Why Kaplow and Shavell's 'Double-Distortion Argument' Articles are Wrong
111 Pages Posted: 19 Dec 2003
Date Written: October 2003
According to an argument that economists denominate the Double-Distortion Argument, transaction-cost considerations aside, in an otherwise-Pareto-perfect world, it will be less-economically-inefficient to redistribute income between earned-income classes by adjusting taxes on earned income than by sacrificing economic efficiency by adjusting liability-rules and damage-rules to make legal outcomes depend on the earned income of the defendant relative to the earned income of the plaintiff or by making the price an individual has to pay for some government-provided good or service depend on the relationship between his earned income and average earned income in the society in question. The conclusion of this argument reflects the fact that, in an otherwise Pareto-perfect world, both the type of legal-rule adjustment and the type of government-pricing policy in question (1) will distort the profitability of labor-leisure choices to the same extent that it would be distorted by the "distributionally-equivalent" tax policy and (2) will in addition distort the profitability of another kind of choice (for example, in the case of an accident-law adjustment, potential-injurer and potential-victim avoidance-choices and, in the case of the government-pricing policy, potential-buyer purchasing choices). Kaplow and Shavell (KS) make two sets of claims that relate to this Double-Distortion Argument. The first is the following, highly-qualified pair of claims: if
(1) one is trying to redistribute resources from the richer to the poorer or between or among classes that are defined by their members' "richness" ("poorness") - put more negatively, one is trying to effectuate a redistribution that is not favored because it instantiates what KS refer to as "an entitlement to payment based on desert,"
(2) one has the option of doing so to the extent desired through tax policy,
(3) transaction-cost considerations can be ignored or do not favor the conclusion that the legal-rule adjustments needed to effectuate the desired redistribution will be more economically efficient than the tax policies needed to do so, and
(4) other Pareto imperfections do not make the relevant legal-rule adjustments economically efficient (for example, by offsetting the extra distortion the relevant legal-rule adjustment generates),
it will always be (A) not only economically efficient but also morally desirable (B) for any public decisionmaker to reject choices that would redistribute income from the richer to the poorer by making liability and damages decisions depend on the defendant's richness relative to the plaintiff's in ways that would generate an extra distortion even if it were desirable to redistribute income from the richer to the poorer. Kaplow and Shavell's second set of claims is that each of the four qualifications that their first set of claims contains can be ignored - i.e., is empirically insignificant.
This article explains why each element of both sets of Kaplow and Shavell's claims is wrong. First, it argues that Kaplow and Shavell's highly-qualified set of claims is wrong because it assumes incorrectly that the greater economic efficiency of a choice guarantees its moral superiority and because it ignores the difference between the authorization of "legislators" and "adjudicators." Then, it argues that even if Kaplow and Shavell's highly-qualified set of claims were correct, that fact would not justify their ultimate conclusion that no government decisionmaker should ever make a non-tax-policy decision that would sacrifice economic efficiency in an otherwise-Pareto-perfect world for distributional purposes because none of the qualifications Kaplow and Shavell's qualified claims contain can in fact be ignored. More specifically, the article demonstrates that only one of Kaplow and Shavell's arguments to the contrary can bear scrutiny and that the only correct (though inchoate) argument they make for ignoring a qualification (an argument about the role of judges in our type of democracy) undermines their assertion that another qualification is unimportant.
Suggested Citation: Suggested Citation