Belief in conspiracy theories: What do the data say?

In a recent Monkey Cage post on belief in conspiracy theories among liberals and conservatives, Alfred Moore, Joseph Parent and Joseph Uscinski criticized liberal pundit Paul Krugman for writing, “Unlike the crazy conspiracy theories of the left — which do exist, but are supported only by a tiny fringe — the crazy conspiracy theories of the right are supported by important people: powerful politicians, television personalities with large audiences.” They responded, “Krugman is mostly wrong that nuttiness is found mainly among conservatives.”

I noticed this and criticized Moore, Parent, and Uscinski for what I saw as oversimplification:  Krugman was making a statement about the attitudes of American elites of the left and right–arguing that conspiracy theories are supported by important leaders on the right but only by a tiny fringe on the left–whereas Moore et al. were giving data on mass opinion.  The data on mass opinion are valuable–and I’m a big fan of the research of Dan Kahan that give us these data–but they don’t address Krugman’s point about leaders vs. fringe.

Yesterday, Parent and Uscinski responded to my criticism as follows, writing that “they [Krugman, Jonathan Bernstein, and me] counter that Republican elites are more approving of conspiracy theories than Democratic elites.”

Before going on let me just clarify that I myself am not, and was not, saying anything about elites.  What I’m saying is that the survey data cited by Moore, Parent, and Uscinski do not directly address Krugman’s points.  Krugman may be correct or he may be incorrect in his claims;  my point is that a statement about what percent of liberals and conservatives have certain attitudes is not the same as a statement about who are the people with these attitudes.  (I also took issue with Moore et al.’s characterization of what is a conspiracy theory, but that’s a separate issue.)

Anyway, Parent and Uscinski responded as follows to Bernstein, Krugman, and me:

Our critics are sharp observers and they are probably on to something.We make the case in our book that out-of-power parties are especially inclined to conspiracy theorizing. So we agree that Republican elites are more prone to conspiracy theories right nowbecause they have more to fear. With a Democrat occupying the most powerful office on the planet, Republican elites are an alarm system sounding warnings about potential plots against conservative values. But that is a temporary, situational argument, and one that none of our critics make.

Their claim is that there is something conspiratorial about conservative elites that is unique and lasting. Circumstances change, but conservatives’ stay the same. Democrats are different.

When they talk about “our critics” here, they can’t be talking about me, as I never wrote anything about unique or lasting.  So I clicked through to see what Krugman and Bernstein had written, and I didn’t see anything about unique and lasting or anything like this.  In his recent post Krugman specifically refers to “these days,” and in a post from 2013 he was even more explicit:

Back in the 1980s, after all, health experts at Heritage [the conservative policy advocacy organization] made a good-faith effort to devise a plan for universal health coverage — and what they came up with was the system now known as Obamacare.

But that was then. Modern conservatism has become a sort of cult, very much given to conspiracy theorizing when confronted with inconvenient facts.

So, yes, Krugman calls conservatism a “cult” but he’s not saying that “something conspiratorial about conservative elites that is unique and lasting.”  Not at all.  It’s the opposite–Krugman is explicitly saying there have been changes in the past few decades.  Krugman’s also written about the “Clinton derangement syndrome” so I think he’d be sympathetic to Parent and Uscinski’s point that out-of-power parties are especially inclined to conspiracy theorizing.

OK, so neither I nor Krugman claimed that there is something conspiratorial about conservative elites that is unique and lasting.  What about Jonathan Bernstein?  I followed the link and he writes, “We also need to have some practical understanding about what’s gone wrong with the Republican Party.”  This doesn’t sound like a claim about a unique and lasting phenomenon.  “What’s gone wrong” implies a change or trend.

Identifying the disagreement

The big difference here is that Krugman explicitly says there have been changes since the 1980s, i.e., that many influential Republican leaders became associated with conspiracy theories during the Clinton years and stayed that way, with no corresponding conspiratorial shift among influential Democrats.  Bernstein doesn’t specify dates but my impression is that he holds the same view.

In contrast, Parent and Uscinski identify conspiracy theorizing as coming from “out-of-power elites” so they might agree that Republican leaders believed in conspiracies during the Clinton and Obama presidencies but that Democratic leaders similarly believed in conspiracies during the Bush years.

Nobody is claiming that there is something conspiratorial about conservative elites that is unique and lasting, not if you define “lasting” as lasting back before 1992.

What do the data say?

I have not looked at any data on the attitudes of Democratic and Republican leaders.  But Parent and Uscinski bring some empirical evidence to the table.  Let’s see what they have to offer:

We tried to remedy this [lack of evidence] in our research and examined over 100,000 letters to the editor of the New York Times and Chicago Tribune as a proxy for public opinion over a century, separating elite from non-elite letters. Our results found little difference between elites and non-elites, or between Republicans and Democrats, or between Republican elites and Democratic elites. The best available data doesn’t support Krugman.

It’s great to see people doing empirical research and ya gotta start somewhere, but . . . No.  Based on the description above, their data are essentially irrelevant to their debate with Bernstein and Krugman.  As I noted above, the debate is all about a purported change during the Clinton era  and continuing since then.  It’s fine to have a century’s worth of data but what’s going on here is not a disagreement about what was happening in 1920 or 1940 or even 1960 but much more recently.

I can’t be sure about this as I haven’t see the book that Parent and Uscinski link to, but through a web-search I did find a couple of articles by them using this dataset (that’s where I got the above pie chart) and I didn’t see attempt to look for changes in recent decades.

Beyond this, I don’t know that letters to the editor are telling us much about elite attitudes at all.

In short, I respect Parent and Uscinski’s research project but I don’t think it has anything much to say about Bernstein and Krugman’s claims, nor for that matter do I think that Bernstein and Krugman are saying what Parent and Uscinski are saying they’re saying.

P.S.  All this might seem like a bit of inside baseball, a debate among some pundits about what some other people believe.  But it’s important.  To the extent that leaders of either party believe in false conspiracy theories, this is a problem.  For that matter it’s a problem when the public has these false beliefs.  So I agree with Dan Kahan that the topic is worth studying, I agree with Krugman that it can have policy implications, and I agree with Parent and Uscinski that we should look at any general claims about liberals and conservatives with a critical eye.  And I agree with Bernstein that if data are not directly available to evaluate claims such as Krugman’s, we shouldn’t be so quick to be dismissive.

Andrew Gelman is a professor of statistics and political science at Columbia University. His books include Bayesian Data Analysis; Teaching Statistics: A Bag of Tricks; and Red State, Blue State, Rich State, Poor State: Why Americans Vote the Way They Do.
Comments
Show Comments
Most Read Politics
Next Story
David Damore · September 3