The Fantasyland of Conflict of Interest

My regular readers are marvelous people. Though I rarely respond to comments, your feedback is a privilege. Whenever my writings overflow my Bubble and reach a mass audience, however, the feedback tends to be hysterical and childish. The modal comment ignores all my arguments in favor of a cheap ad hominem: “You’re funded by the Koch brothers!”

There’s a double insinuation.

The first is obvious: No one could sincerely hold my beliefs.

The second is subtler: My willingness to say things is highly price-elastic.

I can’t help but find the first insinuation insulting. On reflection, however, the second insinuation is barely about me. After all, Conflict of Interest (COI) rules permeate academia, and their core assumption is precisely that humans in general are easily bought.

You could object, “COI rules don’t deny the integrity of the average person. They exist to cover our bases. Even if only a small minority of people are corruptible, we hate corruption so much that we go to great lengths to eliminate it.”

A nice argument, but it contradicts some obvious facts. Namely: There are two mighty non-financial factors that corrupt human judgment, yet COI almost totally ignores them.

1. Ideology is the most obvious. One of the dirty little secrets of academia is that if you know a social scientist’s political learnings, you can predict what his research will “find” with 90%+ accuracy. Sure, the details of the research can be slightly contrarian. But no matter what facts he discovers, the typical leftist academic will never write, “Affirmative action is stupid and we should just embrace color-blindness.” Believing is seeing; if your fundamental philosophy tells you that white racism is the worst thing in the world, even a trace level of white racism warrants a full-blown crusade.

2. Friendship is the other undeniable factor of corruption. Politics aside, everyone is invested in their own past research findings. If your research tells your friends what they want to hear, you’ll stay friends. Otherwise, your publications put your friendships at risk. The words, “If you publish the wrong conclusions, we’re not friends anymore” need never be spoken. In fact, when someone goes out of his way to reassure you that, “Whatever your research finds, we’ll still be friends,” you should still be worried. Maybe even more worried! And once you’re worried, the natural human tendency is to shade the truth — or twist it beyond recognition.

Under current COI rules, researchers are, at minimum, expected to disclose relevant funding sources. But it would make at least as much sense to require researchers to disclose their ideology, plus a list of their closest colleagues.

In fact, mandatory disclosure of ideology and friendships would make a lot more sense than the status quo. Why? At least outside of STEM, ideology is plainly a far stronger predictor of research conclusions. And even within STEM, friendship probably matters a lot more for your career success than funding per se. If successful colleagues personally like you, you really don’t need your own funding, because the people in charge want you on their research teams.

Why do COI rules focus on one relatively minor factor of corruption? Because we’ve “discovered” the “leading cause” of corruption in a totally corrupt way. Instead of calmly measuring the factors that bias human judgment, rule-makers used their simple-minded ideology to “learn” that money is the root of all evil. Blaming social problems on “greed” just sounds a lot better than blaming social problems on “fanaticism” or “cronyism“ — even when the relevant social problem is “the corruption of expert judgment.” COI rules comes from a Fantasyland where evil fatcats habitually bribe your intellectual opponents to feign disagreement with the undeniable truths for which you stand.

To be fair, plenty of free-market economists casually believe that the typical statist social scientist has been corrupted by government funding. This, too, is almost all wrong. The truth is that if all government funding of social science ended tomorrow, researchers’ ideology would barely change. Bias is endemic, but the bias comes from within.