I’m not generally a big fan of blogging to complain about law reviews or the way that my work has been interpreted by others. I’m generally of the view that the risk of having my work misinterpreted within a reasonable range is my own to bear, and that if it happens, it’s probably due to my own failure to write clearly enough. I’m not a big fan of either of those things. With that kind of lead in, you can be pretty sure I’m going to do both in this post.
I recently opened my mail to see a reprint from Professor Alan White at Valparaiso University School of Law from his article, Behavior and Contract, published in the University of Minnesota Journal of Law and Inequality. Unfortunately, the SSRN link above is not the finally published version to which I will refer in this post. I read with some interest because Professor White’s article takes on a claim I make in this article that the behavioral law and economics literature has, at least as applied to the world of consumer contracts, overstated its case in terms of its predictive power relative to vanilla neoclassical economics. My article is essentially a literature review of the real world evidence involving consumer behavior in these markets, evidence both supporting and contradicting claims in the behavioral literature, and concludes that the rational choice models outperform their behavioral counterparts. The goal of the article was to draw attention to the existing empirical evidence, since proponents of both behavioral and traditional law and economics agree that predictive power of the models is of primary importance. I attempted to do so carefully and thoroughly.
As I thumbed through the article, I was struck by the following passage in n. 143 referring to my article:
Wright relies on industry-sponsored research to contend that behavioral theories are not sufficiently predictive of credit card consumers’ choices, because a majority of credit card consumers are observed to make the “rational” (i.e., wealth-maximizing) choice based on the given data. See id.
I was really surprised and disappointed to read this. My article is a literature survey. As such, I cite and “rely” on dozens of studies in the field. Professor White is referring to just one of these studies, a well known study in the credit card literature by Tom Brown & Lacey Plache, Paying with Plastic: Maybe Not so Crazy, 73 U. CHI. L. REV. 63 (2006) which uses a unique dataset (the VISA Payment System Panel Study) provided by VISA (Brown is affiliated with Visa, as is made obvious in the paper). No doubt, the Brown & Plache article is “industry sponsored” in some sense or another. It should be noted that I cite to, and discuss in detail, the findings of a number of studies. To my knowledge, the overwhelming majority of these studies do not involve any industry funding. Nor do I see anywhere else in Professor White’s article where he makes this special designation regarding funding sources for other articles he either critiques or cites.
But that is a bit beside the point, which is that the natural interpretation of the footnote is misleading at best. White implies that I either (1) relied exclusively on industry funded research to support my conclusion that credit card consumers appear to be acting rationally, or (2) tried to mislead readers of my article by emphasizing industry-sponsored research to the exclusion of “unbiased” studies. I note, for the record, that Professor White does not address the merits of Brown & Plache in the slightest nor even acknowledge the other studies coming to similar conclusions (and there have been several more since the publication of my paper back in 2007). The larger point is that I find the first sentence entirely misleading and suggesting something intellectually dishonest about my scholarship. As such, I wanted to clear the record here.
It is literally true that I cite to Brown & Plache and rely on their empirical findings in supporting my conclusion that the rational choice models maintain predictive superiority in credit card markets. But the message of the footnote is, in my view, pretty clearly that “one should dismiss Wright’s critique generally because he relies on industry-sponsored research and is biased”; it is NOT that “Wright cites to a bunch of studies and you should ignore one of them because it is co-authored by a VISA employee and uses VISA data.” As such, I view the reference as misleading and calling for the clarification made here with the request that readers interested in taking a look at the evidence please just read my paper.
Finally, this raises several other tangential but I think interesting points.
One is that I think that, as the title of the post suggests, law review editors should be catching things like this and not letting them slide. This is something that law review editors can and should be doing. It does not require special technical skill in statistics or econometrics, just a basic cite check.
A second is that this raises interesting issues about the probative value of industry-sponsored research in law and economics. Many have written about this. And I do think it can be rational to apply some discount to funded work in appropriate circumstances. But the larger point is that I view it as wholly insufficient to simply point to an article with serious empirical analysis (Brown & Plache in this instance) and dismiss it–as well as papers relying on it–simply because it is industry-sponsored and without taking on the methods, the data, results or anything of substance.
Third, there are other quibbles I have with Professor White’s article. For instance, the rest of n. 143 reads:
On the other hand, he concedes that based on the evidence, consumer behavior is neither 100% rational nor 100% irrational. See id. at 509-10. Wright contends that behavioral economists “assume consistent irrationality,” and thereby set up a straw man. See id. at n.31. Rather than asserting that consumer behavior is always and predictably non-utilitarian, as Wright implies of behavioralists, behavioral economics is better understood as saying that consumer behavior is not entirely predictable by rational choice theory. What behavioralism loses in predictive certainty, it gains in descriptive depth.
White apparently does not understand the behavioral literature he is hoping will inform legal debates involving consumer contracts. Of course behaviorists and the “new” paternalists assume consistent irrationality! If errors from cognitive biases were randomly distributed around some central tendency towards rationality then the average consumer would be acting quite rationally despite there being a distribution that included both rational and irrational individuals. We’d be back to a logical concession that the rational actor model predicted average consumer behavior incredibly well. To make the claim that the insights of behavioral economics can help us do “better,” it must be the case that (and by the way, Sunstein, Thaler, Jolls and just about every behavioralist that I’ve read proudly claims that behavioral economics can do exactly this …) behavioral law and economics can identify systematic, consistent and predictable deviations from rationality. The point of the behavioral literature is not simply to say that the rational choice model doesn’t entirely predict consumer behavior. Its to offer a better alternative. White’s denial on this point is not only puzzling, but undermines the intellectual basis for his reliance on behavioral economics in the first place.