Mainstream Media Use Flawed ‘Hispanic’ Polling to Attack Trump

donald trump
Matthew Busch/Getty Images

The Washington Post and ABC News want you to know Hispanics hate Donald Trump. I mean really hate him – to the tune of 82% viewing him unfavorably.

That was the narrative both outlets harvested, packaged, and propagated last Wednesday when highlighting a poll the two co-sponsored. These outlets helped create a social weapon, of sorts, that was wielded by those looking to bruise Trump for a few news cycles. The funny thing is that just a few years ago, ABC News, the company that co-sponsored the poll, published a report critical of the same Hispanic polling methods it is now using against Mr. Trump. Welcome to the privileges of the mainstream media.

The weapon I am referring to was first given birth through an article published by the Washington Post citing the results of a poll it co-sponsored with ABC News. The author of the piece highlighted Jeb Bush’s advantages over Donald Trump with the Hispanic community, subtly noting that Trump earned an 82% unfavorable rating from Hispanics to make his point. I guess you can say the author selectively edited the data.

CNN saw the number and pulled it to the forefront – repackaging and sharpening the weapon. “Poll: 82% of Hispanics view Donald Trump unfavorably” screamed its headline. The article, co-authored by no less than CNN’s Director of Polling and Election Analytics, who’s likely no slouch with statistics, claimed the poll showed “Hispanic voters” preferred Jeb Bush to Donald Trump. CNN would swing this weapon in its panels, their pundits would shoot it out on Twitter, and ABC News would likewise brandish it in a story on Good Morning America. Megyn Kelly picked it up as well, along with other news sites far and wide.

The problem is, the poll cited by all of these trusted outlets wasn’t actually a “Hispanic” poll. It wasn’t even really a “voter” poll as CNN claimed. And to make things worse, the way it was being used didn’t meet one of its sponsors’ own smell test.

The poll methodology was published by both the Washington Post and ABC News, each providing varying degrees of transparency, but both clearly state the poll was “a random national sample of 1,005 adults” – in other words, not solely Hispanics. It also included respondents that were “not registered” to vote. As I’ll point out, these are both important characteristics of the poll.

I contacted both the Washington Post and ABC News asking for the sample size of Hispanics used and they have yet to respond. But using the published methodology that appears to use Census percentages to simulate the general population, their poll likely only included roughly 171 Hispanics residing in the U.S. — of which all were not necessarily citizens. Putting that rather hefty latter point aside for the moment, the sample of Hispanics was just too small to extract reliable Hispanic data – or at least that was a point discussed and emphasized at length by none other than ABC News just a few years ago.

According to a report entitled “The Problem with Polling Latinos” published by ABC News in 2012, “Sampling Latinos as a small fraction of a national survey can also often [be] problematic, pollsters say.” The report continued (emphasis added):

Often times, national polls also fail to survey enough Latinos or break out the demographics by important indicators such as state, age, foreign born vs. native born, and country of origin. A recent poll from Quinnipiac took a sample of just 143 Latinos as part of a greater national poll.

The ABC News report quoted Latino Decision, a Hispanic polling outfit recently hired by Hillary Clinton, as stating (emphasis added), “National polls are not designed to get an accurate geographic representation of Latinos because they draw a nationally proportionate sample of all Americans, and pick up Latino respondents wherever they surface… the Latino sample is rarely representative of the overall Latino population.”

In other words, ABC News was now using its national poll in a way it once reported as being “problematic.” That’s no bueno.

This is often how the partisan mainstream media works. They harvest information looking for lethal material, then package this information into a social weapon that can be used and modified by other like-minded organizations. And when it loses its potency, it’s discarded. Old standards fall by the wayside if the weapon can be forged sharp enough or used for a quick hit.

In order to get a more credible Hispanic view on Trump, the pollsters should have performed what is called an “oversampling” of Hispanics – meaning they should have supplemented the general population poll by questioning considerably more Hispanics to accurately represent the community’s view on Trump. And that’s not just my opinion — it’s that of Pew Reseach. Even ABC News reported that very fact as it relates to Hispanic polling. That doesn’t appear to have happened in their poll.

Polling Hispanic voters is a well documented challenge, as has also been noted by the New York Times and Huffington Post. But what these outlets fail to emphasize also happens to be the biggest obstacle in acquiring accurate Hispanic opinions – and that is, it’s difficult to know if the randomly selected Hispanic person you are polling is actually a U.S. citizen.

This is no small point.

Depending on which estimate of the illegal Hispanic immigrant population you believe, randomly polling Hispanics could include anywhere from 12% – 35% illegal immigrants. So not only is the Washington Post/ABC News poll likely capturing a very small sample of Hispanics, a significant portion may not even be U.S. citizens. And my guesstimate is that every single illegal Hispanic immigrant would answer “no me gusta” to the question of whether they like Donald Trump. Add to these factors the fact that the pollsters asked to speak with the youngest adult in the house when polling, which typically skews Democrat, it’s easy to see that characterizing Trump’s unfavorables at 82% with Hispanics is highly unreliable.

Which leads me to one question. Given that ABC News didn’t meet the criteria it previously reported was needed to produce a reliable Hispanic poll, will ABC News and its weapon-wielding comrades correct the record? I won’t hold my breath.

Patrick Courrielche has contributed to Breitbart since 2009. He can be followed on Twitter @courrielche.


Please let us know if you're having issues with commenting.