As the presidential campaign gains momentum, the Hispanic vote has become the subject of focused attention by pollsters who are reporting substantially different levels of support for candidates Trump and Biden. 

A recent poll sponsored by the Texas Hispanic Policy Foundation reported a rather surprising result: 47% preferred Biden while 38% preferred Trump – a 9 percent difference.  Based on this outcome as well as comparison’s to Hillary Clinton’s Latino support in 2016, various news sources are portraying Biden as losing ground with the Hispanic electorate or running a campaign that is out of touch with the needs of this important voter segment.  The online poll was conducted by YouGov using their proprietary methodology that has been widely used in past years but not necessarily in Hispanic-targeted polling.

The results from the Texas Hispanic Policy Foundation poll are especially surprising because they contradicted the findings of various recent polls that included sub-samples of Latino voters that generally found that Biden has enjoyed a double-digit lead over Trump. [1] 

Recent Polls Including Latino Sub-Samples

                   (Percent)

 Curiously, this summary of recent polls by Kuff [2] omitted two other key polls conducted by survey organizations that have historically used the highest methodological standards when conducting polls of U.S. Latinos.[3]  For example, recent polling by Latino Decisions in six battleground states – including Arizona, Florida, Texas, North Carolina and Pennsylvania — confirmed a similar double-digit Biden advantage over Trump [4] while a recent poll of U.S. Latinos by Pew Research Center [5] also confirmed Biden’s double-digit lead over Trump. 

Lastly, an ABC/Washington Post poll released on Sept. 23, 2020 of likely voters shows that, among Latinos, Biden maintains a double-digit lead over Trump (61% vs. 34%) and a smaller double-digit lead in Florida (52% vs. 39%). [6]

Who is correct?  The answer to this question has important implications for the political messaging that is conveyed about Hispanic voters between now and election day since it could influence voter turnout. Is Biden really losing ground among Latino voters, or should we consider the poll sponsored by the Texas Hispanic Policy Foundation an anomaly?

Over the past 45 years, I have taught survey research methods at four North Texas universities and developed an interest in conducting post-mortems on the methodology employed in studies of multicultural persons whose results were conflicting or controversial. Over the years, this interest has included challenges to the Nielsen Hispanic television ratings, measurement bias in college admissions tests, biased surveys in municipal elections, estimates of COVID-19 mortality rates for Blacks and Hispanics, biased crime reporting in urban areas, and site location methods that reinforce redlining practices in communities of color.  The goal of these post-mortems is to help the general public understand which one of these conflicting or controversial studies is more deserving of our confidence using criteria that I have discussed elsewhere in some detail. [7]In my opinion, I would place my confidence on the pollsters with a proven track record in measuring Latino sentiments.  Let me explain my reasoning.

While I did not have sufficient information to conduct a detailed post-mortem on the methodology utilized by YouGov in the Texas Hispanic Policy Foundation study of Texas Latinos, there were some concerns that were noted by Huff in his recent summary of Latino polls. [8] For example, the report did not identify the Hispanic sample size and its margin of error, while the survey results may have been distorted due to the use of outdated 2016 data to weight the survey results. Huff also expressed concern about the obvious discrepancy between the YouGov poll results and the results from the various polls including Latino voters. While it has merit, it is my opinion that the methodology used by YouGov lacks the historical track record in measuring Hispanic voter sentiments that has been employed for decades by Latino Decisions and Pew Research Center.

Perhaps the more compelling reason that raises skepticism is the inference by political pundits that Hispanic voter support for Trump is on the rise, a rather untenable position given his repeated attacks and hate-filled speeches directed at U.S. Hispanics.  Indeed, are we to believe that Hispanics suddenly feel compelled to reward Trump with their votes despite these attacks?  I don’t think so. Hispanic families have suffered significantly under Trump’s presidency and are not likely to forget these attacks as they decide to cast their vote this November.

Jason Villalba, President of the Texas Hispanic Policy Foundation, concluded from his organization’s polling results that “neither presidential candidate has yet to completely lock down the Texas Hispanic vote.” [9]  Based on the concerns raised about YouGov’s polling methodology, I would say that his conclusion is pre-mature and lacks confirmation by other trusted sources. Moreover, the report appears dismissive of the substantial Latino support shown for Biden in the various other polls conducted recently and makes little effort to reconcile the contradictory findings. The Biden campaign has accumulated millions in their campaign fund which may be utilized to solidify his standing among Hispanic voters, especially in the battleground states. Indeed, there is room for optimism.

Going forward, I believe that we would be wise to place our confidence on the polling results provided by the more trusted sources in Latino polling with a proven track record such as Pew Research Center or Latino Decisions. 

 End Notes