Judicial independence and social welfare.

AuthorGilbert, Michael D.
PositionII. Optimal Judicial Independence D. Step Four: The Payoff through Conclusion with appendix and footnotes, p. 601-625
  1. Step Four: The Payoff

    The judge's decision, whether legalistic or nonlegalistic, affects social welfare, meaning that it affects aggregate utility. I refer to this effect as the "social payoff' of the decision. This Section describes possible payoffs.

    To begin, suppose the law is indeterminate. Here, the judge is like a policymaker; he chooses from a menu of options. His decision, which is necessarily nonlegalistic, yields a social payoff that reflects the sum of utility gained by the winning party, the utility lost by the losing party, and all the utility gains and losses experienced by individuals affected by, but not directly involved in, the litigation. A "good" or "wise" decision might generate a lot of utility, while a "bad" or "foolish" decision might generate very little.

    I can make these ideas concrete with a concept from political science. The median voter theorem posits that, under certain conditions, the policy that the median voter prefers will defeat all alternative policies in a head-to-head vote. (154) If additional conditions are satisfied, then the policy that the median voter prefers also represents the policy that maximizes aggregate utility. (155) The conditions are somewhat technical, and I will not examine them in the text. (156) Instead, I will illustrate with an example.

    Suppose three voters are choosing from three tax rates: 5%, 10%, and 15%. The "conservative" voter prefers 5%, and his utility declines as taxes go up from there. The "liberal" voter prefers 15%, and his utility declines as taxes go down from there. The "moderate" voter prefers 10%, and his utility declines as taxes go up or down from there. The moderate is the median. His preferred tax rate, 10%, would defeat each alternative rate two to one in a head-to-head vote. (The conservative and the moderate prefer 10% to 15%, and the moderate and the liberal prefer 10% to 5%.) If the utility the conservative gains when the rate approaches 5% exactly offsets the utility the liberal loses and vice versa, then the moderate's preferred rate maximizes the group's utility. If the rate increases beyond 10%, the gains and losses of the conservative and the liberal would wash out, but the moderate would lose utility, leading to a net loss. Likewise if the rate dropped below 10%.

    I assume that the conditions are satisfied such that the median citizen's preferred policy is also the policy that maximizes aggregate utility. This assumption simplifies and lends precision to the analysis. Now the social payoff of the judge's decision depends entirely on the congruence between the judge and the median citizen. (157) The closer the judge's decision is to the median citizen's preferred decision, the greater the social payoff of the judge's decision, and vice versa.

    This assumption might seem implausible for any particular case, but it seems more plausible for a representative case, which, as discussed, is the focus of the framework. For every case on the docket in which a decision "left" of the median would maximize utility, there is an offsetting case in which a decision "right" of the median would maximize utility. In any event, the assumption is not critical to the analysis. If it does not hold, the question simply changes from "which judge is more congruent with the median citizen?" to "which judge is more congruent with the citizen whose ideal outcome would maximize aggregate utility?" (158)

    The variable c represents the social payoff of the decision by an independent judge in a case where the law is indeterminate. It reflects the congruence between the judge and the median citizen. The expression (c + a) represents the payoff of the decision by a dependent judge in this circumstance. The variable a reflects the change in congruence that results from making the judge dependent on an outside authority. The variable a is positive if dependence makes the judge more congruent with the median citizen and negative if dependence makes him less congruent with the median citizen.

    Now suppose the law is determinate, and suppose the judge makes a nonlegalistic decision. As above, that decision yields a payoff reflecting the sum of utility gained by the winner, lost by the loser, and gained or lost by third parties. That sum depends on the congruence between the judge and the median citizen. The intuition runs like this: the judge is not making a legalistic decision but rather a policy choice. Greater congruence between that choice and the median citizen's preferred choice implies a higher payoff, and vice versa.

    One might suppose the payoff also depends on another factor, the nonlegalism itself. This is, after all, a decision that contradicts determinate law. Such a decision might reduce social welfare by undermining reliance interests and predictability in legal obligations, for example. I ignore that complication. As discussed in the last Section, I treat decisions in cases where the law is indeterminate the same as decisions in cases where the law is determinate but the judge does not follow it: all are nonlegalistic decisions. I do so not because such decisions are fundamentally the same but because accounting for the difference would complicate the analysis without changing the insights I am interested in. Again, I prove this assertion in the Appendix. (159)

    Consistent with that approach, the payoff of all nonlegalistic decisions is a function of congruence with the median citizen and congruence alone. The payoff of a nonlegalistic decision by an independent judge is always c, and the payoff of such a decision by a dependent judge is always (c + a).

    Finally, suppose the law is determinate, and suppose the judge makes a legalistic decision. The resulting social payoff depends on two factors. First, it depends on the value of legalism itself. A legalistic decision might, through the promotion of reliance interests and predictability, generate a lot of utility. Second, it depends on the congruence between the decision and the raw policy preferences of the median citizen. By "raw preferences," I mean preferences of the median citizen in the absence of any legal rules. In other words, if the median citizen were fully informed and writing on a blank slate, with no reliance interests or other legal values involved, what policy would he select? The closer the legalistic decision is to that policy, the greater the utility that flows from the decision, and vice versa.

    I capture all of these ideas with simple notation. A legalistic decision yields a social payoff labeled l. (160)

    To be clear, there are two sources of utility in l: adherence to legalism and the result that legalism produces. Because of the former, a legalistic decision that does not align with the median citizen's raw preference may generate more utility than a nonlegalistic decision that does. Society may be better off with an "off-median" decision that furthers legalism. Because of the latter, a legalistic decision that does not align with the median citizen's raw preference may generate less utility than a nonlegalistic decision that does. Society may be better off with an "on-median" decision that undermines legalism.

    This leads to some important clarifications. If a judge makes a nonlegalistic decision, then the welfare-maximizing decision is always the decision that aligns with the median citizen's raw preferences. If a judge makes a legalistic decision, that decision may or may not generate more utility than a nonlegalistic alternative that is more congruent with the median citizen's raw preferences.

    All of this can be summarized in short order. If the judge, whether independent or dependent, makes a legalistic decision, the social payoff is l. If the judge is independent and makes a nonlegalistic decision, the social payoff is c. If the judge is dependent and makes a nonlegalistic decision, the social payoff is (c + a).

    One final note. I express the social payoff of the judge's decision in terms of aggregate utility. I do so because I assume that aggregate utility is an appropriate proxy for social welfare. Of course, that is contestable; utilitarianism has critics, (161) and there are many alternative conceptions of social welfare. (162) To be clear, the framework is not limited to utilitarianism. It could accommodate a variety of conceptions of social welfare. One would have to reformulate values for legalistic decisions (l) and nonlegalistic decisions (c and (c + a)) in light of the new conception of social welfare. It would not be a straightforward aggregation of individual utilities, but the framework itself would not change.

  2. Decision Tree

    Figure 1 captures the entire framework in a decision tree. Readers who do not care for the display and the math can safely skip this short Section.

    Following the tree provides an easy summary. Society selects an independent or a dependent judge. The judge is then presented with a case. There is a d probability that the law as applied to the case is determinate, in which case the judge must choose between making a legalistic and a nonlegalistic decision. The probability that an independent judge makes a legalistic decision is p, and the probability of that judge making a nonlegalistic decision is (1 - p). The probability that a dependent judge makes a legalistic decision is (p + i), and the probability of him making a nonlegalistic decision is (1 - p - i). The payoff of a legalistic decision is always l. The payoff of a nonlegalistic decision is c if the judge is independent and (c + a) if the judge is dependent. There is a (1 - d) probability that the law is indeterminate, in which case both the independent and the dependent judge make a nonlegalistic decision with payoffs of c and (c + a), respectively.

    In addition to providing a visual summary, the decision tree facilitates the translation of concepts into math. For the mathematically inclined, the social payoff of having an independent judge decide the case is

    d(pl +...

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT