One Thing Keeping Democrats Up at Night

By Thomas B. Edsall, NY Times

The composition of the minority electorate in the United States is rapidly changing. This constituency was once dominated by Black voters loyal to the Democratic Party. Now, African American clout has been eclipsed or at least threatened by Hispanic, Asian American and other nonwhite voters whose less firm loyalty to the Democratic Party lowers the party’s Election Day margins among people of color overall.

This multiracial, multiethnic population constitutes one-third of the electorate, according to an article published by the Center for Politics at the University of Virginia, “The Transformation of the American Electorate,” which was written by Alan Abramowitz, a political scientist at Emory.

“Eight months out from the election, polls are still suggesting 2024 will be the largest racial realignment since the Civil Rights Act was passed,” Adam Carlson, a data analyst, posted on X (formerly Twitter) on March 5.

Three days later, John Burn-Murdoch, the chief data reporter for The Financial Times, contended in “American Politics Is Undergoing a Racial Realignment” that

many of America’s nonwhite voters have long held much more conservative views than their voting patterns would suggest. The migration we’re seeing today is not so much natural Democrats becoming disillusioned but natural Republicans realizing they’ve been voting for the wrong party.

On March 15, the polling expert Nate Silver, citing Burn-Murdoch’s racial realignment article, posted “Democrats Are Hemorrhaging Support With Voters of Color” on his Substack.

These claims of a racial realignment in partisan politics have not gone unchallenged.

Brian Schaffner, a political scientist at Tufts who oversees data collection at the Cooperative Election Study, described his views in an email:

What I see is some fluctuation over the past two decades coinciding with unique presidential candidates, no major realignment. A lot of what people are prognosticating about is something that current polls suggest might happen in November, but at this point I don’t think we can say that there has been any kind of major shift yet.

Along similar lines, Jacob Grumbach, a political scientist at the University of Washington, replied by email to my inquiry about racial realignment:

The overall takeaway is that we’ve seen some Latino movement toward Trump in some parts of the country and potentially some Asian American movement as well. It’s an important shift, but it’s uncertain how durable it is, and it’s not unseen in earlier periods, such as George W. Bush in 2004.

There was universal agreement among those I contacted that recent polling data is problematic for the Biden campaign, which is reflected in the RealClearPolitics analysis of the 13 most recent surveys, which, in aggregate, give Trump a 1.7 percentage point lead over Biden, 47.2 to 45.5.

The debate is over whether the adverse trends for Democrats are long lasting and structural or are vacillations unique to the current campaign.

Let’s take a look at the conflicting evidence.

Compare some of the results of the March 10 to 12 Economist/YouGov poll of 1,559 adults with those in the March 9 to 12 Civiqs/Daily Kos survey of 1,324 registered voters.

YouGov found Biden leading Trump 68 to 15 percent among Black Americans, 47 to 36 among Hispanic Americans and 56 to 29 among 18-to-29-year-olds. Civiqs found much higher levels of support for Biden among Black people (79 to 8) and Hispanics (71 to 17), but among 18-to-34-year-olds in the Civiqs survey, Trump had a substantial lead (49 to 36) over Biden.

Carlson has aggregated polling trends for subgroups by combining data collected in February 2024 from 10 polling firms to get a sample size of 11,288 people, including 1,134 Black voters, 1,161 Hispanic voters and 1,003 young voters ages 18 to 29.

The trends in these subgroups provide little comfort to the Biden campaign.

Among Black voters, Biden led Trump by 55 points (73 to 18), far less than his 83-point margin in 2020. Among Hispanics, Biden led by six points (48 to 42), compared with a 24-point advantage in 2020. Among 18-to-29-year-olds, Biden led by eight points (50 to 42) compared with 24 points in 2020.

Despite the erosion of Black, Hispanic and youth support since 2020, Biden remained competitive in Carlson’s data compilation — just two points behind Trump (47 to 45) among all respondents. This was possible because Biden made modest gains among very large subgroups: 1.3 points among 2,014 white college graduates, 0.6 point among 2,103 white non-college grads, four points among 923 voters ages 50 to 64, 1.8 points among the 2,208 voters 65 or older.

In an email, Carlson voiced caution about drawing conclusions based on the aggregated polling data:

We’ve seen zero evidence in recent election results that young voters and Black voters are abandoning voting for Democrats, so all of this is speculation based on polling. Among Latinos the evidence is a bit more mixed, but there’s more electoral evidence from 2020 and some from 2022 that they could be moving right.

Carlson, however, pointed to additional polling trends daunting to Democratic prospects.

Gallup reported on Feb. 7 that

in 2020 Black voters self-identified as +66 Democratic, and in 2023 they’re at +47. They find Hispanics at +12 Democratic now — an all-time low since 2011, but that decline has been more gradual. They’re also seeing a Democratic decline among age 18-29-year-olds (+21 in 2020 to +8 in 2023).

I asked Carlson how he could justify using “realignment” to describe what’s been happening, since that suggests a full-scale partisan conversion of the country or of a major constituency, as in the 1932-36 realignment that saw the electorate go from majority Republican to majority Democratic or the post-civil-rights realignment that saw the white South go from majority Democratic to majority Republican.

Carlson responded:

If what we’re seeing in recent polls regarding shifts among young, Black and Latino voters ends up happening in November, in my view “realignment” is the right term. It won’t be like 1932 or 1964, where the parties essentially swapped coalitions for the New Deal and civil rights, respectively.

Essentially it would be a continuation of the trends we saw in 2020 among Latinos, a sizable but not earth-shattering shift among Black voters (though even in the most pessimistic assessments Biden will still win at least 75 percent of Black voters) and a shift to roughly even among younger voters from a strong Dem advantage.

Carlson had this caveat: “For what it’s worth, I am skeptical that these swings will be this large once all is said and done in November, but that’s neither provable nor falsifiable until then.”

Data from the Cooperative Election Study, which conducts surveys of more than 50,000 voters every election cycle, does not support the case for a realignment of any major voting bloc.

The percentage of Black voters choosing Democratic House candidates does not reveal a consistent downward trend that would signal a slow-motion realignment. Instead, this measure of Black Democratic support shifts back and forth from lows of around 88 percent in 2010, 2016 and 2022 and highs of around 93 percent in 2008, 2012 and 2018.

Asian American support for Democratic House candidates fell from 75.4 percent in 2016 to 64.7 percent in 2022, but the 2022 level of support was higher than it was in the elections of 2008, 2010, 2012 and 2014.

Hispanic support for Democratic House candidates fell from a high of 71.4 in 2018 to 59.6 percent in 2022, but the 2022 percentage was very similar to the margins from 2008 to 2012.

Perhaps most significantly, more detailed election study data breaking down voting trends by race, ethnicity and ideology shows that the defections of Black and Hispanic voters from the Democratic Party to the Republican Party are heavily concentrated among those who describe themselves as conservative.

An estimated 16 percent of Black voters are conservative, and from 2012, when Barack Obama was at the top of the Democratic ticket, to 2022, their level of support for Democratic House candidates fell from 84.2 to 47.7 percent.

Some 32 percent of Hispanic voters identify themselves as conservative. From 2016 to 2022, their support for Democratic House candidates fell from 33.7 to 13.4 percent.

These trends among ideologically conservative minorities lend credence to Burn-Murdoch’s contention that

many of America’s nonwhite voters have long held much more conservative views than their voting patterns would suggest. The migration we’re seeing today is not so much natural Democrats becoming disillusioned but natural Republicans realizing they’ve been voting for the wrong party.

Matthew Blackwell, a political scientist at Harvard, contended that the trend Burn-Murdoch cited is less consequential than it appears to be:

Burn-Murdoch is correct in the column that Black conservatives are increasingly voting for Republicans, but this also misses how the Black electorate has changed since the Obama era.

In 2014, 25 percent of Black voters reported themselves as “conservative” or “very conservative,” and 31 percent reported themselves as “liberal” or “very liberal.” In 2020, only 17 percent of Black voters put themselves in the conservative categories, and 41 percent put themselves into the liberal categories.

So while the connection between ideology and voting has become stronger for Black voters over the last 10 years, the overall voting rate has not changed nearly as much.

There is evidence that a substantial share of Black, Hispanic and other voters from multiracial, multiethnic backgrounds oppose some elements of the Democrats’ liberal social and cultural agenda.

A Jan. 22 to 27 YouGov survey, for example, asked whether it was “morally acceptable or wrong to have a sexual relationship with someone of the same gender?” Forty-four percent of Black voters answered “morally wrong,” and 17 percent said “morally acceptable.”

The same question was posed on the issue of identifying “with a gender different from the gender assigned at birth.” A plurality of Black voters said it was morally wrong, 39 to 13 percent, and Hispanics agreed, 30 to 15 percent.

By even larger margins, Black (51 to 17 percent) and Hispanic (49 to 27) respondents opposed policies that allow transgender athletes to play on sports teams that match their gender identity rather than their sex assigned at birth.

These numbers suggest that some aspects of Democratic liberal orthodoxy contribute to the exodus of conservative minorities from the party.

Marc Hetherington, a political scientist at the University of North Carolina, Chapel Hill, expanded on this line of analysis in an email:

Racial and ethnic minorities are out of step with white Democrats in their worldviews. The former are much more tradition-minded and authority-minded than the latter. White Democrats were enthusiastic about Sanders and Warren, who wanted to blow up the system. Minority Democrats were enthusiastic about Clinton and Biden, whose programs tended to work within the present system.

I suspect this divide has implications for Black and Hispanic men, in particular. For example, the source of traditional authority over generations has been men. However, Democratic leaders argue for greater gender equality, along with equality for all sorts for groups that have generally occupied lower rungs on the societal ladder. This is bound not to sit well with groups that have traditionally held power and influence.

In an email responding to my questions about minority voting, Emily West, a political scientist at the University of Pittsburgh, argued that “nonwhite Americans who previously may have voted Democrat for identity-based reasons are increasingly likely to vote more sincerely according to their conservative ideology or policy preference and thus vote Republican.”

West, Bernard L. Fraga and Yamil R. Velez, political scientists at Emory and Columbia, recently published “Reversion to the Mean, or Their Version of the Dream? Latino Voting in an Age of Populism,” in which they maintained that there have been

significant pro-Trump shifts among working-class Latinos and modest evidence of a pro-Trump shift among newly engaged U.S.-born Latino children of immigrants and Catholic Latinos. These findings, coupled with an analysis of the 2022 C.E.S., point to a more durable Republican shift than currently assumed.

More specifically, Fraga, Velez and West showed that the strongest shift to Trump was among Latinos holding conservative views on crime and punishment, holding restrictionist beliefs on immigration, having the lowest education levels and making the least money.

These shifts do not, in the researchers’ view, constitute a realignment.

“Well over a majority of Latinos still support Democrats, so I hesitate to call it a realignment, but there are segments of the population that have grown more fond of the G.O.P. over time,” Velez wrote in an email elaborating on the paper. “There is much stronger evidence that some segments of the Latino population such as conservatives have shifted toward the G.O.P. and will stay there, but I’m skeptical that this is a broad-based racial realignment.”

Fraga also replied by email to my inquiry: “There is, indeed, a shift toward Republicans among voters of color, but I would characterize it less as a realignment and more a sort of ideological sorting, where the relatively small population of conservative voters of color are now voting for the party more closely aligned with their ideological preferences.”

The three authors concluded their paper by noting that

the Republican gains we describe align with two key processes shaping American politics: ideological sorting and educational polarization. Unlike the general population, these mechanisms have been notably delayed among Latino voters.

Eric Schickler, a Berkeley political scientist and the author of the 2016 book “Racial Realignment: The Transformation of American Liberalism, 1932-1965,” contended by email that “it is premature to say that there is a realignment among nonwhite voters, but the survey results suggest some erosion in Democratic support that is clearly worrisome for the party heading into 2024.”

The polling data “certainly suggests the possibility that something substantial could be happening,” Schickler continued. “I do not think one can just dismiss the evidence that Biden and the Democrats’ numbers with respect to nonwhite voters have been concerning for several months.”

Kirill Zhirkov, a political scientist at the University of Virginia, suggested in an email that one of the factors driving Latinos from the Democratic Party could be Hispanic prejudice against African Americans.

The image of the Democrats as a Black political party, what Zhirkov called “the Democrats-Black schema” has the same effect “among Latinos and Asians,” Zhirkov wrote. “It is possible that as Latinos and Asians assimilate in the United States, some of them also (at least partially) acquire anti-Black prejudice as one of the organizing principles of anti-progressive politics.”

In support of his claim, Zhirkov cited his 2022 paper written with Nicholas Valentino, a political scientist at the University of Michigan, “The Origins and Consequences of Racialized Schemas About U.S. Parties,” and a 2016 paper, “The Political Consequences of Latino Prejudice Against Blacks” by Yanna Krupnikov, a professor of communication at the University of Michigan, and Spencer Piston, a political scientist at Boston University.

The conflict between competing analyses of recent voting and polling trends won’t be resolved until the votes are counted after Election Day, Nov. 5.

In the meantime, what can be made of the data currently available?

There is not yet adequate evidence to proclaim a racial or minority voting realignment. There may be significant and consequential defections from the Democratic Party among Black, Hispanic and Asian Americans, but it is very unlikely that any of these constituencies will cast a majority of their votes or anything close to it for Donald Trump.

It is also possible that the Biden campaign could make up for the losses it incurs among minority voters with improved margins among white liberals and moderates angered by the Supreme Court’s Dobbs decision ending women’s constitutional right to abortion and by the litany of stories describing the inability of women in red states to get medically necessary abortions.

Even so, Biden has his work cut out for him in the coming months. Voting data and polling data are in conflict, which confounds analysis — tiny shifts among white voters can still have an outsize impact. Biden knows he has to raise both the level of his support and the level of turnout among America’s minority voters if Democrats are going to have a decent chance of beating Trump.


Thomas B. Edsall has been a contributor to the Times Opinion section since 2011. His column on strategic and demographic trends in American politics appears every Wednesday. He previously covered politics for The Washington Post. @edsall


Be the first to comment

Leave a Reply

Your email address will not be published.