On bias and conflicts of interest in antitrust academia

Recently, I had the privilege of hosting a one-hour conversation with one of the foremost US academic economists whose research (and name) is today associated with several “theories” in our field and beyond. During our discussion, we talked about several prominent (and not so prominent) antitrust cases. On almost each and every case, our guest provided a disclaimer that he had been retained as a consultant to work on those matters. We proceeded seamlessly (and almost unconsciously) to discuss ideas. I cannot recall any segment in the interview where we felt his thoughts were biased, prejudiced or stained by his consultancy work.

I learnt a few things from this conversation: First, while the disclosure of interests gave us a sense of the direction of the arguments (in a field where private interests and the public interest converge and collide), it had no bearing on their formal, methodological and empirical merit. After all, regardless of where the money to fund research came from, what mattered were the ideas. Second, disclosure ensured accountability: our guest put himself in a position where we could cross examine him on any deviations, extrapolations and falsifications of “prior art,” in particular in the light of his own previous and parallel academic and consultancy work. And because reputation matters, our guest’s disclosure signalled that the views he expressed whilst wearing his consultant “hat” were, and remained, his own personal views. Third, our guest talked perceptively of the industries we were discussing, in a field where practitioners who work in the trenches often deride from the outset academic work as being out of touch with empirical reality. This may be one of the upsides of doing consultancy work. Fourth, our guest also disclosed the consulting work he had done on behalf of the Government. Again, this was a useful reminder that bias (if any) is a two-way street, in particular when antitrust agencies are run and directed by professional (and elected) politicians.

Overall, disclosure assisted our discussion, and provided empirical evidence of the general and specific positions advocated by Luigi Zingales in a paper titled “Preventing Economists’ capture”.

Strikingly, disclosure is not the norm in European antitrust academia, much to the contrary (the press this week got particulary vocal on this).[1] This gives rise to three pernicious effects. First, when disclosure is not the social norm, any academic work — including work that has not been “sponsored” by any private interests but that coincidentally aligns with them — can be stigmatized by an opponent as biased, and this with no evidence other than the opponent’s own beliefs.[2] Second, because academic works containing a voluntary disclosure are the exception, they lend themselves to being systematically stigmatized as biased (including by competing intellectuals who do not disclose their own potential conflicts of interest).[3] Third, when sponsoring is declined, this is not seen and thus not accounted for as a sign of good ethics. I have personally experienced all three effects. For instance, when I wrote some blog posts discussing critically the Google Shopping decision and supporting the Apple tax case, some accused me on Facebook of being part of a “gang” of academics sponsored by Google. The problem is that I have not received a single $, directly or indirectly, from Google or any related organisation. I just had nothing to disclose (although it is true that I have received service in kind through Google Scholar)… On another instance, I wrote a rebuttal of a piece on self preferencing authored by Bo Vesterdorf and sponsored by Google. I disclosed that my article had been written under an unrestricted sponsorship from iComp, a trade association financed by Microsoft that was the main initiator of the early complaints against Google, and I was accused this time of siding with the Microsoft “gang”. At this stage, one can see how absurd such allegations are: how could I possibly be (financially) biased to favour Google, and at the same time write a piece that generates (legal) arguments against it… In other circles, some of my academic articles on patent law have been disparaged as being pro-Qualcomm simply because that company gave a (disclosed) grant to the Liege Competition and Innovation Institute (“LCII”) restricted to the funding of PhD and PostDoc research and events. Lastly, and perhaps even more absurdly, I have been told by some law clerks that some colleagues of the Court of Justice of the EU (“CJEU”) believed that I had done work for Intel, because when I attended the hearing at the CJEU out of personal curiosity, I sat on the side of the room allotted to the defendant’s lawyers…

I guess we antitrust academics should work together to design a code of conduct on conflicts of interests and ethical guidelines for research, which would include systematic disclosure requirements. Lack of disclosure undermines intellectual debate. Ad hominem rhetoric and unverifiable accusations take precedence over proper argumentation. People play the man, not the ball.

When funding, sponsorship and grants are disclosed from the outset, intellectual opponents can move quickly and confidently to discuss the merits of ideas, positions and theories. The identity of those paying the opportunity cost of academics is no longer an issue, because any inconsistency, falsification or manipulation in the research will soon or later be discovered.

As with the Schelling hockey helmets example, this is a collective action problem that academics should try to solve through formal coordination (hopefully, the antitrust agencies will not object to that). As for the writer of this piece, I have decided to soon set up a page on my website which will disclose potential conflicts of interests, drawing inspiration from Luigi Zingales. I know I face the risk that my research will be criticized for this, but unless we do not take such steps individually, we will not make our profession collectively better.

[1] Even though we cannot rule out that academic consulting is as widespread an economic practice as in the US.

[2] Free speech and the absence of defamation rules in the field facilitate the proliferation of such attacks.

[3] See, for instance, https://www.wsj.com/articles/paying-professors-inside-googles-academic-influence-campaign-1499785286