The Real Reason Democrats Can’t Shake Trump in 2024 Polls

Not just the United States, but most of the democratic world has witnessed a major realignment over the past decade.

Political divisions, which once ran along the lines of economic class and occasionally race, increasingly began to reflect educational differences instead. Culture, which in the past had been less important than one’s economic class and derived from specific ethnic and religious traditions, increasingly became defined by class itself. Class, meanwhile, rather than being determined by income, became a matter of education.

Which religion you are also increasingly has begun to matter less than whether you are religious. In 2004, it would have been safe to bet that a devout Muslim mechanic in Detroit and an Evangelical Christian small business owner in Tennessee would vote in opposite ways. Now, however, the fact that they are religious and do not work in the globalized service industry means they have more in common with each other than with their children who may have gone off to Oberlin or Berkeley.

This shift has had multiple impacts, but one of the greatest has been on the gap between the perception of politics among elites and the actual reality of how the electorate at large perceives the political landscape.

Nowhere is this more obvious than when it comes to the cases of Joe Biden and Donald Trump. Perhaps no president since Nixon has produced such polarized views between the elites and the country at large. Elites may have detested Donald Trump, but they understood how popular he was with large swathes of the country, even as they came up with increasingly bizarre excuses for it. Nixon’s popularity came as a genuine shock to those who knew almost no one who liked him within their bubbles.

The same is true of Joe Biden. Within the elite bubble, Joe Biden is seen as a largely successful president who has ushered through key legislation and presided over an activist administration. Inflation is falling, U.S. manufacturing is bouncing back from the pandemic, and Russia is checked in Ukraine, they insist. Things are on the right track.

This view is shared not just by Democrats, but by Republican elites who are apt to adopt a fatalistic view when it comes to their party’s prospects against a president they view as formidable against all evidence.

This is one reason why there has been such a shocked reaction to the recent Times/Siena poll which shows the general election tied between Joe Biden and Donald Trump.

The tie itself is not interesting – other polls have shown similar results. What is interesting is the work the pollsters, commissioned by the Upshot, the paper’s in-house political analytics division, put into trying to figure out exactly why it was tied.

For many modern pollsters, the goal is to get a topline number which can be published. The demographic breakdowns are secondary. The goal with this poll, however, was the opposite. It bent over backwards to get representative breakdowns. In particular, it sought to correct one of the key flaws in polling since 2016: not weighting for education.

The results provide a clear picture of why coverage of the 2024 race is so detached from the data in America today. The electorate is clearly divided into two universes, and the media overwhelmingly inhabits just one of them.

The results for the Times/Siena poll do not paint a picture of an electorate that believes America is on the right track. Only 23% of respondents agreed that it was. 65% felt America was moving in the wrong direction, including 38% of Democrats and two-thirds of independents.

These numbers showed little differentiation among age, income, or even race. Those over 65 were the least apt to feel the nation was on the wrong track, but 62% still believed it was. Hispanics had the highest percentage believing the nation was on the right track, but this only reached 33%. Democrats were the only group to give a net-positive response, and that was only by the narrow margin of 46%-38%.

Joe Biden’s numbers were equally dire. Thirty-nine percent either slightly or strongly approved of his performance as president compared with 54% who did not. Biden was underwater with every age group, and every racial group except for African Americans, where he was up 60%-26%.

There was also a limited income gap. Biden was down 38%-54% among those who made less than $50,000 a year, 38%-57% among those making $50,000-$100,000 per year, and 42%-52% among those making over $100,000 a year. If anything, the Democrats, the traditional party of the “working class,” seem more popular with higher earners.

But there is one demographic breakdown that does provide a clear differentiation in Biden’s popularity. Among voters without a bachelor’s degree, Joe Biden was down 35%-60%. Among those who had received a bachelor’s degree, Joe Biden had a net positive approval rating, 47%-45%.

There is a hint here which explains the major media narratives about American politics. While the media without a doubt lean left, the greatest homogeneity comes from education. Even Republicans who work in the media and staff outlets like Fox inhabit a world almost entirely made up of those with college degrees, which may explain why coverage, even on a network like Fox, seems to exist in a universe where Joe Biden has a 47%-45% approval rating.

The narrative that Biden, if divisive and not loved, enjoys a solid base of support better reflects the world of the college-educated, where he does indeed have substantially more support. In that world, Joe Biden is favored for reelection in general, and by a lopsided margin in a rematch against Donald Trump. Biden, so the narrative goes, may not be loved, but Donald Trump is outright hated.

The Times/Siena numbers do not support this. Donald Trump’s numbers are not good (although Trump has repeatedly outperformed polls) but they are marginally better than Biden’s approval rating, at 41%-55%, and very close to Biden’s own personal numbers of 43%-54%.

The answer to the question “which universe is the media reporting on?” once again lies in the educational gap. Among voters without college degrees, Trump is viewed favorably, 49%-46%. But among those with bachelor’s or above, Trump is viewed unfavorably by lopsided 27%-70%.

The media narrative thus makes perfect sense in a world where 70% have an unfavorable view of Donald Trump, compared to only 45% who view Biden that way. The problem for the media, as in 2016 and 2020, is that their world is not the entire world.

When combined, the two worlds seem to almost cancel out, at least in this poll. That may be a product of the Upshot’s interest in crosstabs. As previously noted, they were far more interested in who was supporting each candidate than in whether Trump or Biden led by 1.5%, which after all is guesswork based on unknowns about turnout. The result is that we can get some idea of what a tied race would look like.

Such a race is one where the educational gap is replicated. Among voters with a bachelor’s degree or higher, Joe Biden leads by 22%, 54%-32%. Among those without a bachelor’s degree, Donald Trump leads by 13%, or 49%-36%.

These gaps are higher than the margins among any age group, the most lopsided of which, 18-29-year-olds, favor Biden by 10%, 47%-37%. It also entirely dwarfs income groups. The largest margin for either candidate is Joe Biden’s 47%-41% lead among those making over $100,000 a year.

The only group where the margins are higher are among African American voters and Evangelical Christians. In the former case, however, Biden’s 71%-12% lead nevertheless has Trump receiving a higher share of the African American vote than he did in 2020. In fact, the wider trends in the poll could be seen as a continuation of the 2022 midterms, where Democrats held their own or even made gains in areas in suburbs and college towns with service-based economies, but Republicans made gains, even in the bluest parts of the country, among working-class Democrats who found themselves uncomfortable in an increasingly elitist party. This manifested in large GOP gains in California and New York, while Democrats did better in the Upper Midwest, Georgia, Arizona and Colorado.

One thing which is relevant to the increasingly dominant role of education is that the new culture wars do not actually seem to be built around policy differences. While there is some correlation between degree attainment and “social issue” positions, it is small and seems to be more correlated with age.

When it comes to a proposed 15-week federal abortion ban, those with bachelor’s degrees oppose it by a margin of 36%-58%, while those without also oppose it, but only by 40%-50%. Same-Sex marriage is supported by 79% of those with bachelor’s degrees, but also 65% of those without. Military support for Ukraine is favored by 65% of those with bachelor’s degrees, but also 51% of those without. While 68% of voters without bachelor’s degrees believe America is on the wrong track, so do 60% of those with bachelor’s degrees or above. More than 70% of voters with bachelor’s degrees believe the economy is fair or poor.

There is more agreement among these groups on issues than many might expect.

The obvious conclusion is that what we are seeing is a new type of identity politics, whereby the Democrat Party is winning voters with bachelor’s degrees because it is messaging to those voters that their education means they should vote Democrat. It is a new type of class solidarity.

This may also explain the solidarity within the media and the federal civil service, including those at the DOJ and FBI, when it comes to Donald Trump and U.S. politics.

This is not the product of some “Deep State” conspiracy. The trouble is much simpler. Those positions draw almost entirely from a demographic on one side of the great cultural chasm which divides politics in the 2020s, and we should expect them to behave homogenously as a result.

There are likely to be limits to this trend in the long run. The educational gap is a relatively “white” phenomenon. While higher rates of education correlate with Democrat support among white voters, they have a much more limited correlation among non-whites in a Trump-Biden matchup, and a negative one when it comes to Biden’s approval rating. While whites with bachelor’s degrees disapprove of Biden 45%-49%, and those without 29%-67%, with non-whites the numbers are 47%-46% and 50%-34%, respectively.

While Biden does 34% better among whites with degrees than among those without, he does a net 15% worse when it comes to non-whites. If attending college and entering the “woke force” makes white voters more Democrat, it seems to have the opposite effect on non-whites.

This might seem to be a uniquely American phenomenon if not for the prior example of the United Kingdom. The United Kingdom saw the same educational polarization under Tony Blair, with professionals in the service sector and degree-holders moving towards Labour and the Liberal Democrats.

However, the inverse happened with many of the children of African and Asian immigrants, with the result that much of the professional base of the Conservative Party is made up of non-white individuals who are on average much further to the Right than their white counterparts. These include the Prime Minister Rishi Sunak, the Home Secretary Suella Braverman, her predecessor Priti Patel, and the Trade Secretary Kemi Badenoch, who has been one of the leading social conservatives in government.

We already see signs of a similar trend with the prominence of Asian, Latino, and African American conservatives. It may well be that educational polarization will result in a world where much of the elite are Democrats, but the Republican professional class is actually more reflective of America as a whole.

In the meantime, it goes a long way to explain the bubble much of America’s political coverage takes place in.

Daniel Berman is a frequent commentator and lecturer on foreign policy and political affairs, both nationally and internationally. He holds a Ph.D. in International Relations from the London School of Economics. He also writes as Daniel Roman.