No best answer?

• Author: Kornhauser, Lewis A.
• Position: Symposium: Law and Incommensurability

INTRODUCTION

In the late 1960s, Ronald Dworkin first advanced the controversial claim that there is a right answer to every legal question.(1) Debate over this claim generally has focused on the question of judicial obligation in "hard cases." When a promisor contests the validity of a contract, for instance, is it always the case that the contract is either valid or invalid? Or, is it possible that no right answer exists to this question? The controversy began with Dworkin's initial advancement of the claim,(2) persisted through the publication of Law's Empire,(3) and was then relegated, without resolution, from the mainstream legal literature to the literature on legal philosophy.

Recently, this debate has been revived in the context of conflicting values, rather than in the context of the validity of conflicting legal claims. The focus of debate thus has shifted from the existence of right answers in matters of principle to the existence of best answers in matters of policy. In these latter circumstances, the legal decisionmaker often must balance conflicting values; the "no best answer" thesis asserts that there is no best way to do this. Questions of the appropriate balance of conflicting values are omnipresent in public policy. Although on some accounts of adjudication these questions may arise in adjudication when a court seeks to determine the validity of a legal claim,(4) they usually arise in the context of either administrative rulemaking or judicial review of agency actions. Courts, for example, have had to review decisions by the Environmental Protection Agency (EPA) to set standards for lead additives to gasoline,(5) decisions by the Occupational Safety and Health Administration (OSHA) to set standards for cotton dust emissions,(6) and decisions by the EPA and the Department of the Interior (DOI) restricting the above-ground use of pesticides containing strychnine.(7)

When one turns from conflicts of rights to conflicts of value, the claim that a right answer exists is transformed into a claim that there is a best answer to every collective decision problem. This claim is often attributed to economic analysts of law, in general, and to advocates of cost-benefit analysis, in particular. Opponents of the claim generally rely on the literature in ethics that denies the existence of a best answer to many individual decisions. These opponents, however, too often simply adopt, without modification, the arguments for incommensurability in ethics to problems of law. In this Article, I argue that such straightforward adoption is inappropriate.

To begin this analysis, one needs to understand the nature of the debate in ethics. Part I thus offers a characterization of this debate that identifies two distinct strands of the debate. Each strand leads to different consequences for the implications that incommensurability has for law. I identify these strands as "instrumental incommensurability" and "expressive incommensurability." The two Parts following Part I address the implications of each form of incommensurability for law. Part 11 considers the problem of instrumental incommensurability in the context of law. I suggest first that the case for instrumental incommensurability of values in law has not been adequately made. Subsequently, I accept, for purposes of argument, the claim that legal decisionmakers must choose in the face of instrumentally incommensurable values and ask what consequences these "choice" situations might have for legal decision. Part III investigates the relation between the expressive aim of action and the expressive aspect of law discussed in recent legal literature. The Article ends with a brief conclusion.

1. INCOMMENSURABILITY IN PRIVATE DECISION

   Theories of practical reason differ in the role they assign to reason in an agent's choices. Rationalists assign reason a comprehensive role in decisionmaking; the agent is both instrumentally rational and ends-rational. Rationality identifies the ends that the agent ought to pursue and then dictates that she choose the option that best promotes her rationally given ends. In contrast, decision theory, as practiced by economists, requires only that the agent be instrumentally rational, in that she must choose the best means to further her subjectively endorsed ends.

   Claims of incommensurability arose as a challenge to decision theorists and (consequential) rationalists.(8) As the following discussion demonstrates, the incommensurability claim delineates two "substantive"--as opposed to instrumental--roles that reason may play in choice. First, as ends-rationalists claim, an agent's ends are subject to rational criticism on substantive, as opposed to merely coherence, grounds. At a minimum, then, reason would dictate that some ends not be pursued. Most strongly, reason might dictate which ends ought to be pursued. Second, reason might constrain the way in which the agent integrates her ends, whether or not they are rationally mandated.

   Decision theory constrains this integration in that it requires the agent to have a complete and transitive all-values-considered ordering, generally called the agent's "preferences." A more stringent integration--rationality would impose additional constraints on the relation between an agent's underlying ends and her all-values-considered ordering. On this account, then, a fully rational agent would be required to integrate all her rationally defined ends into a single, rationally defined, all-values-considered ordering, and to choose the option that ranked highest on this ordering.

   More generally, this analysis defines a three-dimensional classification scheme for theories of prudence. The first dimension, that of ends-rationality, addresses the extent to which the individual's ends are rationally required. Integration-rationality, the second dimension, concerns the extent to which rationally required constraints on integration restrict the set of admissible all-values-considered judgments. The final dimension, instrumental rationality, mandates that the agent adhere to her all-values-considered ordering.

   Claims of incommensurability, when properly understood, challenge the ability of consequentialist reasoning to meet the demands of integration-rationality. This characterization of the incommensurability debate, however, requires defense because, on its face, the incommensurability debate presents a bewildering set of conflicting characterizations. This confusion has arisen for several reasons. First, the scope of the incommensurabilist's critique varies. At times, the claim of incommensurability encompasses only conflicts among values (defined more precisely below(9)); at other times, the claim of incommensurability encompasses not only conflicts among values, but also conflicts between values and rights, or among rights.(10) In this Article, however, I restrict attention to the more circumscribed question of conflicts among values alone.(11) Second, as elaborated more fully below, authors writing about the conflict among values have adopted different definitions of incommensurability. Third, and most important, the incommensurabilists agree in their aversion to standard decision theory, but they disagree about the theory of value and rational choice that should underlie decisions. Different underlying theories of value have varying consequences for the implications of incommensurability and, sometimes, for one's entire understanding of incommensurability.

   In this Part, I adopt a distinction suggested by Elizabeth Anderson's theory of value between the instrumental and expressive aims of choice.(12) On her account, consequentialists admit only an instrumental aim of choice, as choices are meant to promote ends. For Anderson, however, choice may also (or instead) express the end rather than promote it. Claims of incommensurability, including Anderson's, view incommensurability of value as problematic for this instrumental aim, but the different conceptions of value endorsed by the varying critiques imply different consequences and remedies to the problem.(13)

   Each of these two aims of action has an associated argument concerning incommensurability. I shall call these two arguments "instrumental incommensurability" and "expressive incommensurability." In Part I.A, I argue that the best interpretation of the core set of arguments concerning instrumental incommensurability introduces a third role for rationality--he specification of the conditions imposed on the integration of an agent's multiple ends. In Part I.B, I turn to the relation of the expressive aim of choice to discussions of incommensurability. Some discussion here is important because the applications of the incommensurability claim to legal settings appeal to expressive interests.

   1. Defining Instrumental Incommensurability(14)

      Arguments for incommensurability begin with two observations: first, that agents pursue multiple values, and second, that these values sometimes conflict. When such conflict arises, the agent cannot (always) identify an option that best promotes her projects, all values considered. A "value" (V) for these purposes is a criterion for assessing "options." For most, if not all, purposes, these assessments are comparative. Hence, I am concerned with the associated value relation on options that has the structure of "better than (with respect to V)." This relation satisfies the two properties of completeness and transitivity: (1) (completeness) for any two options A and A either A is better than B in terms of V, B is better than A in terms of V, or A and B are equally good in terms of V, and (2) (transitivity) if, for any three options A, B, and C, A is better than B (in terms of V) and B is better than C (in terms of V), then A is better than C (in terms of V).(15)

      The relation "A is incommensurable with B," however, applies not to values but to options; that is, to the choices available to an individual...