POLITICS: Trump’s Electoral College Edge Seems to Be Fading
The early polls show Donald J. Trump and President Biden tied nationwide. Does that mean Mr. Trump has a clear advantage in the battleground states that decide the Electoral College?
It’s a reasonable question, and one I see quite often. In his first two presidential campaigns, Mr. Trump fared far better in the battleground states than he did nationwide, allowing him to win the presidency while losing the national vote in 2016 and nearly doing it again in 2020.
But there’s a case that his Electoral College advantage has faded. In the midterm elections last fall, Democrats fared about the same in the crucial battleground states as they did nationwide. And over the last year, state polls and a compilation of New York Times/Siena College surveys have shown Mr. Biden running as well or better in the battlegrounds as nationwide, with the results by state broadly mirroring the midterms.
The patterns in recent polling and election results are consistent with the trends in national surveys, which suggest that the demographic foundations of Mr. Trump’s Electoral College advantage might be fading. He’s faring unusually well among nonwhite voters, who represent a larger share of the electorate in noncompetitive than competitive states. As a consequence, Mr. Trump’s gains have probably done more to improve his standing in the national vote than in relatively white Northern states likeliest to decide the presidency, like Pennsylvania and Wisconsin.
Midterm results typically don’t tell us much about the next general election. Polls taken 15 to 27 months out don’t necessarily augur much, either. But the possibility that Republicans’ Electoral College advantage is diminished is nonetheless worth taking seriously. It appears driven by forces that might persist until the next election, like Mr. Biden’s weakness among nonwhite voters and the growing importance of issues — abortion, crime, democracy and education — that play differently for blue and purple state voters.
Of course, there is more than a year to go. Mr. Biden may regain traction among nonwhite voters or lose ground among white voters, which could reestablish Mr. Trump’s Electoral College edge. Perhaps his Electoral College edge could grow even larger than it was in 2020, as some Democrats warned after that election.
But at this point, another large Trump Electoral College advantage cannot be assumed. At the very least, tied national polls today don’t mean Mr. Trump leads in the states likeliest to decide the presidency.
There are three basic pieces of evidence suggesting that Mr. Trump’s key advantage might be diminished today: the midterms, the Times/Siena polls and state polls.
The 2022 midterms were a surprise. Republicans won the national vote, just as the polls anticipated. With Republicans usually faring better in the battlegrounds in recent cycles, a national popular vote advantage might have been expected to yield a “red wave.”
But Democrats held their ground in the battleground states, allowing them to retain the Senate and nearly hold the House. Nationally, Republican House candidates won the most votes by about two percentage points (after adjusting for uncontested races). The margin was almost identical in the presidential battlegrounds, like Arizona, Pennsylvania and Wisconsin, where Republican House candidates also won by two points.
The shrinking gap between the key battleground states and the national popular vote wasn’t just because of Democratic resilience in the battlegrounds. It was also because Republicans showed their greatest strengths in noncompetitive states like California and New York as well as across much of the South, including newly noncompetitive Florida. Democratic weakness in these states was just enough to cost them control of the House of Representatives, but did even more to suppress Democratic tallies in the national popular vote, helping erase the gap between their strength in the battlegrounds and the national vote.
Does the House popular vote tell us much about the Electoral College two years later? Possibly, though not necessarily. The 2018 midterms showed House Republicans running well in key battleground states, foreshadowing Mr. Trump’s expanded Electoral College advantage two years later. Republican strength by state in the House mirrored the presidential race in 2020 as well. Perhaps it should be expected to foreshadow the presidential vote by state again.
But today, it’s harder than it was at this time in the last cycle to connect voter attitudes about the House with presidential preference. One major issue: The House results weren’t highly correlated with Mr. Biden’s approval rating. In contrast, the tight relationship between the House vote and Mr. Trump’s approval rating back in 2018 made it reasonable to believe the distribution of the House vote told us something about his strength heading into 2020.
The midterms are an important clue, but additional data is probably needed to connect what happened last November to what might happen next November.
Times/Siena polling over the last year offers additional evidence of such a connection.
Overall, Mr. Trump has gained in the places where Republicans fared well in the midterms, while Mr. Biden is holding up well in the states where Democrats fared well in the midterms, based on a compilation of 4,369 respondents to Times/Siena polls.
On average, Mr. Biden continues to match his 2020 performance in the states where Democrats fared better than average in the midterms, a group that includes every major battleground state. Instead, all of his weakness in Times/Siena national polling is concentrated in the states where Democrats fared worse than average last November.
In the sample of 774 respondents in the battleground states, Mr. Biden leads Mr. Trump, 47-43, compared with a 46-44 lead among all registered voters nationwide. On the other hand, Mr. Biden leads by 17 points, 50-33, in a sample of 781 respondents in California and New York — the two blue states that primarily cost Democrats the House last November — down from a 27-point margin for Mr. Biden in 2020.
In general, I am loath to look at geographic subsamples in our polling; results by state are just so sensitive. For this analysis, it makes a huge difference whether Mr. Biden is tied in the battlegrounds or up by five points.
But in this particular case, the specific findings are part of the broader pattern supported by larger samples. Splitting our sample into two groups, we have over 2,000 respondents in states where Republicans did well and states where Democrats held up. The trends in both groups align with those of the midterms, and, while the sample is small, the pattern also appears to filter down to the crucial battlegrounds.
There aren’t too many polls of the key battleground states at this early stage. But the available survey data doesn’t show any sign of an Electoral College advantage for Mr. Trump, either.
Over the last year, Mr. Biden leads by 1.3 points in national polls, while he leads by at least one point in the average of polls taken in Michigan, Pennsylvania and Wisconsin — three states that would probably be enough to re-elect him.
In contrast, Mr. Biden won the national vote by 4.5 points in 2020 while winning Wisconsin by just 0.6 points. The key measure of Electoral College strength, relative to the national vote, is the difference between the national vote and the “tipping-point state” — the state that pushes a candidate over the Electoral College threshold. That difference was roughly 3.8 percentage points in Republicans’ favor in 2020 and 2.9 points in 2016, with Wisconsin the tipping-point state in each case. In the state polling today, that gap is essentially nonexistent.
On the other end of the competitiveness spectrum is New York, one of the most solidly blue states in the country. Mr. Biden will surely win the state, but he may not do as well there as he did in 2020. He holds a 48-35 lead in eight polls over the last year, including a 47-34 lead in a Siena College poll last month. For what it’s worth, you can add a 49-36 margin in the Times/Siena compilation of 256 respondents in New York.
In one sense, New York was the worst state in the nation for House Democrats in 2022, based on their mere nine-point aggregate House win compared with Mr. Biden’s 23-point win in the state in the 2020 presidential election. The state numbers today look as reminiscent of the midterms as the last presidential election. Results like these in blue states will hurt Mr. Biden in the national polls and popular vote, but won’t do anything to hurt his chances in the Electoral College.
The new issues
Together, the midterms, the state polling and the Times/Siena polls offer three serious if imperfect data points suggesting Mr. Trump isn’t faring much better in the battleground states compared with nationwide, at least for now.
But why? Broadly speaking, there are two major theories: the issues and demographics.
First, the issues. In the aftermath of the midterms, Democratic strength in key battleground states appeared attributable to specific issues on the ballot, like abortion, crime and democracy. This helped explain some aspects of the election, including the failures of anti-abortion referendums and stop-the-steal candidates — and perhaps New York Democrats.
It’s possible these new issues are helping to shift the electoral map heading into 2024 as well. New issues that have emerged since 2020 — abortion rights, trans rights, education, the “woke” left and crime — are primarily state and local issues where blue, red and purple state voters inhabit different political realities, with plausible consequences for electoral politics.
Moderate voters in a blue state — say around Portland, Ore. — have no need to fear whether their state’s conservatives will enact new restrictions on transgender rights or abortion rights, but they might wonder whether the left has gone too far pursuing equity in public schools. They might increasingly harbor doubts about progressive attitudes on drugs, the homeless and crime, as visible drug use among the homeless in Portland becomes national news.
But moderate voters in a purple state — say those who live around Grand Rapids, Mich. — might have a different set of concerns. The “woke” left could be a very distant worry, if they understand what it is at all. They’ve probably never heard of the gender unicorn. Their city’s crime, homelessness and drug problems don’t make national news.
What does make national news is the conduct of their state’s Republican Party, which not only tried to ban abortion last fall but also embraced the stop-the-steal movement. The “threat to democracy” is not an abstraction for Biden voters here: It was their votes that Mr. Trump and his allies tried to toss out.
This is a plausible explanation, if one that’s hard to put to the test. The apparent relationship between the midterms and presidential polling is perhaps the best piece of evidence, if we stipulate that the pattern in the midterms was indeed explained by the varying salience of these state and local issues.
Shifts among demographic groups
Mr. Trump’s Electoral College advantage was built on demographics: He made huge gains among white voters without a college degree in 2016, a group that was overrepresented in the key Northern battleground states. It let him squeak by in states like Wisconsin, Michigan and Pennsylvania, even as his weakness among college-educated voters cost him votes — and ultimately the popular vote — in the Sun Belt and along the coasts.
The polls so far this cycle suggest that the demographic foundations of Mr. Trump’s advantage in the Electoral College might be eroding. Mr. Biden is relatively resilient among white voters, who are generally overrepresented in the battleground states. Mr. Trump, meanwhile, shows surprising strength among nonwhite voters, who are generally underrepresented in the most critical battleground states. As a consequence, Mr. Trump’s gains among nonwhite voters nationwide would tend to do more to improve his standing in the national vote than in the battleground states.
Overall, 83 percent of voters in Wisconsin, Michigan and Pennsylvania were white in the 2020 election, according to Times estimates, compared with 69 percent of voters elsewhere in the nation. Or put differently: If Mr. Biden struggled among nonwhite voters, it would do a lot more damage to his standing outside of these three states than it would in the states that make up his likeliest path to 270 electoral votes.
Is this enough to explain Mr. Trump’s diminished advantage? It could explain most of it. If we adjusted Times estimates of the results by racial group in 2020 to match the latest Times/Siena polls, Mr. Trump’s relative advantage in the Electoral College would fall by three-quarters, to a single point.
In this demographic scenario, Mr. Biden would sweep Wisconsin, Pennsylvania and Michigan. He would lose Arizona, Georgia and Nevada, just like in the state polls conducted so far. It would be a narrow Biden win if everything else went as expected: He would earn 270 electoral votes, exactly the number needed to win.
There’s also a chance that maybe, just maybe, Democrats might defy these unfavorable national demographic trends in states like Arizona and Georgia. After all, these two states lurched leftward in 2020, even though nonwhite voters shifted to the right nationally in that election as well. Clearly, other state-specific trends canceled out Mr. Trump’s gains among nonwhite voters: White voters moved more toward the left than elsewhere in the country; the nonwhite share of the electorate grew more than it did elsewhere; and Democratic support among nonwhite voters appeared relatively sturdy, for good measure.
If those state-specific trends prevail over the national ones again, perhaps Mr. Biden can hope to get the best of both worlds: good results in the Northern battlegrounds, thanks to his national strength among white voters, with resilience in the blue-trending Sun Belt states where idiosyncratic factors might cancel out unfavorable national demographic trends.
With more than a year to go, none of this is remotely assured to last until the election. But at least for now, a tied race in the national polls doesn’t necessarily mean that Mr. Trump has a big lead in the Electoral College.