POLITICS

All The Problems With This Anonymous Poll About Alexandria Ocasio-Cortez

Be smart: Don't bother with unsourced data.
Rep. Alexandria Ocasio-Cortez (D-N.Y.) is the subject of a questionable poll published by Axios.
Rep. Alexandria Ocasio-Cortez (D-N.Y.) is the subject of a questionable poll published by Axios.

The 2020 general election is still more than a year away, but it’s never too early for some truly unhelpful polling coverage.

As part of the latest salvo of Democratic infighting, Axios published Sunday what they called an “exclusive poll” from May of white likely voters without college degrees. The results, Axios’ Mike Allen wrote, have been circulated by top Democrats to show “that one of the House’s most progressive members — Rep. Alexandria Ocasio-Cortez — has become a definitional face for the party with a crucial group of swing voters.”

Just don’t ask where it came from. 

“The group that took the poll shared the results with Axios on the condition that it not be named because the group has to work with all parts of the party,” Allen wrote.

Why it matters: With public trust in polling frayed, and pollsters adopting an increasingly diverse array of methods to keep up with the times, transparency has become more crucial than ever both for the firms conducting surveys and the outlets reporting on them ― making the idea of reporting on an “anonymous poll” an unfortunate outlier.

Here are some of the Associated Press’s guidance for survey reporting: “Reputable poll sponsors and public opinion researchers will disclose the methodology used to conduct the survey, including the questions asked and the results to each, so that their survey may be subject to independent examination and analysis by others…. Polls paid for by candidates or interest groups may be designed to produce results that are beneficial to that candidate or group, and they may be released selectively as a campaign tactic or publicity ploy.”

Here’s an excerpt from CNN’s new standards for writing about polls: “Evaluating the validity of a poll is only possible if pollsters are transparent about what they are doing. Transparency has become a critical part of how those in the survey research world assess the validity of other research ... A poll’s existence alone does not make it news, and journalists shouldn’t be any more credulous about numbers than they are about words.”

In this case, readers have no way of knowing who commissioned the poll, who conducted it, how they identified the voters they surveyed, what methodology they used to interview them or what exactly respondents were asked. That makes it basically impossible to evaluate the survey in any meaningful way. 

Go deeper: Beyond the questionable sourcing, the polling data provided doesn’t make a clear case that white voters who aren’t college graduates see Ocasio-Cortez as the face of the party. The numbers included ― name recognition and favorability for Ocasio-Cortez and Rep. Ilhan Omar (D-Minn.), as well as opinions on capitalism and socialism ― suggest that both congresswomen have unusually high national profiles for freshman representatives and that neither is popular among the subgroup surveyed.

But those numbers don’t establish whether they’re seen as defining all Democrats. That’s an eminently pollable question: A 2017 CNN survey, for instance, asked, “Thinking about Democratic leaders today―which one person best reflects the core values of the Democratic Party?”

It also doesn’t establish whether they’re actually less popular than Democrats in general. The bloc polled ― white voters who aren’t college graduates― is Republican-leaning in general. And in one publicly available poll, a February Economist/YouGov survey, Ocasio-Cortez’s net favorability rating (that is, the share who rated her favorably minus the share who rated her unfavorably) among white voters without degrees was roughly in line with ratings for House Speaker Nancy Pelosi and for Democrats as a whole.

That’s obviously one datapoint, and a dated one at that. But it suggests that these voters tend to be consistent either in liking the Democratic Party and its politicians or in disliking them.

There are a few additional hazy details in the Axios story. The headline implies that its findings reflect “swing states,” but the article doesn’t make it clear whether the poll was conducted nationally or only in battleground states. And although the article provides exact percentages on how few of the voters polled viewed Ocasio-Cortez and Omar favorably, it doesn’t include numbers on how many viewed them outright unfavorably, rather than having no opinion.

The bottom line: Basing a story on unsourced, thinly contextualized polling data isn’t savvy; it’s deeply credulous.

CONVERSATIONS