A month before the 2024 U.S. presidential election, former President Barack Obama called out “the brothas”—Black men—for their apparent diminished enthusiasm for the Democratic nominee, Vice President Kamala Harris. Instead of entertaining the idea that eroding support for Harris was in part due to policy concerns, in Obama’s view, Black men were “coming up with all kinds of reasons and excuses” because they “just aren’t feeling the idea of having a woman as president.”
Obama’s pointed assumption that the dip in enthusiasm from Black men stemmed from a moral or ethical deficiency felt unusual, particularly given that this demographic (by gender and race) is the Democratic Party’s second-most loyal constituency—Black women being the most supportive. Oddly enough, Obama could not be found giving the same message to other key groups, such as Hispanic and young White men, who were projected to support President-elect Donald Trump in significantly larger numbers than Black men.
Nevertheless, following Trump’s victory on Nov. 5, The Associated Press’s VoteCast survey showed that Harris captured the vote of approximately 74 percent of Black men and 89 percent of Black women—the highest percentage of any racial group in the election. Given the rapidly changing contours of what it means to be “Black” in America, perhaps Obama’s critique of Black men could have benefited from some specificity, a nuance that Obama himself embodies as the multiracial child of a Kenyan immigrant and a white American.
While Trump’s gains with nonwhite male voters will raise broader questions about sexism and racism—particularly anti-Blackness—the key issue that remains sorely overlooked is this: Amid the complex array of reasons underpinning the shifting allegiances of nonwhite voters, there lies a deeper question about how we discuss and analyze the voting behaviors of individuals within these broad and imprecise U.S. racial groups.
In other words, the identities of both white and nonwhite Americans are highly diverse and require proper segmentation. Take, for example, public discourse on “Black Americans,” a category that encompasses generational African Americans—those who descend from U.S. slavery or who have resided in the United States for multiple generations—as well as recent immigrants from Caribbean countries such as Jamaica or Trinidad or sub-Saharan African countries such as Nigeria or Ghana. Many of these immigrants are highly educated before they move to the United States, leading to different social outcomes than their U.S.-born counterparts in some instances.
The same holds true for Latinos, a category that collapses individuals from Caribbean islands (such as Cuba, Puerto Rico, and the Dominican Republic) and individuals from Central and South American nations (such as Mexico, Venezuela, and Honduras) into the same category. Each of these nations has distinct historical experiences, patterns of selective migration to the United States, and systems of racial classification.
Analyzing these groups through single-dimensional categories—such as white, Black, Hispanic/Latino, and Asian—is visually appealing in graphics, but it obscures important nuances within each racial group, namely ethnicity, national origin, and relationships to U.S. empire and its domestic political parties.
The same logic can be applied to the European American majority, colloquially referred to as “white” or “white Americans.” As recently as the mid-20th century, whiteness functioned like an exclusive social club, with membership largely limited to Anglo-Saxon Protestants. Over time, this category expanded to include other groups, such as Catholics from Southern and Eastern Europe (including Italians, Poles, Spaniards), the Irish, and Jews.
Even before these groups were fully incorporated into the hallowed halls of “whiteness,” they were often understood as distinct voting blocs with unique political behaviors. This history underscores the continued value of disaggregating racial categories by national origin to capture important variations in voting patterns.
The flattening of nonwhite racial groups into monolithic blocs perpetuates the assumption that large numbers of these individuals would be naturally repelled by the Republican Party’s racial bigotry and thus find refuge in the ostensibly tolerant, multicultural Democratic Party. The era of Trump and his “Make America Great Again” movement, however, is stress-testing these assumptions.
Additionally, the Democratic Party’s emphasis on race-specific domestic policies and the elevation of minorities into political leadership—while historic, as seen with Obama’s election—has simultaneously fostered a culture where voters are often expected to align with candidates based solely on their immutable racial characteristics.
Take, for instance, Obama’s browbeating of Black men in October, or outgoing President Joe Biden’s comments during the 2020 election, when he told Black American podcast host Charlamagne tha God that Black voters who were unsure about supporting him and his Black and South Asian running mate, Harris, that they “ain’t black.”
This raises an important question: Would Biden have had the gumption to declare that white Americans “ain’t white,” or that Indian Americans “ain’t Indian” if they didn’t vote for him?
Such identity reductionism does two things: It fails to treat minority groups as individuals capable of analyzing their material conditions and exercising political agency—whether by voting Republican, supporting a third party, or abstaining from the electoral process—and it erases nuances within racial groups and thus the ensuing unique voting behaviors of these communities.
Within each racial category are individuals with distinct heritages, lineages, sociopolitical realities, and unique relationships with the U.S. government and its political parties.
Take, for instance, those individuals who are classified as “Black” in the United States. Generational Black Americans—those potentially descended from American slavery and shaped by the legacy of Jim Crow laws and mass incarceration—may relate to U.S. politics and its two major parties differently compared to individuals classified as “Black” who recently immigrated from African or Caribbean nations.
A descendant of Black Southern migrants living in Detroit—whose ancestors languished under American slavery and weathered Jim Crow, redlining, and mass incarceration may feel a stronger affinity toward the Democratic Party and its race-based social policies compared to, let’s say, a wealthy, naturalized Nigerian immigrant in Houston, who may prioritize other issues—such as low taxes and pro-business policies—over policies aimed directly at tackling systemic racial inequality.
Similarly, among Latinos, white Cuban immigrants in Miami who fled Fidel Castro’s socialist regime may find the anti-communist rhetoric of the Republican Party appealing. In contrast, Puerto Ricans—subjects of U.S. empire—living in New York often face similar issues to African Americans (high unemployment, poor housing quality, and police brutality) and therefore may be more attracted to the Democratic Party’s support for social welfare programs and racial equality.
Current data underscores the importance of more precise delineation when examining the voting behaviors of Black and Latino Americans. For instance, a poll published in October by the Carnegie Endowment for International Peace, conducted with a sample of 1,100 registered Black voters, revealed significant differences in voting patterns along ethnic lines—a nuance that is often overlooked in popular discussion about the voting behaviors of the individuals who are categorized and constructed as “Black” or “African American” in the United States.
For example, top-line data showed that 78 percent of respondents had already voted for or intended to vote for Harris, 11 percent went for Trump, and the remaining 11 percent intended to vote for another candidate or sit out the election. When examining the data by gender, it confirms the suspicion that Black men leaned slightly more toward Trump than Black women (15 percent versus 7 percent).
However, analyzing the data along ethnic lines presents interesting divergences. Respondents self-identified their ethnic group (such as generational Black American or descendent of Black American slavery, Black Caribbean, Afro-Latino, Black African, or other). Without controlling for gender, 79 percent of generational Black Americans intended to vote for Harris, compared to 9 percent for Trump. Conversely, among nongenerational Black respondents or other Black ethnic groups—taken together as one category—74 percent planned to vote for Harris, while 15 percent supported Trump. (Pre-election polling data on candidate preferences by ethnicity is not available in the main survey top-line results but can be derived through cross-tabulations.)
When controlling for both ethnicity and gender, partisan support became more pronounced. Among Black American men, 76 percent intended to vote for Harris, while 13 percent supported Trump. Among men from other Black ethnic groups, 68 percent planned to vote for Harris, while 24 percent supported Trump.
These findings align with a recent Brookings article that found that in the 2020 election, foreign-born Black men were more likely to vote for Trump (30 percent), compared to U.S.-born Black American men (11 percent). To be sure, the Carnegie survey was conducted before the election and asked about likely votes rather than actual votes, but the results point to an important phenomenon: There is a slight variance in the voting behavior of Black people of recent foreign heritage compared to U.S.-born or generational Black men.
There is also a strong need for careful disaggregation of Latino voters. AP VoteCast revealed that 43 percent of Latinos voted for Trump, and 55 percent voted for Harris, largely in line with preelection polling that signaled that Trump might mirror former President George W. Bush’s success with Latinos.
While pundits, perhaps dismayed and perplexed by the notable shift in Latino voting patterns, may search for explanations, it is crucial to remember that “Latino” is a broad category encompassing individuals from countries as varied as the Dominican Republic and Venezuela—each with distinct historical contexts, racial compositions, and patterns in voluntary migration.
Preelection polls that disaggregated the Latino vote by national origin reveal these subtle differences. For instance, polling from the Americas Society and Council of the Americas found that Cuban Americans (58 percent) and those from the broadly defined region of South America (41 percent) were more likely to vote for Trump compared to Puerto Rican Americans (37 percent) and Mexican Americans (33 percent).
The complex racial identities within the Latino community further complicate our attempts to place a diverse group of people into neat U.S. racial classifications, as highlighted by a 2019 Urban Institute report that revealed that in the 2010 Census, more than half of Latino people self-identified as “white,” 36.7 percent identified as “some other race,” and 2.5 percent categorized themselves as “Black.”
My research, along with the findings from Brookings, points to the need to include ethnicity or heritage and national origin in any serious analysis of voting behavior. As polling from Pew Research has revealed, the nature and meaning of Black identity in the United States is changing rapidly. Since 2000, the number of foreign-born Black people has nearly doubled, with almost 9 in 10 of those recent arrivals hailing from the Caribbean or Africa.
While Black people in the United States may share phenotypes associated with Africa, it is important to acknowledge that each ethnicity has distinct historical experiences that inform policy preferences and political behavior.
The same holds true for Latinos; as the share of individuals from Latin America continues to grow, collapsing diverse groups obscures important differences. While broad racial labels such as “Black” and “Latino” may serve a purpose in helping demographers categorize and sort a diverse and rapidly changing electorate, much is lost by not acknowledging ethnicity within these larger racial categories.
Just as white voters are not treated as a monolithic bloc—evidenced by the regional and class split in their votes this election—treating Black and Latino voters as a unified bloc without understanding ethnic distinctions could lead to serious miscalculations and oversights in future political strategy.
The post U.S. Election Polls Can’t Tell Us Much About Race appeared first on Foreign Policy.
