One consequence of the November elections is a renewed discussion of the Latino vote, specifically whether it is trending back to the Democrats. Despite claims in the news this week that Latinos are abandoning the Republicans or repudiating Trump, the election results are more equivocal than the headlines proclaim.
From a data analysis perspective, the relatively small number of Latino voters in a limited sample of blue states are not comparable to all Latinos. They may not even be representative of Latino voters in those states in 2024, as turnout is lower in odd-year elections. And to believe that the 2025 election returns are indicative of the Latino vote in 2026 and 2028 is to rely on crystal balls as much as social science.
More theoretically, we need to understand why the Latino vote has been in flux for a decade. Rather than imagining Latinos as constituting a minority group that prioritizes collective concerns, we need to understand them as individuals who are moving to the mainstream. Their voting decisions will therefore increasingly reflect the values of all Americans, as we saw in 2024 and 2025, when the economy and the cost of living were front and center.
This is bad news for the Democrats because it suggests that Latinos are following the assimilation path of the Italian, Irish, Polish, and other ethnic immigrant groups of the previous century. While many academics and activists thought and hoped that Latinos would adopt the “linked fate” perspective of African-Americans, which has anchored this electorate in the Democratic Party by promoting a group rather than an individual lens on politics, the reality has disappointed progressives.
If I were a headline writer, I would choose “Latinos Move with the Mainstream” rather than “Latinos Repudiate the Republicans” to describe the recent elections. Only the former encompasses the fundamental changes we are seeing in the Latino electorate. Latinos who voted Democratic in Virginia and New Jersey were not permanently rejecting the GOP or Trump; they were reacting to the real problems of the economy, just as many did in 2024. This does not constitute a renewed loyalty to the Democratic Party—a loyalty that was taken for granted and badly repaid for decades.
Over time, the Latino electorate will likely find its level in the center to center-right political space, because that is the American mainstream. Nevertheless, from election to election, it may move toward one party or the other depending on the economic and political context.
None of this is to deny that 2025 was a bad year for Republicans or that the party may be heading for “a thumping” in the midterm elections. The November defeats were also consistent with the results of this year’s special elections at the federal, state, and judicial levels, where Democrats outperformed Kamala Harris’s November 2024 tally by about 15 percentage points. That Latino voters are moving with this emerging blue wave, but not declaring their metaphysical loyalties to either party, will be cold comfort to defeated Republican candidates.
Nevertheless, we should differentiate between long-term vs. short-term trends in the Latino vote. Because some Latinos who previously voted Republican chose to stay home or vote Democratic in New Jersey and Virginia does not mean this electorate has stopped its pro-GOP glide or that Latinos are returning to the Democrats. Both parties overinterpret the 2025 elections at their peril.
What can we learn from the 2025 races?
If Democrats made meaningful gains among Latino voters, this would be joyous news for a party that has encountered many disappointments and defeats in the new century. It might even revive a flicker of hope in the theory of “demography as destiny,” which held sway for decades in American politics. It assumed that demographic change would lead to a critical mass of minority voters who joined with white progressives in a “rainbow coalition” that ushered in a new era of Democratic Party dominance.
However, in a classic example of “the slaying of a beautiful hypothesis by an ugly fact”—at least from the progressive perspective—election returns showed growing Latino support for Trump and the Republican party since 2016. My colleague Álvaro Corral and I challenged the conventional wisdom and documented this electoral shift, which has been a disaster for progressive hopes of reshaping America.
This change has been evident in both presidential and midterm elections. After years of denials by Democrats and doubts by Republicans, the argument that “Latinos are moving to the GOP” has now become part of the conventional wisdom.
The 2025 elections may have created a “vibe shift” in the punditry after a year of Democratic despair, but off-year elections are few in number and not easily comparable to other contests. If the 2021 Virginia and New Jersey gubernatorial elections had been followed in 2022 by US Senate elections, which they were not, we might draw some tentative conclusions about whether the voting patterns in Virginia and New Jersey off-year contests continue into the next statewide election. While some are contrasting the recent elections with the 2024 federal elections in those states, these two sets of contests feature different candidates and issues, not to mention variations in voter turnout, political expenditures, and media coverage.
In the following sections, we move beyond media and pundit hot takes to learn what we can from the two main contests of 2025: the Virginia and New Jersey gubernatorial elections. While other elections also took place, particularly in New York City (mayor), California (Proposition 50), Georgia (Public Service Commissioners), and Pennsylvania (retention of State Supreme Court justices), either the data do not exist for useful subgroup analysis or the elections were too unusual to have broader implications.
Virginia governor
According to the CNN-SSRS exit polls, Democrat Abigail Spanberger won 67 percent of the Latino vote vs. 33 percent for Republican Winsome Earle-Sears. This 2-1 ratio will sound familiar to those who follow Latino politics, as before the Trump era, it was the traditional rule-of-thumb estimate for the Latino vote. Even when Republicans such as Ronald Reagan and George W. Bush did unusually well among this electorate, the media-pundit-academic class saw this ratio as the baseline to which the Latino vote would ultimately return.
The actual Latino vote, to the degree we have reliable data on it, was always more complicated than this implied. Today, many hope that Latinos will return to this ratio, and that the Trump years will prove to be outliers, but the steady growth of Latino support for the GOP across the past five federal elections has no precedent and cannot be considered a blip.
Was 67 percent for Spanberger high or low for a Democrat? It is problematic to make comparisons with the 2024 federal election, which had higher turnout (70.5 vs. 54 percent). Even as many state elections have become “nationalized” in recent decades, a presidential contest will draw more and different voters than an off-year gubernatorial election.
A better comparison might be the 2021 election, in which the Democrat Terry McAuliffe won 66 percent of the Latino vote while Republican Glenn Youngkin won 32 percent. From this perspective, the Latino vote changed little across the past four years. Rather than a tale of Latinos “abandoning” the GOP, this comparison suggests continuity, even in a bad year for Republicans.
Exit poll data for 2025 indicate it was whites, who constituted over 70 percent of voters, who made the difference. McAuliffe received 38 percent but Spanberger won 47 percent. In addition, the black vote (17 percent of the electorate) increased 7 percentage points for Spanberger, and the Asian vote (4 percent of the electorate) increased by 13 percentage points to reach 80 percent for Spanberger.
In sum, every major Virginia electorate except Latinos moved in a pro-Democratic direction from 2021 to 2025.
We might also note that the Latino vote was similar in the race for lieutenant governor. Democrat Ghazala Hashmi received 2 percentage points fewer Latino votes than did Spanberger (65 vs. 67 percent) while the Republican John Reid received 2 percentage points more than did Earle-Sears (35 vs. 33 percent). Reid also did slightly better among black and Asian voters.
The Democratic candidate for attorney general, Jay Jones, not unexpectedly received the lowest margin of the three Democratic statewide candidates, including among Latino voters. He won 61 percent while incumbent Republican Jason Miyares won 39 percent. This may well be due to the Jones texting controversy and suggests that even in a very good year for Democrats, the candidates themselves can make a difference for Latino and other voters. We should also note that Miyares is of Cuban heritage, which may have increased his Latino vote share.
Lastly, why are pundits even discussing the Latino vote in Virginia? Not because it affected the actual outcome. According to the exit polls, Latinos were 5 percent of the electorate, and as Spanberger won by 15 percentage points, she would have won in a landslide even if no Latinos voted. If the reason is that Virginia Latinos are seen as indicative of Latinos nationwide, we should note that while 60 percent of all US Latinos are of Mexican origin, Central and South Americans are predominant in Virginia. If the reason is that their 2025 votes are thought to say something about the Latino vote in 2026 and 2028, then we might be getting ahead of ourselves.
As every Virginia county shifted Democratic in 2025 in comparison to 2021, and almost every county did so in comparison to 2024, this suggests that the election results do not need any Latino-specific explanations. A broader wave was at work, which Latinos surfed but did not cause.
New Jersey governor
In contrast to the nail biter in 2021, Democrat Mikie Sherrill easily defeated Republican Jack Ciattarelli by 56 to 43 percent. The exit poll found that while the white vote was almost split (47 to 52 percent), the margins were much higher among black (94 to 5), Latino (68-31), Asian (82-17), and “other” (54 to 43) voters.
Without exit poll data in 2021, it is not easy to understand how the Latino vote may have changed. One helpful question in the 2025 exit poll is previous vote in the 2024 presidential election. This helps to determine how much the recent results reflect changes in voter turnout.
As Nate Cohn suggested, “The Democratic rebound in heavily Hispanic parts of New Jersey, for instance, could be mostly or entirely driven by changes in turnout.” This is because the exit polls show that the Latinos who voted in 2025 indicated that they supported Kamala Harris by a much larger margin (25 percent) than did the actual Latino vote at the time (9 percent). This suggests that quite a few Latinos who voted for Trump last year decided to stay home this year, a pattern that may or may not repeat in 2026 or 2028.
Another approach is to examine voting changes in cities and counties with large Latino populations, but this raises the ecological inference problem, as we cannot be certain which individual voters are responsible for aggregate shifts. Nevertheless, such jurisdictions did vote more Democratic this year than last year, although a comparison with 2021 reveals mixed results. Among nine municipalities with populations at least 60 percent Latinos, three showed no substantive partisan change (within 5 points) from 2021 to 2025, three showed Democratic gains, and three showed Republican gains.
Lastly, every county in New Jersey swung to the Democrats in comparison to both 2021 and 2024, which again suggests that treating Latinos as a sui generis group is problematic. Many demographic groups undoubtedly swung for the Democrats, but the big story was not any subgroup movement but the economic factors that worked against Joe Biden in 2024 and now Republicans in 2025.
The economy or immigration?
As I observed in May of 2025, “Democrats have been paying the price since 2016 for assuming that demographic change would automatically accrue to their benefit. Republicans might think that recent trends are the wave of the future, but they may find that a recession, inflation, and mass deportation can change the direction of travel fast.”
In the recent election, the Democrats clearly reaped the benefits of voter concerns about the economy and the cost of living, just as Trump did in 2024. While the party is interpreting the election results as great news, they cannot “celebrate” (to use a favorite progressive term) an electorate that shifts with economic performance; this will work against them as much as for them. What they need, but do not have, is a Latino electorate that behaves much as do African-Americans: overwhelming Democratic, and with only relatively small shifts depending on economic and political circumstances.
What about the immigration issue, which Democrats have long hoped would attract Latino votes? With polling indicating Latino unhappiness with Trump’s immigration policies, did this help the Democrats? One caveat is that the overall American public has become more favorable toward immigrants and less supportive of the administration’s migration policies over the past year. In other words, Latino views have moved in the same direction as those of non-Latinos, which is in line with the thermostatic model of politics in which Americans react against policies that are seen as too far from the mainstream.
While Latinos are more likely to be personally affected by ICE enforcement policies, especially in light of a Supreme Court “shadow docket” decision that gives a green light to certain types of profiling (for now), a pre-election survey sponsored by UnidosUS (formerly the National Council of La Raza) found Latinos are responding to economic issues far more than immigration and the border. This is consistent with broader public opinion trends; an Associated Press story noted that “Voters prioritize the economy above immigration and crime in Election 2025.”
Once again, rather than finding that Latinos hold unique perspectives that are contrary to the mainstream, we see how their views often move in tandem with those of the American public. I have previously termed efforts to understand such parallels as “The Politics of Similarity,” which is contrary to the standard academic-media-pundit attempts to find and highlight differences between Latinos, whites, and other groups.
Beyond the bloc
We now see many post-election analyses that are reminiscent of the GOP’s reaction to the defeat of presidential candidate Mitt Romney in 2012. This loss, a shock to many in the party, led to the famous “autopsy” report. The media and pundit class are now conducting their own mini-autopsies, often featuring the Latino vote.
Jonah Goldberg recently asked, “Has Trump Already Lost the Latino Vote?” in the sense of “Who Lost China?” But trying to understand Latinos as a constituency that can be “won” or “lost” is like asking who won or lost the Italian-American vote in the twentieth century. Over time, Italian-Americans moved into the mainstream and made up their own minds, weighing the same issues and candidates considered by all Americans, just as Latinos are doing now.
