The Polls Were Still Wrong (But Also Sorta Right)

So, of course, my predictions in “The Polling is Wrong” in Issue 13, Volume 114 and “The Polls are Still Wrong (But Getting Better)” in Issue 2, Volume 115 of The Spectator regarding the Democratic Party’s performance in the 2024 election did not pan out. I predicted when President Joe Biden was in the race that he would manage to win, and when he withdrew, that Vice President Kamala Harris would win. Former President Donald Trump outperformed expectations once again to win a second nonconsecutive term—only the second president after Grover Cleveland—and swept every battleground state, winning the popular vote by just over 1.5 percent thanks to middling Democratic performances in New York, New Jersey, California, Florida, and Texas. In the wake of these results, I’ve been asking myself what went wrong.

The guiding principles of my predictions were the 2022 environment and inflation being even less salient than in 2022 due to wages catching up to inflation. The problem, of course, was assumptions made about the 2024 electorate in comparison to 2022 that should have been fatally obvious, in hindsight. I assumed the 2024 electorate was far more favorable to Democrats than it really was, due to the nature of the Democratic base being more low-propensity (infrequent voters). But the low-propensity nature of this electorate and their lower education levels makes them turn out less than the high-propensity electorate, and Republicans’ highly populist persuasion is actually far more effective with lower education levels. Populist rhetoric can often also easily misconstrue the reality into misinformation, and Republican operatives have heavily invested in purchasing local media outlets and maximizing foreign-language media, such as WeChat in ethnically Chinese communities. This allowed Republicans to break into the traditional Democratic base, while Democrats were not able to, due to voters’ lower education levels, resulting in racial and age depolarization. Hence many suburbs stagnated—especially diverse ones that have been key to Democratic gains—while urban areas were nightmares for Democrats. This less educated electorate also values the economy far more, often viewing inflation in simple terms—inflation meaning prices go up. Ineffective and inept Democratic governance in safely Democratic areas may have also contributed to perceptions that Democrats were struggling to care about kitchen table issues like costs of housing and crime, especially given the indictments of New York City Mayor Eric L. Adams and former United States Senator Bob Menendez of New Jersey on corruption charges. Republican tactics of busing undocumented migrants into safe Democratic cities were successful, in particular due to the housing crisis and many of these cities being sanctuary cities. Yet, they could not afford to house the migrants due to the cost of living, hence creating anti-immigrant hysteria among immigrant communities, many of whom saw undocumented immigrants as “skipping the line.”

There are many other reasons for Harris’s loss that people will assign blame to, such as campaigning with unpopular Liz Cheney or sexism—hard to argue though because female Democratic candidates held up impressively downballot—but overall, it was these shifts in party coalitions that did Harris in. I hate to say it, but Republican strategist Patrick Ruffini was right in his book Party of the People: Inside the Multiracial Populist Coalition Remaking the GOP. And, so were Susie Wiles and Chris LaCivita, with their strategy of “For every Karen we lose, we’re going to win a Jamal and an Enrique.” The Trump coalition enabled the rise of racial—and to a lesser extent, age—depolarization that I had previously dismissed. Trump has received record numbers with Hispanics and Latinos this year—only rivaled by George W. Bush in 2004—along with Asian-Americans. Only, his gains with African-Americans were far more muted. The youth are quite closely divided for the first time in two decades while the old are even more divided, likely for the first time since the mid to late 20th century (the New Deal era). This reshaping of the electorate has the potential to be a profound realignment, as Trump has created a fragile yet formidable coalition that is powerful enough to win presidential elections.

All of the above may seem like a vindication of the polling, but the polling was still wrong, just not in the ways one would expect. The national and battleground polls were directionally quite accurate—leaning Republican within a reasonable polling error—but safe state polling was complete garbage. Polling slightly overestimated Harris’s margins with white voters and underestimated quite a bit Trump’s inroads with voters of color and younger voters. The polling underestimated Trump by double digits in many blue states, such as a Rutgers-Eagleton poll of New Jersey that was off by over 15 percent. Though pollsters may have corrected (or possibly overcorrected) for 2016 and 2020 polling errors, there remain issues as the Trump coalition evolves.

Predicting elections is difficult, and sometimes you do get it wrong. You have to rely on the wealth of knowledge that the past provides, including trends and whatever limited samples you get of the present, and use your “priors” and the samples of the present to predict the election. Sometimes, the outcome is clear; sometimes, it is not; sometimes, there is an upset. Trump’s victory this time around was reasonably predictable, but I did not predict it myself. I did predict, however, that all the battleground states would break for one candidate since they were likely to correlate, and they ended up doing so, just not for the candidate I anticipated. The thing that caught me off guard the most—more than even Trump’s victory—were these levels of racial and age depolarization.

With Trump’s victory, questions remain on what Democrats can do now that they are once again in the minority. It can be argued that many of the Democratic Party’s problems are concentrated in the safe states where they have bad reputations due to poor governance. Democratic governments have grown complacent and overrun with corruption and Democrats in Name Only, and they will likely need to face a reckoning of their own, as this has grown unsustainable. Democrats need to tackle the cost of living, especially by taking measures to increase the housing supply. Ineffective state and local party leadership and governments need to be purged and replaced with people who won't hog gold bars from Egypt (as Senator Menendez did!). Ideology is a lesser issue here, but the causes that make for a great Democratic Party need to be tailored and sold to the communities that they have lost, instead of letting Republicans control the narrative. Representatives Marie Gluesenkamp Perez of Washington and Alexandria Ocasio-Cortez of New York both substantially overperformed Harris, despite having different ideological alignments in the United States House of Representatives—Perez is a member of the Blue Dog Coalition, progressive conservative Democrats, while Ocasio-Cortez is a member of the left-leaning Congressional Progressive Caucus—because their populism successfully marketed the Democratic Party’s platform to skeptical Trump voters while rejecting the ineffectiveness and complacency that many Democratic Party officials hold in safe states. A path forward in the 2026 midterm elections, the numerous off-year elections in 2025 and 2027, and especially the 2028 presidential election must take this anti-establishment ideology and use it against Republicans. Democrats have the keys to success; they just have to be willing to tap them.

From failure comes the initiative to learn from one’s mistakes and refine these predictions further and further. More personally, I was disappointed by Trump’s victory, but it will not deter my interest in politics nor should it deter people from being willing to fight for what is right, which contributes, in part, to my interest. Euphemistically after Trump won, I said “Congratulations to Democrats on their victory in the 2026 midterm elections.” One election loss is not the end; there are off-year elections and downballot races and midterm elections where political capital can be earned and banked before the next presidential election in yet another four years. For many others who are disappointed, much like me by Trump’s victory, it is never joever in America.