How the Left’s Favorite Pollster Is Leading Democrats Into a Trap

Look, for example, at a loaded question Data for Progress asked about the For the People Act.

Rep. Alexandria Ocasio-Cortez (D-New York) says she has figured out why Sen. Joe Manchin (D-West Virginia) is voting against the For the People Act. That’s the sweeping overhaul of election laws that Manchin deep-sixed last week with his announcement that he will support neither the bill nor dismantling the 60-vote threshold that would allow it to pass with Democratic votes alone.

“[The bill] stands up against lobbyists and dark money. I would reckon to think that this is probably just as much a part of Joe Manchin’s calculus as anything else. Because when it comes to this bipartisan argument—I don't buy it,” Ocasio-Cortez told MSNBC. The congresswoman, keeper of many positions unpopular in her own party, cannot summon the moral imagination necessary to understand that another person could sincerely dissent from the majority. Nor is she afraid to make such claims when her own campaign finance conduct could more easily be called into question.

This kind of motivated reasoning is nothing new. And certainly it is not the domain of one party or sect. I assume Ocasio-Cortez would have raised as much money as she could to unseat incumbent Democrats, even if the funds didn’t flow through her former chief of staff’s firm. Similarly, I assume Manchin opposes the bill for his stated reasons—chiefly that it would be acidly divisive at a time of paranoiac partisanship—not because it would limit his access to “dark money.” If you intend to have a real discussion, assume sincerity in your adversary unless proven otherwise.

The human mind is like a miniature Mozart, pouring out masterpiece symphonies of rationalization, one after another. Our powers of reason usually serve to defend our own turf, not choose on which turf to stand. Nothing has changed on that front since David Hume taught us 283 years ago: “Reason is, and ought only to be the slave of the passions.”

And right there, at the sweet spot where confirmation bias meets every politician’s desire to be popular and remain in power, is junk polling. If you look hard enough, you can always find a poll that will support your position. You can cherry-pick favorable numbers out of an otherwise ugly poll from a reputable pollster, or, more commonly, you can just lean on the dubious findings of a partisan or otherwise slipshod firm. If you can say that your preferred position is not only right, but backed by public sentiment, how could your opponents legitimately disagree?

That’s the trap into which Ocasio-Cortez and her fellow Democrats have fallen when it comes to the For the People Act. And leading them into the snare is the new favorite pollster on the left, Data for Progress.

The firm’s poll on the legislation has become a staple of Democratic talking points. Here’s the Los Angeles Times from last week: “[Mariah] Carey became the latest on that list to tweet the hashtag #CallOutYourSenators as part of a viral campaign to pressure certain elected representatives into voting for the bill, which has strong support, according to polls (69 percent in favor, says a May Data for Progress poll).” Who else but a crook (or maybe a racist) could oppose legislation supported by more than two-thirds of Americans—including more than half of Republicans? Unless, of course, that 69 percent is a mirage.

A weekend piece in the New York Times says Data for Progress is popular among Democrats because its work is “targeted, cheap and fairly accurate,” citing the firm’s “B” grade in FiveThirtyEight’s pollster rankings, the same as Rasmussen Reports and CNN. But even the faint praise of “fairly accurate” can only be measured in relation to elections. Yes, Data for Progress is about as good at accurately polling elections as other cheapo online or robo-call outfits.  But that doesn’t apply to issue polling. We know when an election is over how well the pollsters did. But how do you know if a poll has accurately gauged public sentiment on an issue?

In public policy, very seldom does an issue break through to sufficient public awareness to be subject to reliable public surveys. But even on those issues—e.g. abortion, military interventions, and immigration—a great deal depends on how pollsters phrase their questions. Unlike activists, most Americans have complex and often seemingly contradictory opinions. Doing issue research right is hard, expensive, and seldom satisfying to partisans. But doing it wrong is easy. To wit, here’s the whole question Data for Progress asked about the voting law: The For The People Act has been introduced in Congress. Supporters of the bill say it would limit the influence of big money in politics by empowering small donors; make voting easier and more secure; end gerrymandering; and give the public more information about who is lobbying our government. Opponents say it would be an overreach by the federal government and that states should control their own elections. Do you support or oppose the For The People Act?

Sheesh. Leaving aside the limited value of online-only polls, that question would be a stinker in any methodology. You ask respondents, many of whom have never heard of the bill, whether they support something that empowers “small donors,” secures elections, ends gerrymandering, and cracks down on “big money” and lobbyists. Then you tell them the only concern is states’ rights. Presto, 69 percent approval. This dishonest, Trumpian approach to public opinion is bad for the whole country, but particularly dangerous for Democrats as they contemplate ways to jam through legislation that would not only be incendiary but—in short order—very unpopular. The bill is overbroad, radically transformative, and would face serious constitutional challenges. They’d lose all of the Republicans and half of the independents for starters, and decline from there.

Republicans made a similar mistake in their understanding of polls on Obamacare. They saw legitimate numbers that showed the new health-insurance entitlement was unpopular, but failed to ask why. They would have found that while about a third or so of respondents didn’t like it for conservative reasons, another substantial group consistently felt the program did not go far enough. That’s how the GOP ended up in a humiliating defeat after 10 years opposing the law. They only saw what they wanted to see. Now, Data For Progress is making it even easier for Democrats to do the same.