, Anusha Kumar of the Yale University School of Medicine and Julius Lagodny of the German media company El Pato. At the time, polls from the seven swing states showed Trump and Harris tied, mostly within their margin of error, indicating that either one of them had to win the election.
Rather than focusing on candidates’ support across the country or in swing states, Enns and his co-authors built a model that combines two types of data: presidential approval ratings from all 50 states using data from Verasight, the research agency he co-founded. among others, and an index from the Federal Reserve Bank of Philadelphia that measures real income, production and labor market conditions at the state level. Both sets of data were collected more than 100 days before the vote.
Enns first used the model in the 2020 presidential election, where it correctly predicted the outcome in 49 states, excluding Georgia. This year, Enns and his co-authors wrote that Harris, who took over as Democratic nominee for Biden in late July, was on track to lose both the popular vote and the Electoral College, including the battleground states of Michigan, Pennsylvania, Arizona and Nevada. , Wisconsin, North Carolina and Georgia.
“If Harris wins the election, we won’t know exactly why, but we will know that her victory overcame circumstances so detrimental to the Democratic party that the incumbent president dropped out of the race. She will have given a major boost to the Democratic campaign and/or Trump and the Republican party will have lost a significant advantage,” Enns and his co-authors wrote.
The prediction turned out to be correct, but as vote counting continues in a few states, Trump appears poised for a majority victory in the popular vote, not the 50.3% majority they predicted.
Then there’s the question of whether Biden would have done better if he had stayed in the race. The 82-year-old president has been unpopular for most of his term as Americans endured the highest inflation rate since the 1980s even as the labor market rebounded strongly from the Covid pandemic. Biden was also dogged by concerns about his age and fitness for office, which culminated in a terrible debate against Trump in June that forced him to drop out of the race weeks later.
“Given Biden’s low approval ratings and economic circumstances, our model predicted a less than one in 10 chance of a Biden win had he remained in the race. Even taking into account Harris’s approval ratings, which are significantly higher than Biden’s, Democrats face an uphill battle,” the authors wrote.
If Harris had had a chance to overcome the disadvantages she faced entering the race, Enns said it would have had to convince voters she would be a very different president than her boss — which she apparently failed to do.
“There are some economic headwinds, there are headwinds from the Biden establishment. And what that suggests to me is that, given the headwinds that Harris faced, the optimal strategy would have been to differentiate himself more from Biden,” Enns said.
But the vice president’s fate may have been sealed in the years leading up to her bid for the White House, when she failed to build the kind of public profile that would have brought her approval ratings to the level she needed .
“If she had been more popular, you can imagine what could have happened to make our prediction wrong. So the fact that our 100-day forecast was so accurate, which really benefited the campaign, had minimal effect on the outcome,” Enns said.
“The task ahead was to exceed the forecast, and her campaign failed to do that.”
