Alan Schwartz of Yale has written Regulating for Rationality, Forthcoming in the Stanford Law Review. Here's the abstract:
Traditional consumer protection law responds with various forms of disclosure to market imperfections that are the consequence of consumers being imperfectly informed or unsophisticated. This regulation assumes that consumers can rationally act on the information that it is disclosure’s goal to produce. Experimental results in psychology and behavorial economics question this rationality premise. The numerous reasoning defects consumers exhibit in the experiments would vitiate disclosure solutions if those defects also presented in markets. To assume that consumers behave as badly in markets as they do in the lab implies new regulatory responses. This Essay sets out the novel and difficult challenges that such “regulating for rationality” — intervening to cure or to overcome cognitive error – poses for regulators. Much of the novelty exists because the contracting choices of rational and irrational consumers often are observationally equivalent: both consumer types prefer the same contracts. Hence, the regulator seldom can infer from contract terms themselves that reasoning errors produced those terms. Rather, the regulator needs a theory of cognitive function that would permit him to predict when actual consumers would make the mistakes that laboratory subjects make: that is, to know which fraction of observed contracts are the product of bias rather than rational choice. The difficulties exist because the psychologists lack such a theory. Hence, cognitive based regulatory interventions often are poorly grounded. A particular concern is that consumers suffer from numerous biases, and not every consumer suffers from the same ones. Current theory cannot tell how these biases interact within the person and how markets aggregate differing biased consumer preferences. The Essay then makes three further claims. First, regulating for rationality should be more evidence based than regulating for traditional market imperfections: in the absence of a theory the regulator needs to see what actual people do. Second, when the facts are unobtainable or ambiguous regulators should assume that bias did not affect the consumer’s contracting choice because the assumption is autonomy preserving, administerable and coherent. Third, disclosure regulation can ameliorate some reasoning errors. Hence, abandoning disclosure strategies in favor of substantive regulation sometimes would be premature.