Natural reasoning in the common law

Under the natural model of common law decision-making, courts resolve controversies as ordinary reasoners do, through moral and empirical reasoning. We assume that moral reasoning follows the Rawlsian method of wide reflective equilibrium (Rawls 1951; Daniels 1979; Klepper 1995).

correct, by reconsidering her judgments about its application to other cases. In other words, the reasoner moves between principles and par­ticulars to reach a better understanding of both the moral values at stake and their implications for the case she must decide.

Suppose, for example, that a case arises in which a homeowner sues to enjoin a proposed halfway house for newly released prisoners from opening in a residential neighborhood. The homeowner's evidence sug­gests that a halfway house may increase traffic and that newly released convicts may engage in behavior that provides bad examples for local children. At the time of the case, the halfway house is in a planning stage and its sponsors have invested no funds. Evidence presented by the halfway house indicates that it will house only non-violent crimi­nals such as minor drug offenders and will help these former prisoners transition successfully into society.

The judge's initial sense of the case is that the halfway house should be allowed. Its burden on landowners is not great, and the halfway house will have difficulty finding an auspicious location if homeowners are given a legal veto. To support this outcome, the judge formulates a principle: land use that does not pose a significant threat to health or safety of neighbors should be permitted.

The judge then tests her principle against a wider range of decided and imagined cases. In her view, a pet bear, a rifle range, an indoor trampoline fun-house, and a mortuary should be enjoined, but a day care center and a sewage treatment plant should be permitted. Her initial principle, that property owners should be allowed to engage in activities that pose no significant danger to health or safety, confirms her judgment about the bear (risky), the rifle range (risky), the day care center (low risk), and possibly the sewage treatment plant (aes­thetically displeasing but probably low risk). But this principle fails to confirm her judgment about trampolines and mortuaries.

At this point, the judge must then either reformulate her principle or change her assessment of particular cases. She might, for example, expand the principle to hold that land uses that do not pose signifi­cant danger and also serve important community interests should be allowed. This principle fits her initial judgment that the halfway house should be allowed and that the trampoline fun-house should not. The mortuary case still presents a problem, because the community has an interest in disposing of human remains; in this case the judge may need to reconsider her initial response.

Empirical reasoning also plays an important part in natural decision­making by judges, who are often called on to estimate the probable consequences of actions or the best means of implementing agreed ends. To resolve these questions, judges must gather data, form and test hypotheses, and assess risks.

The most significant way in which prior decisions affect current decision-making in the absence of a rule is by creating expectations. Parties to the original dispute will, of course, expect courts to treat a past decision as final, particularly when procedural rules support this assumption. Others who observe the legal outcome of the dispute may also form the expectation that future courts will reach similar results in like cases, and adjust their behavior in advance. Suppose, for example, that a court has held that a day care facility is not a nuisance in a resi­dential neighborhood. Day care providers may take the prior decision into account in deciding where to locate their businesses.

Without more, third party expectations of this kind are not necessarily justified expectations. The reasonableness of an investor's expectation depends on whether the investor has sufficient reason to think that future courts will take non-party reliance on the prior decision into account as a reason to reach a similar result. Nothing in the natu­ral model of decision-making itself supports this assumption: one court's balance of relevant practical and moral considerations does not ensure that future courts will in fact take the investor's expectations into account as a sufficient reason for decision in accord with those expectations.

There is, however, a social interest in facilitating coordination among private decision-makers (Schauer 1991a, Raz 1986, Postema 1982). The benefits that a course of action will provide for one actor often depend on the actions others are likely to take.

The desirability of generating expectations that will enable actors to coordinate their actions should not be confused with another reason sometimes given for following precedent under a natural model of reasoning: treating like cases alike (see Moore 1987; Greenawalt 1983; Coons 1987). The principle that similarly situated parties should be treated alike has intuitive appeal. It seems obvious to many that if a prior court has held that a day care facility in a residential neighbor­hood is a nuisance, later courts have a moral reason to hold that a similar day care facility, in a similar neighborhood, is a nuisance.

We disagree, for several reasons. Equal treatment comes into play only when there are no other reasons in play, or when other reasons are in equipoise. If the first day care facility was held to be a nuisance because it generated too much traffic and noise, and the second day care facil­ity also generates too much traffic and noise, the second court has a reason to reach the same result. The reason, however, is not equal treatment but similar traffic and noise. Equal treatment of day care facilities is itself of no moment.

It follows that equal treatment becomes a potential reason for deci­sion only when the current judge believes both that the prior decision was wrong, and that it has not generated reasonable expectations of consistency in future decisions.

Because the argument from equal treatment is seductive, we will elabo­rate further. There are certainly many instances in which justice is comparative and equal treatment is morally required. In the distribu­tion of collective resources or opportunities, and possibly in matters of retribution, it matters that people who are identical in terms of their distributive or retributive desert be treated alike. In the cases we are considering, however, a current judge believes that one of the parties to a prior case was treated wrongly. No sound principle of morality requires that if one party has been treated wrongly in the past, similarly situated parties should be treated wrongly in the future. Killing half an ethnic group as an act of genocide does not create an equality-based reason to kill the other half. The problem we are considering is far less dramatic but no different in its moral structure. If equal treatment of past and present litigants were a moral imperative in itself, morally incorrect decisions would corrupt the morality of law.

There are cases in which equal treatment of litigants is a legitimate consideration for judges. If a change in the common law will place prior litigants at a competitive disadvantage - for example, if a day care facility previously excluded from residential neighborhoods will be at a competitive disadvantage to a similar facility now allowed to locate in a residential neighborhood - this may count as a reason for consistency.

In sum: courts following the natural model of common law decision­making engage in unrestricted moral and empirical reasoning to deter­mine what outcome is best. Past decisions may nevertheless be relevant in several ways. A judicial practice of honoring final judicial determina­tions of the rights of litigants enables those litigants to manage their subsequent activities and investments more confidently. A judicial practice of consistency with prior decisions enables third parties to predict the future pattern of decisions, form expectations, and coordi­nate their conduct around those expectations. For those who reject our views about equal treatment, there may be intrinsic good in equalizing results between past and present litigants. The natural model, how­ever, does not treat past judicial decisions as legally authoritative: the overall balance of reasons for decision, including whatever weight is given to supporting expectations or equalizing the treatment of parties over time, determines the outcome of judicial reasoning.

5.2