|公 法 评 论
et revelabitur quasi aqua iudicium et iustitia quasi torrens fortis
Three Roles for a Theory of Behavior in a
Theory of Law*
a talk by Lewis Kornhauser
Professor Lewis Kornhauser gave this presentation to Stanford Law School faculty on February 24, 1999. In the presentation, Professor Kornhauser focused on the third section of a four-part project, laying out a taxonomy for instrumental approaches to the law. The three types of instrumentalism that comprise Professor Kornhauser’s conception of the law are systemic instrumentalism, institutional instrumentalism, and rule instrumentalism.
Instrumentalism assumes that law is created with a particular result or goal in mind. If the particular goal is not achieved, then the legal rules and institutions designed to forward it are not effective. Instrumentalism assumes that legal rules and institutions can shape the behavior of private individuals, which in turn assumes that these private individuals will react in a predictable way to the incentives created. A common form of instrumentalism is the economic analysis of law, which uses economic theory to describe how people behave in response to the law. This does not necessarily mean that people will try to obtain the most wealth, only that they will try to maximize their subjective preferences. For example, people who value the environment may go out of their way to use recycled goods, even if these cost more than non-recycled goods. Instrumentalism tries to determine what preferences people have in order to create laws that shape their preference-maximizing behavior in a way that achieves the laws’ goals. Thus one main question instrumentalists ask is what theory of behavior should the instrumental designer of legal rules and institutions use in order to create effective laws.
The behavioral theory used to design a law or institution varies according to the desired goal. Kornhauser argues that this goal can be directed through differing legal structures operating at different levels of the legal system. These different levels of the legal system comprise his taxonomy. As will become clear, one can be instrumental at one level and unconcerned about the result at the other two levels. Moreover, the different levels are not hierarchical, in that the existence of instrumentality at one level does not imply instrumentality at another level. However, this does not mean that one cannot be instrumental at a combination of levels. In order to achieve one’s goal, one must first know at which level that goal is directed and then must choose an appropriate behavioral theory so that the law can be shaped to achieve the desired result.
Systemic instrumentalism focuses on the legal system as a whole. Different legal systems can promote different goals and may therefore need to be structured differently. For example, if one believes that the purpose of government is to promote stability at all costs, then one might endorse an authoritarian form of government as necessary to keep the unruly masses in check. The goal of stability at all costs might lead to the creation of a system in which the military both legislates and enforces the law, with the courts being an afterthought at best. If one sees the purpose of government as enabling people to pursue self-selected ends in a democratic fashion, then one might design a system containing an elected legislature held in check by a separately elected executive and a relatively independent system of courts meant to protect individual rights. Alternately, one might design a parliamentary system with a combined legislature and executive to promote public ends. These same presidential and parliamentary systems may not be created to achieve democratic goals, but instead because accommodating public desires and tensions is seen as the best way of promoting stability. The instrumentalist tries to design a legal system that best fits the desired goals, even though the goals of a legal system and the means to achieve them may change.
Legal institutions, and for that matter legal rules, operate within an overall legal system, but they need not share the same goals as the system. Institutional instrumentalism addresses the goals of particular legal institutions and specifically designs these institutions to fulfill those particular goals. The motivation behind the formation of an institution does not have to be shared by the people within the institution. Kornhauser describes Ronald Dworkin’s theory of adjudication,1 which argues that the purpose of courts is to promote integrity in the law, as an instrumental approach to a particular view of justice, even though Dworkin expects judges to act formalistically, not instrumentally. Kornhauser sees these two positions as being consistent with each other despite their seeming contradiction. Formalism is the opposite of instrumentalism because it presumes that one will interpret the law "in its best light" independently of one’s own personal goals. The institution of the courts may be designed with the instrumental purpose of promoting the integrity of the law; however, judges promote the integrity of the law only when they interpret the law in a formalist way. In other institutions, such as the Internal Revenue Service, the purpose of the institution may be more synchronous with the personal goals of its participants—but this only shows that the behavioral approach of those who execute the functions of the institution must be taken into account when the institution is first designed.
Rule instrumentalism operates similarly to institutional instrumentalism, but only at the level of specific rules or laws. It is therefore not surprising that the rulemakers’ purpose in designing a rule may be different from the motivations driving compliance with it. For example, a law may be passed for the purpose of protecting the environment by controlling factory emissions. This law might achieve its goal, though, by providing financial incentives to those corporations that meet certain standards, or alternately, by imposing financial penalties on those who do not. It does not matter in this case if company managers care about the environment, only that they care enough about corporate profits to comply with regulations. Some corporations may care about the environment enough that they would comply with recommended standards regardless of related financial incentives, but presumably those passing the law do not believe that there are enough corporations with such motivation, otherwise there would not have been the need to include the financial incentives in the law. Additionally, institutional and systematic restraints inhibiting the passage of certain laws may exist. If no institution could enforce environmental protection laws, then it would be pointless to pass the law in the first place. This would be an institutional restraint on legislative power. A systemic restraint would be the refusal of American legislative bodies to pass laws which they believe violated the U.S. Constitution, for example, by creating an "environmental czar" with absolute powers.
We see that the analysis of systems, institutions, and rules can take place at different levels, even though these levels may at times interact with each other. Furthermore, when designing legal systems, institutions, and rules, one must be aware that the goals one has in mind for the legal entity that one is designing may not be the same as the motivations of the participants who implement this design.2 Indeed, it is often necessary to assume different purposes on the part of the participants and shape one’s design accordingly. In as much as one is being instrumental in designing laws, the evaluation of the effectiveness of one’s design depends upon whether it achieves the desired goal.
In this respect, it is important to distinguish between regular and institutional legal designs. Each type of legal design faces different constraints. Regular designs consist of rules that directly command or prohibit a certain action by the general population. For example, a state legislature may pass a law making it illegal to sell pornography within five miles of a school. This is a direct prohibition applying to everyone in the state. This law is then enforced through the police and the courts. The decision that a judge makes in a particular case is likewise an example of a regular design, though of a different nature than the one passed by the legislature. Adjudicatively, the judge may have some discretion in deciding what is pornography, though the judge’s discretion is likely limited by past judicial decisions. Furthermore, if a judge decides that the accused is in fact selling pornography, the judge cannot say that a distance of one mile away from a school is "good enough" to avoid punishment. The judge is constrained by what the law actually says.3 Institutional designs consist of the structures that establish the framework by which regular designs can be created and enforced. In this example, the institutional design would examine the state legislative body itself and the various rules that govern its procedure when passing such laws. The judge might also look to see if the proper institutional procedures were in fact followed, and invalidate the law if they were not. More broadly, the judge may look to the U.S. Constitution to see if the anti-pornography law violates the principles on which the legal system as a whole was founded. While the judge may not strike down the law because of personal opposition to it, the legislature may change its mind and pass a contrary law at will. To this extent, those who make the law are less constrained than those who interpret it because the latter are always constrained by existing law—as elaborated in both regular and institutional designs—whereas the former are only constrained by existing institutional designs.
The difference between regular and institutional designs leads Kornhauser to question the types of behavior the instrumental designer of laws should choose to impact in order to create effective laws.4 We have already seen suggestions of how this may differ according to the level—system, institution, or rule—at which one’s legal design is directed. The distinction between regular and institutional designs elaborates upon the earlier taxonomy by describing how different legal designs may be appropriate for different levels of the taxonomy. Systematic instrumentalists focus on institutional design, whereas rule instrumentalists operate through regular designs. Institutional instrumentalists combine regular and institutional designs. The different designs that one uses and the correspondingly different intended targets may require different assumed behavioral theories to achieve the desired result.
Kornhauser outlines three different theories of behavior often used by the institutional instrumentalist. "Best behavior theory" assumes people will do their jobs to the best of their ability, either morally or efficiently. "Worst behavior theory" assumes people will try to take advantage of the situation or that they will be incompetent. "Best predictive theory" chooses either of the above theories depending on how the instrumentalist predicts most people will behave in a particular situation.
The best behavior theory seems overly optimistic, as it would create the danger that those who wanted to take advantage of the rule, institution, or system could do so without fear of reprisal. At the same time, though, one may want to design systems, institutions, and rules that reward those who do act for the good of the whole. Public officials often choose their careers out of a desire to further the public good. Therefore, it may be appropriate to use a best behavior approach when engaged in institutional design, keeping in mind that this approach might not necessarily apply to the laws created as part of regular designs because the general public may not share the same motivations as public officials. In designing rules for the general public as part of a regular design, it may be appropriate to adopt the best predictive approach, since this would most closely accommodate how the general public as a whole would respond to the law. The best predictive theory approach and the best behavior approach run the risk of creating a system of laws that allow individuals to take advantage of the situation. However, assuming the worst from people may lead to laws that are so over-protective as to accomplish little. Furthermore, it is unclear what exactly "worst" would mean. At first glance, one may assume it means that people will act in a self-interested fashion. Yet, instrumentalism is based on this very assumption. As far as instrumentalism is concerned, the "worst behavior" would be behavior that is irrational and random. However laws that attempt to counteract randomness are likely to fail because it is impossible to know what to counteract.
In the end, the decision of which theory of behavior to adopt is highly contextual. An understanding of Kornhauser’s taxonomies will help guide us in choosing the appropriate theory of behavior to create the most effective laws possible.