The Taste for Fairness

In a recent article (and new book), Louis Kaplow and Steven Shavell argue that analysts of legal policy should rely exclusively on welfare economics in making decisions, and give no independent weight to fairness, justice, or several other such noneconomic considerations they treat as roughly synonymous. The authors say that fairness is worth worrying about to the extent that people have a taste for it, since in that case a policy that is fairer makes them better off by their own lights. But this exception does not make arguments about fairness as such relevant to legal policy; rather, it calls for purely empirical demonstrations, perhaps through surveys of public opinion, that a policy will upset people because it will offend their taste for fairness. This essay raises a series of objections to Kaplow and Shavell's arguments. First, their acknowledgement that policies should take fairness into account if people have a taste for it has more implications than they recognize. It implies as well that if people have a taste for officials who take arguments about fairness seriously, then they should have officials who pay attention to those arguments (and not just to survey data) in making policy. In principle, the authors are open to the counting of any preferences, presumably including these; in practice, the extent of anyone's taste for fairness, presumably including their taste for officials who worry about fairness, is treated as an empirical question on which convincing data evidently will be hard to obtain. This puts appeals to such tastes at an unwarranted practical disadvantage, and overlooks the role of politics as a register of tastes for various notions of fairness. Second, economics does not provide answers in situations where the contested issue is how much certain goods should be valued - what should count as a cost and as a benefit, and in what proportions. This becomes important when, as is common, a question of legal policy arises in an institutional setting where market measures of value are unavailable and direct empirical testing of people's tastes for fairness in various senses is infeasible and perhaps undesired by (distasteful to) the people themselves. Third, even if our ideas about the fairness of a decision express beliefs or intuitions about the costs and benefits at stake in it, this does not necessarily make cost-benefit analysis a superior procedure for decision than an appeal to fairness. It may be that beliefs about fairness capture costs and benefits in a way different from, and preferable to, the way they are likely to be captured in an economic analysis.