December 12: What Do Trump’s Latino Gains Mean for Democrats?
Amid all the conflicting takes on how Donald Trump won the presidency after losing it in 2020, there’s a strong consensus that gains among Latino voters mattered a great deal. I examined this CW at New York:
Definite judgements about how the 2024 presidential election turned out should await voter-file based data that won’t be available for some time. But it’s pretty clear one of the biggest and most counter-intuitive shifts from 2020 was Donald Trump’s gains among Latino voters. Yes, there’s a lot of controversy over the exact size of that shift. Edison Research’s exit polls (which have drawn considerable criticism in the past for allegedly poor Latino voter samples) showed Kamala Harris winning Latinos by a spare 51 to 46 percent margin, while Edison’s major competitor, the Associated Press VoteCast, showed Harris’s margin at a somewhat more robust 55 to 43 percent. Other estimates range up to the 62 to 37 percent win claimed for Harris in the American Electorate Voter Poll.
But most takes showed sizable Republican gains from 2020, and for that matter, Trump did measurably better among Latinos in 2020 than in 2016 (Pew’s validated voter studies showed Trump winning 28 percent in 2016 and 38 percent in 2020). As Equis Research puts it, “this looks and sounds like a realignment.” And while close elections lend themselves to exaggerated focus on specific voter groups, the size and potential future magnitude of the Latino vote make it a natural source of deep concern for Democrats and optimism for Republicans. A New York Times analysis of the startling losses in vote share by Democrats in urban core areas in 2024 concluded that the most consistent pattern was significant Latino populations, which also showed major Republican gains in non-urban areas as well.
To state the obvious, it’s surprising that a politician so associated with nativist rhetoric and policies as Donald Trump is setting records for support in what has traditionally been a Democratic “base” constituency. Is this a trend that would have occurred without Trump leading the GOP, and if so, could it actually intensify once he’s left office for good?
It’s important to understand that this isn’t the first time a pro-GOP Latino “wave” seemed to be developing. While there was immense controversy over the exact numbers (in part because of uniquely flawed exit polls in that particular year), George W. Bush appears to have won about 40 percent of this vote, beating Ronald Reagan’s earlier record of 37 percent in his 1984 reelection landslide. According to the more reliable exit polls in subsequent elections, the GOP share of the Latino vote dropped to 31 percent in 2008 and then to 27 percent in 2012. Some reasons for this reversal of the trend that appeared in 2004 weren’t that hard to discern: the Great Recession that appeared late in Bush’s second term hit Latino households really hard, even as Republicans retreated rapidly from Bush’s support for comprehensive immigration reform (by 2012, Republican nominee Mitt Romney was promoting policies to make life so unpleasant for undocumented immigrants that they would “self-deport”).
But it’s possible that what we are seeing now is the resumption of a slow drift towards the GOP among Latinos that was temporarily interrupted by the Great Recession and a nativist uprising among white Republicans. Whatever unhappiness Latinos felt towards Trump’s immigration views was pretty clearly offset by economic concerns, especially among younger Latino men, who broke towards Trump most sharply. As happened during the Great Recession, the economy mattered most, and the combination of inflation (especially in housing costs) with tight credit eroded already-thin Democratic loyalties. As the above-mentioned Times analysis showed, defections to Trump happened all across the landscape of the Latino electorate, not just among more traditionally Republican-prone groups as Cuban Americans or South Americans. The question as to whether this is a party accomplishment rather than a personal accomplishment by Trump is an open one; Democrats did significantly better among Latinos in down-ballot races in 2024.
A general trend towards a more politically diverse Latino voting population makes some intuitive sense. As former immigrants slowly give way to native-born citizens, particularly those who are entering the middle-class en masse, it’s logical that identification with “the party of immigrants” will decline. Latinos who embrace conservative evangelical–and especially hyper-conservative pentecostal–religious practices also has helped intensify right-leaning cultural attitudes. We may never return to the days of reliable two-to-one Democratic advantages in this community, particularly as young voters who are especially alienated from traditional party loyalties move into the electorate.
While Democrats should be worried about the future of Latino voting behavior, Republicans have no reason for complacency. It’s now Trump and the GOP who are fully responsible for economic conditions which could turn out to be much worse than vague positive memories of the first Trump administration might suggest. And while (as some polling indicates) Latino citizens may have a negative attitude towards the recent surge of migrants that has become so central to Trump’s grip on his MAGA base, it’s less clear the mass deportation regime Trump has pledged to undertake immediately is going to go over well among Latinos, even those who voted for him. A recent Pew survey showed that Latinos were significantly less supportive of a major deportation program than other voters. And if the Trump administration pursues deportation round-ups in a cruel and ham-handed way (which elements of Trump’s base would welcome as a virtue rather than as a vice), or by methods that affect Latino legal immigrants and native citizens (most likely via ethnic profiling by law enforcement officials), we could see a pretty significant Latino backlash.
In other words, while some Latino trend towards the GOP may be inevitable all things being equal, it’s hardly guaranteed and could be sharply reversed. For their part Democrats need to get more serious about Latino voter outreach (particularly among young men) and identify (and learn to explain!) an economic agenda that prioritizes the practical needs of middle-class folk from every background.
Given the wide variety of methods used to choose delegate in the Democratic primary (open primary election, closed primary election, caucuses, and even a mixture of these methods within the same state), how can we determine an accurate popular vote total? If the Democratic party had a national primary election in all states on the same day, with the same rules regarding party registration for participation in each state, then the popular vote total would be meaningful.
Changing from a primary open to all registered voters to a closed primary where only voters registered as Democrats can participate, for example, has a significant impact on the total number of votes cast for each candidate. Likewise, changing from a primary election where voters have 12 hours to cast their ballots to a caucus where voters must arrive at a specific time would change the total number of votes cast. And Mr. Kilgore states that vote totals are not even available for several caucus states. Voter participation levels would also be different for an election held on a Tuesday in January in New Hampshire compared to an election held on a Saturday in May in Guam due to such factors as weather and work schedules.
The presidential nominee will be chosed by delegates at the national convention in Denver. When Sen. Clinton or Sen. Obama can win the support of a sufficient number of delegates, then she or he will be the nominee.
I understand that focusing attention on the “popular vote total” is an attempt by Sen. Clinton’s campaign to persuade undeclared superdelegates that she deserves the nomination. Because of the diversity of election and caucus rules and primary election dates, I would argue that the “popular vote total” is an arbitrary, and therefore meaningless, number.