In the 2020 US Census, the percentage of Hispanics identifying as white alone dropped by over half. I'm curious about the differences between those that checked white as their race, "some other race" (usually a write in for "Hispanic" or related terms), or neither of those (most often "two or more races" or "Native American").

In the United States as well as parts of Latin America from what I've read, identifying as non-white is often on average associated with more liberal views because of history with European colonialism and other things. Has there been any research within the Hispanic category in the US on political leanings along these lines?

Note: I'm not specifically asking "what percentage of Hispanic voters who selected some other race on the 2020 Census voted for Biden in 2020's presidential election" or anything else like that though such a statistic could be helpful.

1 Answer 1


In the United States as well as parts of Latin America from what I've read, identifying as non-white is often on average associated with more liberal views

Your intuition is correct. Racial identification drives voting tendencies in the direction you suspect. Religious adherence, and whether someone is native born or not, are also significant drivers of voting tendencies.

One of the notable trends in the 2020 election was a shift away from Democrats and towards Republicans among Hispanic voters in several specific areas among non-college educated Hispanic voters, especially, the Miami, Florida metro area, near the Texas border with Mexico, in certain already conservative leaning areas in California, and arguably in Southern New Mexico (although more than one facto was at play in that case). According to the second link in this paragraph (to the New York Times):

Mr. Trump’s most sizable gains outside of Miami were in the Rio Grande Valley in the predominantly Hispanic areas along the border with Mexico, including Hidalgo County, home to McAllen.

enter image description here

For example, in the 2020 U.S. House races, Democrats lost 1% of their "safe Democratic" seats (as rated by Cook's Political Report) and about half (± one seat) of the "Vulnerable Democrat" seats. Five of the seats (three in California, two in Southern Florida and one in New Mexico with a somewhat more complicated story) seem to reflect GOP strength in 2020 with certain Hispanic communities. Democrats lost CA-21 (open seat, Fresno, 71% Hispanic), CA-39 (safe, suburban LA, 34% white, 33% Hispanic, 29% Asian), CA-48 (Laguna Beach, 59% white, 20% Hispanic, 18% Asian), FL-26 (Miami, 72% Hispanic), FL-27 (safe, suburban Miami, 72% Hispanic), and NM-2 (Southern NM including Las Cruces, 27% Hispanic, 6% Native American).

In all three of those places, the Hispanic voters shifting to vote for Republicans overwhelmingly identify as white first, and Hispanic second, even in cases where they are fully bilingual (a particularly good profile can be found at Politico). (It isn't entirely clear if this self-identification reflects predominantly European ancestry as it does for Cubans in Florida, or if it is a matter of self-perception among people with significant Mestizo ancestry.)

Religion And Foreign Born Status

These populations also tend to have a low share of first or 1.5 generation immigrants. (A "1.5" generation immigrant is someone who like a first generation immigrant is foreign born, but migrated to the U.S. as a child where they were raised by first generation immigrants, but are fully assimilated in U.S. culture and usually fluent in both the language of their parents and American English.)

Another factor distinguishing the voting tendencies of Hispanic voters both in Latin America and in the U.S. is religious affiliation. On average, Hispanic Protestants (most often Pentecostals) are significantly more conservative politically than Hispanic Catholics, and non-religious Hispanics are less liberal than Protestant Hispanics but only moderately so, although this last distinction in confounded by the fact that foreign born Hispanics are overwhelming Christians of some sort or another, while non-religious Hispanics are much more likely to be U.S. born.

Latino voters were 50% Catholic, 25% Protestant/Other Christian, 6% Other Religious Affiliation and 20% No Religious Affiliation in the CNN Exit polls in 2020 and according to the same poll, Latino support for Biden (in the 18 states with statistically significant samples) ranged from a low of 53% in Florida to a high of 75% in New York State, while Latino support for Biden was highest among Catholics 71%, then those with no religious affiliation 55%, and then Protestants 51%. According to the Associated Press Votecast Exit polls 65% of Latino Catholics voted for Biden.

A significant share of the U.S. born Hispanic population in Texas, New Mexico, and Southern Colorado have ancestors in the same place who lived there before the area was part of the United States. Indeed, the dialect of Spanish spoken in New Mexico and Southern Colorado (see also here) by these populations retains certain words and grammatical features from the colonial era that have not been retained in any other Spanish speaking place. Also, a century ago, these pre-U.S. origins Hispanics included a significant share of Sephardi Jews, although that religious affiliation is now extremely rare in these places due to migration and conversion.

  • "Racial identification drives voting tendencies". Or is it the other way round? Voting tendencies (i.e. political beliefs) drive self-identification...
    – Zeus
    Commented Mar 24, 2022 at 1:59
  • @Zeus Correlation is not necessarily causation, so you could make that argument, looking at the 2020 election alone. But the racial identification is temporally prior to the voting act by generations, and the voting tendency changed between 2016 and 2020 based upon how those issues interacted with racial identification differently in the different elections. We know, factually, that there has not been a big change in racial identification away from identifying as racially white in any of these communities in recent years. See particularly, the Politico article that I link discussing this.
    – ohwilleke
    Commented Mar 24, 2022 at 2:03

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .