century. Statutory and case law by that time had well specified the meaning of theft of assets and the conditions for holding someone liable for deceit. Increasingly, however, objects such as electricity, ideas, and telephone lines became subject of appropriation that differed from cases for which the law had been designed Similarly, the legal principles governing fraud and deceit were developed for cases where asymmetry between the parties was limited and the truth of the matter was easy to verify. With the growth of markets for financial instruments, the asymmetry of information between seller and buyer increased, as did the value of information Someone with the power to change and adapt law had to decide whether existing legal principles that had been developed with different cases in mind should be used to resolve these cases, or whether different principles were needed, and if so, stipulate such principles Given the incompleteness of law, a crucial question is who should hold the power to interpret or make new law in the future and to resolve questions about the application of existing law to new cases. Unlike contracts, where the parties to the original contract or their assignees have the power to renegotiate, for law the question who holds residual lawmaking power is less obvious. Thus, legal systems must allocate these rights. In doing so, legal systems must address two questions: who should hold these rights in order to ensure effective law enforcement. and what factors should be considered in allocating these rights to different agents The notion that law is ambiguous or indeterminate-concepts that are close to our term"incompleteness-has long been recognized in the legal literature(Hart 1961 Solum 1999). In addition, a substantial literature has analyzed the optimal choice between standards and rules(Kaplow 1992; Kaplow 1995; Kaplow 1997). Thus, the corporations. In fact, some jurisdictions, including Delaware, do not distinguish among these two type
10 century. Statutory and case law by that time had well specified the meaning of theft of assets and the conditions for holding someone liable for deceit. Increasingly, however, objects such as electricity, ideas, and telephone lines became subject of appropriation that differed from cases for which the law had been designed. Similarly, the legal principles governing fraud and deceit were developed for cases where asymmetry between the parties was limited and the truth of the matter was easy to verify. With the growth of markets for financial instruments, the asymmetry of information between seller and buyer increased, as did the value of information. Someone with the power to change and adapt law had to decide whether existing legal principles that had been developed with different cases in mind should be used to resolve these cases, or whether different principles were needed, and if so, stipulate such principles. Given the incompleteness of law, a crucial question is who should hold the power to interpret or make new law in the future and to resolve questions about the application of existing law to new cases. Unlike contracts, where the parties to the original contract or their assignees have the power to renegotiate, for law the question who holds residual lawmaking power is less obvious. Thus, legal systems must allocate these rights. In doing so, legal systems must address two questions: who should hold these rights in order to ensure effective law enforcement, and what factors should be considered in allocating these rights to different agents. The notion that law is ambiguous or indeterminate – concepts that are close to our term “incompleteness” – has long been recognized in the legal literature (Hart 1961; Solum 1999). In addition, a substantial literature has analyzed the optimal choice between standards and rules (Kaplow 1992; Kaplow 1995; Kaplow 1997). Thus, the corporations. In fact, some jurisdictions, including Delaware, do not distinguish among these two type
laim that law is incomplete is not a novelty to most lawyers. What our theory seeks to add is that incompleteness of law is not merely a matter of choice but at a fundamental level law is intrinsically incomplete, that is, lawmakers cannot write a complete law. To enhance the efficacy of law, legal systems must therefore allocate LMLEP. The key contribution of the incompleteness of law theory is its emphasis on properly designing enforcement institutions as a response to incompleteness of law The concept of fiduciary duty discussed in this paper is an example of a highly incomplete law. This broad principle encompasses all actions that might violate the rights of principals by fiduciaries. To avoid imposing the risk of excessive law enforcement on fiduciaries however. the law must be able to exclude actions from its applicability that do not warrant liability, and must be able to do so with sufficient certainty ex ante. Specific applications of fiduciary duty can, and indeed have been carved out and codified. Examples include the duty to disclose information to shareholders and to notify them in advance of shareholder meetings, and conflict of interest provisions in corporate statutes that specify circumstances when directors may not act on their own, but must seek shareholders approval or abstain from voting Indeed, closer inspection might reveal that there are more cases where codification might be possible and desirable. Where rules can be sufficiently specified codification can safe costs for individual actors as well as law enforcers The codified parts of fiduciary duty would not form part of the residual anymore. Yet, they would still share the same value judgment and should carry comparable sanctions of companies when it comes to the application of fiduciary duty principles
11 claim that law is incomplete is not a novelty to most lawyers. What our theory seeks to add is that incompleteness of law is not merely a matter of choice, but at a fundamental level law is intrinsically incomplete, that is, lawmakers cannot write a complete law. To enhance the efficacy of law, legal systems must therefore allocate LMLEP. The key contribution of the incompleteness of law theory is its emphasis on properly designing enforcement institutions as a response to incompleteness of law. The concept of fiduciary duty discussed in this paper is an example of a highly incomplete law. This broad principle encompasses all actions that might violate the rights of principals by fiduciaries. To avoid imposing the risk of excessive law enforcement on fiduciaries, however, the law must be able to exclude actions from its applicability that do not warrant liability, and must be able to do so with sufficient certainty ex ante. Specific applications of fiduciary duty can, and indeed have been, carved out and codified. Examples include the duty to disclose information to shareholders and to notify them in advance of shareholder meetings, and conflict of interest provisions in corporate statutes that specify circumstances when directors may not act on their own, but must seek shareholders approval or abstain from voting. Indeed, closer inspection might reveal that there are more cases where codification might be possible and desirable. Where rules can be sufficiently specified, codification can safe costs for individual actors, as well as law enforcers. The codified parts of fiduciary duty would not form part of the residual anymore. Yet, they would still share the same value judgment and should carry comparable sanctions. of companies when it comes to the application of fiduciary duty principles
B. The allocation of residual lawmaking and law Enforcement Powers Once the notion that law is intrinsically incomplete is accepted, the question arises who should hold residual lawmaking and law enforcement powers. We argue that this should be determined by the lawmaking and law enforcement functions different agents perform. Legislatures are agents that make law ex ante, but typically do not exercise any law enforcement powers. Courts usually make law ex post, that is, after the critical facts of a case have been revealed. however once made case law also has ex ante implications for actions taken in the future. In addition, courts exercise law enforcement powers. An important feature of courts is that they enforce law only after some other party brings an action. This party may be the victim, or it may be a state agent, such as a prosecutor or administrative agency. The reason that courts do not act on their own initiative follows from the rule of law notion that courts should act as neutral arbiters Similar to courts but unlike legislatures, regulators combine lawmaking and law enforcement functions. Like legislatures. they make law ex ante. Regulators however, are typically vested only with limited lawmaking powers defined by certain activities or sectors, yet within the scope of their lawmaking powers, they can change the law more flexibly and with fewer procedural requirements. This allows them to be more responsive to socioeconomic or technological change than legislatures However, a similar function could be achieved by setting up a special parliamentary committee to deal with a specialized area of the law. The distinctive feature of regulators thus lies not in greater flexibility and greater expertise than legislatures, but n combining lawmaking and a particular type of law enforcement power that
12 B. The Allocation of Residual Lawmaking and Law Enforcement Powers (LMLEP) Once the notion that law is intrinsically incomplete is accepted, the question arises who should hold residual lawmaking and law enforcement powers. We argue that this should be determined by the lawmaking and law enforcement functions different agents perform. Legislatures are agents that make law ex ante, but typically do not exercise any law enforcement powers. Courts usually make law ex post, that is, after the critical facts of a case have been revealed. However, once made case law also has ex ante implications for actions taken in the future. In addition, courts exercise law enforcement powers. An important feature of courts is that they enforce law only after some other party brings an action. This party may be the victim, or it may be a state agent, such as a prosecutor or administrative agency. The reason that courts do not act on their own initiative follows from the rule of law notion that courts should act as neutral arbiters. Similar to courts but unlike legislatures, regulators combine lawmaking and law enforcement functions. Like legislatures, they make law ex ante. Regulators however, are typically vested only with limited lawmaking powers defined by certain activities or sectors, yet within the scope of their lawmaking powers, they can change the law more flexibly and with fewer procedural requirements. This allows them to be more responsive to socioeconomic or technological change than legislatures. However, a similar function could be achieved by setting up a special parliamentary committee to deal with a specialized area of the law. The distinctive feature of regulators thus lies not in greater flexibility and greater expertise than legislatures, but in combining lawmaking and a particular type of law enforcement power that is
different from the courts' law enforcement powers. What distinguishes regulators from courts is that they can enforce law proactively. In contrast to courts, regulators can launch an investigation, enjoin actions, or impose fines on their won initiative These particular features make regulators potentially very powerful law enforcers These very same features raise concerns, as regulators may use these powers excessively and thus suppress potentially beneficial actions or engage in rent-seeking activities. To optimize law enforcement it is therefore important to identify the conditions under which the benefits of lmlep allocation to regulators outweighs its potential costs C. Allocating LMLEP for Resolving Fiduciary Duty Cases According to our theory, the choice between regulators and courts depends on the degree of incompleteness of law, the possibility of standardizing, at reasonable cost, actions that may result in harm, and the degree of harm that may result from harmful actions(Pistor and Xu 2002a; Pistor and Xu 2002b) When law is highly complete, the law can determine appropriate sanctions ex ante, and reactive enforcement by courts is sufficient to enforce the law effectively When law is highly incomplete, the optimal allocation of LMLEP is determined by the degree of expected harm and the costs of standardizing actions that might result in harm An example where levels of expected harm are matched with reasonabl costs of standardizing actions are disclosure requirements for firms issuing shares to the public, or safety standards imposed on producers of pharmaceuticals, automobiles, This is the classic objective raised against regulators. See Stigler(1964); Stigler(1971); and Posner
13 different from the courts’ law enforcement powers. What distinguishes regulators from courts is that they can enforce law proactively. In contrast to courts, regulators can launch an investigation, enjoin actions, or impose fines on their won initiative. These particular features make regulators potentially very powerful law enforcers. These very same features raise concerns, as regulators may use these powers excessively and thus suppress potentially beneficial actions or engage in rent-seeking activities.6 To optimize law enforcement it is therefore important to identify the conditions under which the benefits of LMLEP allocation to regulators outweighs its potential costs. C. Allocating LMLEP for Resolving Fiduciary Duty Cases According to our theory, the choice between regulators and courts depends on the degree of incompleteness of law, the possibility of standardizing, at reasonable cost, actions that may result in harm, and the degree of harm that may result from harmful actions (Pistor and Xu 2002a; Pistor and Xu 2002b). When law is highly complete, the law can determine appropriate sanctions ex ante, and reactive enforcement by courts is sufficient to enforce the law effectively. When law is highly incomplete, the optimal allocation of LMLEP is determined by the degree of expected harm and the costs of standardizing actions that might result in harm. An example where high levels of expected harm are matched with reasonable costs of standardizing actions are disclosure requirements for firms issuing shares to the public, or safety standards imposed on producers of pharmaceuticals, automobiles, 6 This is the classic objective raised against regulators. See Stigler (1964); Stigler (1971); and Posner
or aircraft. Disclosure rules capture only one particular aspect of the relation between firms(and their agents) and investors. This is, however, an area where past experience suggests that lack of disclosure may result in substantial harm not only to current shareholders or future investors in a particular firm, but also to investors more broadly and ultimately the functioning of the financial market as a public good. When firms come to the market, investors face a lemons problem(Akerlof 1970) Incidences of misrepresentation of information may seriously discourage investments in shares, as is evidenced by market crashes in response to the revelation of stock fraud schemes (Milgrom and Stokey 1982). Thus, the expected degree of harm is high. Fortunately, however, disclosure rules can be standardized at reasonable cost Lawmakers can define the type of information that must be disclosed, and adapt these ules over time as market behavior changes or as it becomes apparent that investor require different information. They can also use this information to determine whether further action is needed, such as the initiation of proactive enforcement activities in the form of investigations By contrast, when individual actions are not expected to generate much harm, or if standardization of such actions entails high costs, allocating lawmaking and law enforcement powers to the courts is superior even when law is highly incomplete Law enforcement related to fiduciary duties is an example where the level of incompleteness is high, standardization is possible only for some areas- leaving a large undefined residual -and the expected harm is relatively contained. Fiduciary duties govern the relationship among stakeholders in a particular undertaking (management vS. shareholders, block holders vs. minority shareholders). The harm done when these duties are violated is typically confined to a subset of these same (1974)
14 or aircraft. Disclosure rules capture only one particular aspect of the relation between firms (and their agents) and investors. This is, however, an area where past experience suggests that lack of disclosure may result in substantial harm not only to current shareholders or future investors in a particular firm, but also to investors more broadly and ultimately the functioning of the financial market as a public good. When firms come to the market, investors face a lemons problem (Akerlof 1970). Incidences of misrepresentation of information may seriously discourage investments in shares, as is evidenced by market crashes in response to the revelation of stock fraud schemes (Milgrom and Stokey 1982). Thus, the expected degree of harm is high. Fortunately, however, disclosure rules can be standardized at reasonable cost. Lawmakers can define the type of information that must be disclosed, and adapt these rules over time as market behavior changes or as it becomes apparent that investors require different information.7 They can also use this information to determine whether further action is needed, such as the initiation of proactive enforcement activities in the form of investigations. By contrast, when individual actions are not expected to generate much harm, or if standardization of such actions entails high costs, allocating lawmaking and law enforcement powers to the courts is superior even when law is highly incomplete. Law enforcement related to fiduciary duties is an example where the level of incompleteness is high, standardization is possible only for some areas – leaving a large undefined residual – and the expected harm is relatively contained. Fiduciary duties govern the relationship among stakeholders in a particular undertaking (management vs. shareholders, block holders vs. minority shareholders). The harm done when these duties are violated is typically confined to a subset of these same (1974)