Legal Transitions, Rational Expectations, and Legal Progress

Abstract

This paper was written for a symposium on legal transitions. The central question in the transitions literature is whether or to what extent the government should provide compensation for (or seek in some way to ameliorate) the losses of wealth occasioned by unexpected changes in the law or social policy. This Article argues that the prevailing normative framework for evaluating legal transitions - the consequentialist or economic framework - relies critically on two assumptions: the assumption that private parties (whose incentives are sought to be affected by the choice of transition norm) will behave with rational expectations and the assumption that the law will tend to change in a desirable direction, that there will be "legal progress." Perhaps the most important, and certainly the most well-known, insight of the consequentialist framework has been that, if both of these assumptions hold, a powerful case can be made in favor of a norm of applying legal change retroactively. To the extent, however, that these assumptions do not apply, the case for retroactivity becomes considerably weaker, at least on incentive- or anticipation-grounds.

One conclusion of the Article is that those assumptions are quite likely to apply - and hence a retroactivity norm will likely make good sense - with respect to two fairly narrow types of legal change: 1) expansions in products liability law (or expansions in product safety or environmental regulation generally) that are based on the incorporation of the latest (and adequately confirmed) scientific discoveries; and 2) moves, by Congress, courts, or the IRS to eliminate abusive tax shelter transactions. In both of these admittedly unrelated cases, there is a strong case to be made that the relevant parties - manufacturers in one and sophisticated taxpayers in the other - can rationally predict the future course of law and its application to them. And there is a strong case that these types of legal change will tend to be desirable. One way of putting my conclusion is that, if retroactivity is ever going to make sense on incentive grounds, it will make sense in these cases.