There’s a famous quote attributed to President Harry Truman:
Give me a one-handed economist! All my economists say, ‘on the one hand . . . on the other.’
The natural ambivalence of economists recently surfaced around conflicts of interest (COI) in the economics literature, as studies supporting high CEO pay were found to be published in high-impact journals by authors with a lot of experience consulting for big companies, yet there was plausible skepticism that such circumstantial evidence showed cause and effect, partly because the evidence is sketchy and partly because the logic supports both possibilities (i.e., the best economists generate the best evidence and are more widely sought after as consultants).
On the one hand, on the other.
However, with the backdrop of the movie “Inside Job,” the 2008 financial meltdown, and activists in the community, events recently came to a head, and a group of influential authorities in the field knew it was high time for a change.
The economists’ name for the COI problem being addressed is “capture,” which is sexier than “bias” and more nefarious than “conflict.” It suggests an obeisance and possession relationship that bias alone doesn’t. “Bias” suggests malleability; “conflict” reflects an individual’s relationship to an entity or topic; “capture” suggests the powerless held until the powerful relent. It’s not an entirely fair word, but it’s compelling.
Without rules in place, the concern is that economists may have recently been taking advantage of a relatively lax regulatory environment to cash in on lucrative consulting and board contracts, something exposed in the recent documentary “Inside Job,” which won the 2010 Academy Award for Best Documentary Feature, and which Robert Redford recently praised (along with other documentaries) as an example of investigative journalism done as documentary.
Perhaps like their cousins in the audit community (who grew too close to their clients, yielding the Enron and Worldcom scandals, among others), economists grew too close to their clients, essentially blinding themselves to the calamities that befell us in the late 2000s (rampant speculation, credit-default swaps, and so forth).
. . . adopt a code of ethics that requires disclosure of potential conflicts of interest that can arise between economists’ roles as economic experts and as paid consultants, principals or agents for private firms.
Apparently, economics has been one of the few fields without ethical guidelines defining appropriate behavior when it comes to conflicts of interest or bias in publications. In fact, the field seems to be in the early stages of portraying the problem.
A recent attempt may be illustrative. Luigi Zingales, a professor at the University of Chicago’s Booth School of Business, recently presented on the topic at the same meeting of the American Economic Association at which the new COI rules were articulated. In his presentation (as reported on Bloomberg View):
Zingales looked at the 150 most-downloaded papers that had been done on executive pay. He found that papers supporting high pay for top executives were 55 percent more likely to be published in prestigious economic journals. They were also much more likely to be cited in other papers.
I would love to see the full data here, because while these top-level data are mildly suggestive, they’re far from conclusive. In fact, there are fewer data here than meets the eye. For instance, the n of papers is unknown, but it’s less than the 150 papers on the topic — it’s the subset that concluded that high pay for executives was a good idea. That unknown number of papers is then 55% more likely to be published in prestigious economic journals. What does that mean? There are two major potential meanings — first, that everyone’s in on the scam, or that such papers are more rigorous and likely to be true. We aren’t sure which. Then, there’s the fact offered that papers published in the most prestigious journals are “much more likely” to be cited.
We all know that citations suffer from a Matthew effect — the big and powerful are cited more because they’re prominent, and they’re prominent because they’re cited more. A citation association like that portrayed above is entirely predictable. In fact, the surprise would be if those facts weren’t true.
As for the other data, we don’t know if the n of papers is 149 or 14, so it’s impossible to know. We have no p-value or other statistical measure. And the logic isn’t at all clear about cause and effect.
I emailed Zingales requesting the presentation or paper, but it was late in the day, and I didn’t receive a response. If I do, I’ll write a follow-up, because I don’t know what to make of what evidence he has, and a lot of searching didn’t reveal what I was after. Everyone’s pointing back to his little data pastiche in Bloomberg View.
On the one hand, the AEA has changed its COI rules. That’s for the better, and overdue. The rules are very similar to those used in medicine, engineering, and other fields, but only apply initially to the seven journals published by the AEA. Other economics journals are expected to adopt them as well.
On the other hand, these are rules economists should have embraced long ago. We may be living through some of the consequences of conflicts unexposed and unexplained.