*821 THE SOUND OF SILENCE: DEFAULT RULES AND CONTRACTUAL CONSENT

Randy E. Barnett [FNa1]

Virginia Law Review (May, 1992)

78 VALR 821

. . . I had grasped the significance of the silence of the dog, for one true inference invariably suggests others. . . . Obviously the midnight visitor was someone whom the dog knew well.

--Sherlock Holmes [FN1]

INTRODUCTION: INCOMPLETE CONTRACTS AND CONSENT

By now everyone acknowledges the legal realist insight that all contracts are, by necessity, incomplete to some degree. The inevitability of incompleteness reflects, to borrow a distinction from H.L.A. Hart, both our "relative ignorance of fact" and "our relative indeterminacy of aim." [FN2] These two causes of contractual incompleteness *822 reflect in turn two fundamental problems addressed by liberal principles of justice: the problems of knowledge and of interest. [FN3]

Parties drafting a contract confront a serious knowledge problem. Because they cannot foresee every future event or know precisely how their own purposes may change, they cannot negotiate terms specifically to cover all contingencies. As a result, their manifested agreement will be silent as to these matters. As the duration of a contract is extended, the knowledge problem facing the parties is likely to increase and the completeness of their agreement to decrease.

Incompleteness of contracts is also a function of the parties' interests. Settling in advance even those contingencies that can be foreseen is costly. Many foreseeable contingencies, given their low probability, are better left unnegotiated ex ante in the hopes that they will not materialize or will be handled cooperatively ex post if they do. And strategic considerations may lead one or both parties to remain silent about a particular issue. [FN4]

The received wisdom concerning contractual incompleteness is reflected in the concept of "gap-filler," a legal realist term that has dominated contract discourse for several decades. According to this view, no contract is without its "gaps." Much of what is taught as the law of contract can be conceived as publicly provided "background" rules or principles that fill the inevitable gaps in the private law made by contracting parties.

This gap-filling conception of contract law has tended to undermine support for consent as the basis of contractual obligation. Gap-filling terms-- sometimes referred to as "implied-in-law"--are said to be imposed by the legal system for reasons of principle or policy rather *823 than consented to by the parties. That such implied-in-law terms are based on the parties' consent has long been thought to be pure fiction. [FN5] If the publicly provided rules of contract law almost always operate where there is a gap in the manifestation of assent, then consequently a gap-filling provision must be coercively imposed on parties who have not, by assumption, consented to its imposition. Furthermore, if the central problem of contract theory is to decide how gaps in manifested assent should be filled, then the consent of the parties plays no role in the choice of gap-fillers. In this manner, the pervasiveness of contractualincompleteness and the concept of gap-fillers makes consent look quite irrelevant to the main issues of contract theory. [FN6]

That this is the prevailing wisdom of contract theory is evidenced by the fact that scholars as disparate as Ian Macneil, [FN7] Subha Narasimhan, [FN8] and Charles Fried [FN9] adhere to it. Recently, however, in an almost imperceptible shift, the rhetoric of gap-filling has been increasingly supplanted by a new and powerful heuristic device: the *824 concept of default rules. This concept has been employed by an ideologically diverse group of contract theorists including Ayres and Gertner, [FN10] Goetz and Scott, [FN11] Coleman, Heckathorn, and Maser, [FN12] Haddock, Macey, and McChesney, [FN13] and even some scholars who have failed to overcome the transaction costs of forming a writing partnership, such as Douglas Baird, [FN14] David Charny, [FN15] Richard Craswell, [FN16] Richard Epstein, [FN17] Clayton Gillette, [FN18] and Jason Johnston. [FN19]

The default rule approach analogizes the way that contract law fills gaps in the expressed consent of contracting parties to the way that word-processing programs set our margins for us in the absence of our expressly setting them for ourselves. A word-processing program that required us to set every variable needed to write a page of text would be more trouble than it was worth. Instead, all word-processing programs provide default settings for such variables as margins, type fonts, and line spacing and leave it to the user to change any of these default settings to better suit his or her purposes. [FN20]

*825 What makes the default rule approach to gap-filling distinctive in both word processing and contract law is that default rules are binding in the absence of manifested assent to the contrary--which means that a manisfested assent to the contrary will displace the default rule. Any gap-filling rule that cannot be displaced by manifested assent is not properly called a default rule at all, but is what Ayres and Gertner have called an "immutable" rule-- that is, some other kind of contract law background norm that may fill a gap in assent or may even displace the manifested assent of the parties. [FN21]

By this criterion, many of the provisions of Article 2 of the Uniform Commercial Code (U.C.C.) are default rules, because they apply "unless otherwise agreed." [FN22] Eighteen of the twenty-eight provisions of Part 3 of Article 2, which specify the "General Obligation and Construction of Contract," contain this or comparable language. [FN23] Moreover, section 1- 102(4) specifies that " t he presence in certain provisions of this Act of the words 'unless otherwise agreed' or words of similar import does not imply that the effect of other provisions may not be varied by agreement under subsection (3)." [FN24] The default rule approach is not, of course, limited to the U.C.C. Outside the areas of fraud, duress, and unconscionability, few of the cases in law school casebooks would have arisen had the parties included an express clause in their agreement to govern the problem that arose *826 between them. In almost every case, the silence of the parties creates a gap that the default rules of contract law are used to fill.

The rhetoric of default rules is in the process of becoming the terminology of choice for contract theorists. [FN25] Yet the implications of this subtle conceptual shift for the debate concerning the underlying basis of contractual obligation--especially the role of consent--have yet to be explored. The idea of filling gaps in manifested consent with default rules calls into question the neat dichotomy between terms that are assented to and those that are imposed by the legal system. It suggests yet another category of terms that, although supplied [FN26] by courts or legislatures when there is a gap in manifested assent, nonetheless reflect the consent of the parties.

Terms supplied by default rules are not a product of the expressed or implied-in-fact consent of the parties as these two notions have traditionally been understood, and may therefore be considered genuinely implied-in-law. But neither are terms supplied by default rules invariably imposed on the parties by the legal system for reasons of principle or policy as are terms supplied by immutable contract rules or by tort law. In a very real sense, such terms can be and often are indirectly consented to by parties who could have contracted around them--but did not.

In this Article, I want to challenge the received wisdom of "gap-filling in the absence of consent" by showing how the concept of default rules bolsters the theoretical importance of consent. I will accomplish this by expanding and refining my analysis of a "consent theory of contract." [FN27] I propose that the concept of default rules *827 reveals consent to be operating at two distinct levels of contract theory. First, the presence of consent to be legally bound is essential to justify the legal enforcement of any default rules. Second, nested within this overall consent to be legally bound, consent also operates to justify the selection of particular default rules.

With consent operating on both of these levels, contract terms supplied by default rules can be seen to occupy a previously undifferentiated middle ground between the traditional categories of implied-in-fact and implied-in-law terms. Under certain circumstances, it is not at all fictitious to characterize a choice to remain silent and let default rules operate as an expression of consent, albeit a more inchoate or indirect expression than what we associate with expressed or implied-in-fact terms. [FN28] And, even when parties cannot be said to have consented by their silence to the enforcement of particular default rules, enforcement may still be justified on the grounds of consent when default rules are chosen to reflect the commonsense or conventional understanding of most parties.

As will become apparent, the concept of default rules entails a very practical implication for how contract law is taught in law school. We should replace the traditional dichotomy of expressed or implied-in-fact contract terms that are "really" consented to and implied-in-law contract terms that are "really" imposed on the parties with a more realistic trichotomy. In the first category are terms that are a product of direct consent (expressed or implied-in-fact terms), in the second are terms that are a product of indirect consent (implied-in-law default rules), and in the third are terms that are imposed upon *828 the parties without any consent (implied-in-law immutable terms). [FN29] With this reconfiguration, the set of implied-in-law terms that represent a genuine imposition on the parties is much smaller than is commonly assumed.

In Part I, I elaborate in some detail on the essential social functions that consent plays in the liberal conception of justice and the rule of law. I discuss how the consensual element of contract that comprises the liberal principle of freedom of contract addresses the pervasive social problems of knowledge and of interest. Freedom of contract entails both freedom to contract--the power to effect one's legal relations by consent--and freedom from contract--the immunity from having one's rights to resources transferred without one's consent. Working together, these two components of contractual freedom harness the personal and local knowledge possessed by individuals and associations by enabling them to put their own knowledge into action while taking into account the vast knowledge possessed by others of which they are necessarily ignorant. Moreover, these components of contractual freedom also address the problem of interest by providing incentives both to use the knowledge in one's possession and to take into account the knowledge of others.

In Part II, I use this functional analysis of consent to explain the important justificatory role played by manifestations of consent. I show how, by manifesting their intention to be legally bound, contracting parties are implicitly committing themselves to the jurisdiction of a legal system that is thereby justified in using the background rules of contract law to fill the gaps in their agreement. Under certain circumstances, this "consent to jurisdiction" would also justify the enforcement of any promulgated set of default rules. In sum, a consent to be legally bound provides a necessary but only sometimes sufficient justification for enforcing whatever set of default rules may be promulgated by a legal system.

In Part III, I argue that, within a consent theory of contract, silence can be meaningful and that its meaning should influence our choice of default rules when the circumstances described in Part II are absent. I defend my "consent theory of contract" against Richard *829 Craswell's recent criticism that, whatever other justificatory virtues it may have, a consent theory of contract cannot assist in selecting among possible default rules. [FN30] In particular, I show how a consent theory provides two compelling reasons to choose default rules that reflect the conventional or commonsense understanding existing in the relevant community of discourse. First, conventionalist default rules are likely to reflect the tacit subjective agreement of the parties and thereby facilitate the social functions of consent. Second, when parties have asymmetric access to the background rules of contract, enforcing conventionalist default rules will reduce subjective disagreements by providing parties who are rationally informed of the background rules with an incentive to educate those parties who are rationally ignorant of these rules. This, too, facilitates the social functioning of consent. In sum, where a consent to jurisdiction exists but is insufficient to justify the enforcement of any promulgated set of default rules, consent justifies the enforcement of conventionalist default rules.

In Part IV, I take up the issue of why objective or manifested consent can be considered as "real" or genuine as subjective consent. I discuss how the concept of default rules not only reconciles the idea of gap-fillers with consent as the basis of contractual obligation, but also may justify a more radical change in the legal system that makes contract law. In particular, it may support a horizontal legal order composed of competing legal systems, in contrast to the relatively vertical, monopolistic legal order we live in today. Finally, in Part V, I conclude by briefly discussing the difficult methodological problem of determining the commonsense or conventional meaning of silence.

I. THE SOCIAL FUNCTIONS OF CONTRACTUAL CONSENT [FN31]

Contract cannot completely be understood apart from the role it plays in a broader scheme of entitlements. [FN32] Entitlements are concepts*830 that enable us to handle the pervasive social problems of knowledge, interest, and power. Each of these problems is multifaceted and cannot be fully explored here. [FN33] I shall confine my discussion to those aspects that bear on the question of contractual consent. [FN34] In Section *831 A, I consider how contractual consent helps to solve what I call the "first-order problem of knowledge." In Section B, I discuss how contractual consent is necessary to address two problems of interest. In Section C, I explain how the need to handle what I call the "second-order problem of knowledge" explains the proper functional relationship between objective and subjective conceptions of consent. [FN35]

A. Using Resources: The First-Order Problem of Knowledge [FN36]

Human beings confront myriad ways of using physical resources, including their own bodies. Knowledge of the uses to which resources may be put is both radically dispersed among the persons and associations that comprise a society, and highly contingent upon the particular circumstances facing each person. The problem of a person or association making knowledgeable choices among alternative uses of physical resources is compounded by other persons and associations striving to make their own choices based on their own knowledge. This radical dispersion of knowledge gives rise to what I have called the first-order problem of knowledge.

1. The Limited Accessibility of Personal and Local Knowledge

The first-order problem of knowledge arises because access to the vast range of knowledge possessed by individuals and associations is restricted. For example, each person has knowledge of his particular situation--including knowledge of his abilities, his interests, his preferences, and his opportunities--but access to this knowledge by *832 others is extremely limited. [FN37] When persons seek to act on the basis of the knowledge in their possession, such action necessarily involves the use of physical resources (including the use of their bodies). Many of these actions will conflict, in the sense that attempts by some to use physical resources to put their knowledge into action will inevitably interfere with the efforts of others to do the same.

No one has placed greater stress on this particular knowledge problem than Friedrich A. Hayek:

The peculiar character of the problem of a rational economic order is determined precisely by the fact that the knowledge of the circumstances of which we must make use never exists in concentrated or integrated form but solely as the dispersed bits of incomplete and frequently contradictory knowledge which all the separate individuals possess. The economic problem of society is thus not merely a problem of how to allocate "given" resources-- if "given" is taken to mean given to a single mind which deliberately solves the problem set by these "data." It is rather a problem of how to secure the best use of resources known to any of the members of society, for ends whose relative importance only these individuals know. Or, to put it briefly, it is a problem of the utilization of knowledge which is not given to anyone in its totality. [FN38]

For Hayek, the problem of knowledge does not arise because one person necessarily has "better" information than another (though this may well be the case). Rather, Hayek points to "the dispersed bits of incomplete and frequently contradictory knowledge which all the separate individuals possess." [FN39] The first-order problem of knowledge unavoidably arises because each person and association of persons possesses knowledge that is inaccessible to others.

The radically dispersed knowledge described by Hayek may be either "personal" or "local," depending upon the degree of its accessibility. Individuals have access to their own personal knowledge that *833 others necessarily lack. [FN40] Only I know what I am thinking and feeling as I write this passage; only I can observe the room I am working in from this vantage point; only I know that I would now like something to eat. A list of one's personal knowledge would be both endless and impossible to compile. [FN41] One's personal knowledge includes knowledge of one's personal preferences, to be sure, but it also includes a knowledge of one's physical and emotional needs, one's particular circumstances, and one's opportunities or alternative courses of action. Indeed, it is impossible to enumerate precisely all the types of knowledge each of us possesses, much less the knowledge itself. [FN42]

In contrast to personal knowledge, local knowledge is knowledge that is publicly accessible, though this access is still quite limited. Instead of being confined to a single person, access to local knowledge is limited to particular associations of persons. A dinner conversation between two people in a crowded restaurant is accessible to both conversants, but not to everyone in the restaurant. Even persons at the next table may be hard-pressed to understand what is being said. Knowledge need not be limited to a few people to be local. Sixty thousand people viewing a football game in a stadium have local knowledge of the game in the sense that the rest of the world does not have access to what they can observe about the game in progress. Even if millions more are watching on television, such knowledge would still be local because billions lack access to it.

In sum, as used here, personal knowledge refers to an individual's knowledge that is inaccessible to others. Local knowledge refers to knowledge to which only certain associations of persons have access. *834 These terms refer not to the substance or origin of knowledge, but to the limited access that persons have to it. This radical dispersion of personal and local knowledge creates what I will call the first-order problem of knowledge. This problem has two aspects or dimensions: First, one must be able to act on the basis of one's own personal knowledge or the local knowledge one has access to as a member of an association. Second, when so acting one must somehow take into account the knowledge of others of which each person is hopelessly ignorant.

The dispersal of personal and local knowledge can be pictured as a "knowledge glass" that is both half-full (what each of us knows) and half-empty (what each of us is ignorant of). The first-order problem of knowledge facing any society is how to put to use the half that is full while at the same time taking into account the half that is empty. For persons to survive and flourish in society with each other, they must be able to develop and act upon their own personal and local knowledge. Their actions, however, are likely to affect others in ways that can scarcely be known. And the first-order problem of knowledge is magnified because what each of us knows and wishes to put into action is dwarfed by our ignorance. Rather than being half-full, each of our knowledge glasses contains but a precious drop.

The first step in addressing this problem is to recognize that it is the use, not the mere possession, of personal and local knowledge that creates a problem when persons live in society with others. Only a person's actions, not his or her knowledge, can interfere with the ability of others to act on the basis of their own knowledge. [FN43] Ideally, what is sought is a relational or social order in which the use of everyone's knowledge is possible. Although differing preferences and opinions can lead to conflicting actions, we need not control preferences and opinions themselves to handle the problem of conflicting action. We need only control actions, and only those actions that impede the ability of others to put the knowledge in their possession to good use. *835 In sum, to solve the first-order knowledge problem requires only a relational "order of actions," [FN44] not an order of preferences.

An order of actions requires some scheme in which conflicts among actions are minimized. Human action must occur during particular periods of time and in particular physical spaces; this imperative is reflected in the term "order" itself. An order of actions initially suggests a scheme of temporal priority. ("First her actions, then his.") But spatial priority is another dimension of order. ("She acts over here; he acts over there.") Thus, to achieve an order of actions one must regulate the use of physical resources. A relational order of actions is achieved when the individual or associational uses of physical resources are temporally and spatially coordinated so as to reduce or eliminate the possibility that two persons or associations will attempt to use the same resource at the same time. If human actions can be suitably regulated, then we need not attempt to remold or coordinate personal or local knowledge itself. Not just any relational order will do, however. We want an order of actions in which personal and local knowledge can be developed, disseminated, and acted upon, and some ordering methods will perform this function better than others.

2. Two Methods of Attaining a Relational Order of Actions

Two quite different methods of achieving a relational order of actions are centralized and decentralized ordering. Although both methods are essential to attaining a relational order of actions in any society, they are not equally suited to address the first-order problem of knowledge.

a. Centralized Ordering [FN45]

The idea of centralized ordering of society as a whole is both attractive and plausible in light of its familiarity. The family is organized in this way, with parents making decisions about the disposition of family *836 assets among the family members. Larger commercial firms are organized this way as well, with a hierarchical association of persons called "management" making decisions about using the resources of the company, subject to the approval of a board of directors. The military, with its extremely well-defined chains of command, is perhaps the paradigm of centralized ordering.

Moreover, centralized ordering is undoubtedly a valuable method of capitalizing upon both personal and local knowledge. One individual acting as a central director or planner can effectively order the actions of other persons so as to capitalize on the planner's personal knowledge. For example, centralized ordering can harness a parent's personal knowledge of the needs of her child, an entrepreneur's personal knowledge of an unfulfilled demand in a market, or a field officer's personal knowledge of a tactical situation in combat. Or centralized direction can capitalize upon the local knowledge of an association. For example, it can use the local knowledge of a husband and wife, the talented managers of a corporation, or a military command.

Yet, despite its undeniable advantages, centralized ordering is completely unsuited to handle the first-order problem of knowledge. Suppose we delegated to some person or association the responsibility for coordinating resource use in accordance with the diverse knowledge of all persons and associations in society. To achieve an overall orderof actions with such a strategy, some person or identifiable set of persons would somehow have to (a) obtain the personal and local knowledge of all persons and associations, (b) incorporate this knowledge into a coherent or coordinated plan of human actions, and (c) transmit instructions on resource use consistent with this plan to everyone in the society so that persons could act accordingly. Intractable problems arise at each step in this process.

The very strength of centralized direction in capitalizing on the personal and local knowledge of central directors is at once its weakness as a strategy for solving the first-order problem of knowledge. Centralized ordering is especially effective when those in charge of the ordering scheme have access to useful personal or local knowledge. But, although central directors have access to their own personal and local knowledge, they plainly lack access to the ever-changing totality of personal and local knowledge dispersed throughout an entire society.

*837 In sum, centralized direction cannot solve the first-order problem of knowledge in society at large because central directors cannot possibly have access to the personal and local knowledge that such an ordering strategy requires. They are hopelessly ignorant of the knowledge needed to achieve an order of actions that would permit persons to put to use their personal and local knowledge. Moreover, they lack the capacity to integrate the necessary knowledge into a coherent plan and to communicate to all their allocated roles. Together, the three essential elements required by a centralized ordering strategy are nonexistent when this method is applied to govern resource use in an entire society. In this regard, the impossibility of central planning or true socialism stems not from the impossibility of centralized direction simpliciter, but from the impossibility of using centralized direction writ large to handle the first-order problem of knowledge. It is for this reason that centrally-planned economies have, without exception, failed miserably to serve the public welfare. [FN46]

b. Decentralized Ordering

How could the first-order problem of knowledge possibly be addressed by anything except central direction without immediately descending into chaos or disorder? The answer involves the concept of jurisdiction. A jurisdictional strategy attempts to handle the first-order problem of the radical dispersion of personal and local knowledge by using the idea of "bounded individual and associational discretion." This method of social ordering defines a jurisdiction or domain within which an individual or association is free to act on the basis of personal and local knowledge.

Implicit in this jurisdictional strategy is a crucial distinction between the judgment maker and the judgment to be made. To use the language of American sports, such a strategy distinguishes the jurisdictional question, "who makes the call?" from the substantive question, "what is the correct call?" [FN47] To answer each question *838 requires substantially different knowledge. Answering the second question requires personal and local knowledge of particular circumstances--knowledge that is inaccessible to centralized mechanisms. Answering the first question requires only that we know who is in the best position to have this knowledge.

This quality of "being in the best position to know" is one important aspect of institutional or personal competence. The knowledge needed to answer the second of these questions differs substantially from that needed to answer the question of competence. Even when we do not know the correct call, we may know who is most likely to have the knowledge that such a call requires. Instead of gaining access to the personal and local knowledge needed to make the decision in question, such an assessment requires only that we determine who is in the best position to obtain this knowledge. In baseball, for example, we may know that the umpire is in the best position to assess whether or not a pitched ball is in the strike zone without knowing anything about a particular pitch. [FN48]

The earlier discussion of personal and local knowledge suggests that individuals and associations have a comparative advantage over centralized mechanisms. [FN49] They have access to types of knowledge that centralized mechanisms lack. The fact that individual persons and institutions are generally in the best position to make the right call does not, however, mean that they will always make good use of their access or that others are never in a better position to make a particular call. Nor does it mean that an analysis of personal and institutional competence would never benefit from a substantive assessment of the right call to make. We may, in fact, bolster our assessment of personal and institutional competence by sampling a few decisions to see if they appear to reflect the knowledge we expect these persons and institutions to possess. A pattern of egregious decisions would call into question the competence of the decision maker.

*839 Still, the possibility of second-guessing the decisions of those in the best position to make a call does not change the basic analysis. Given that no decision maker is perfect, we need to make a comparative and generalized judgment when determining the appropriate jurisdictional allocation. A persistent bias in favor of centralized decision making results from an apparent ability to second-guess the wisdom of the decisions of others on occasions when these decisions go awry. Such a bias is an instance of the fallacy of the whole. It falsely assumes that what is unquestionably true about individual decisions--that others can sometimes know better--is also true of systematic decision making--that others are more competent generally. An institutional competence to second-guess the correctness of another's call on occasion does not entail an institutional competence to make correct calls for others systematically. The concept of competence does not rest on an ability to make every decision better than anyone else; it rests on being in a better position than anyone else to make knowledgeable decisions.

The idea of jurisdiction based on "bounded individual and associational discretion" is, of course, far too general to define actual conduct as permissible or subject to prohibition. It says nothing about the nature of the domain or the extent of the boundary. Nonetheless, even at this extremely general level, such a strategy is theoretically revealing in several ways. First, it identifies discretion--or liberty--as a means of capitalizing on knowledge that cannot be transmitted through a chain of command to central directors. Second, it gives discretion to individuals, who are most likely to possess personal knowledge, and to associations, which are most likely to possess local knowledge. Finally, it immediately suggests that discretion must somehow be bounded, albeit in a manner that does not undermine the purpose for adopting the strategy. The boundaries of this discretion are defined by two distinct conceptual components: (a) decentralized jurisdiction over physical resources, and (b) consensual transfers of these jurisdictions. Both concepts are reflected in the liberal conception of justice.

3. The Liberal Conception of Justice

The liberal conception of justice represents a two-part approach to the first-order problem of knowledge. First, it recognizes the jurisdictions of diverse individuals and associations over physical resources so *840 as to permit them to act on the basis of their own personal and local knowledge. Second, it allows the transfer of a person's or association's jurisdiction only with its manifested consent. This permits changes in jurisdictions to reflect changes in knowledge, while making possible a price system that enables persons to take the knowledge of others into account when deciding how to act.

The liberal conception of justice regulates the use of force in society by the concept of entitlements or rights. The first part of the strategy-- decentralized jurisdiction--is reflected in the nature and scope of these rights. Within the classical liberal approach, the rights that concern jurisdiction over physical resources are called property rights. To have property in a physical resource--including one's body--means that one is free to use this resource in any way one chooses provided that this use does not infringe upon the rights of others.

Because this concept of property protects the discretionary use of resources by private persons, as opposed to government officials, this idea is often referred to as "private property." For present purposes, however, I prefer "several property," a term also favored by Friedrich A. Hayek and the figures of the Scottish enlightenment. [FN50] The term several property makes it clearer that jurisdiction to use resources is dispersed among the "several"-- meaning "diverse, many, numerous, distinct, particular, or separate" [FN51]--persons and associations that comprise a society, rather than being reposed in a monolithic, centralized institution.

The second part of the strategy--consensual transfers only--is reflected in the concept of "freedom of contract." Freedom of contract is comprised of two distinct principles: freedom to contract [FN52] and *841 freedom from contract. [FN53] Freedom to contract holds that persons may consent to legally enforceable transfers of their property rights; freedom from contract holds that transfers of property rights should not be imposed upon them without their consent. In other words, freedom to contract permits consensual transfers, whereas freedom from contract requires them. Against a backdrop of several property, these two principles regulate the transfers of several property rights. The manifested consent of the "rights-holder" is, under normal circumstances, sufficient to transfer a property right; and property rights may not normally be transferred without the consent of the rightsholder.

Let us now consider in greater detail exactly how the two-part strategy of decentralized jurisdictions and consensual transfers addresses the first-order problem of knowledge.

a. Decentralized Jurisdiction and the First-Order Problem of Knowledge

The first-order problem of knowledge has two dimensions: human beings need (a) to be able to develop and act upon their own personal and local knowledge, while (b) somehow taking into account the inaccessible personal and local knowledge of others. Decentralized jurisdiction is the principal means of coping with the first aspect of the problem.

*842 Before explaining further, however, an important caveat is in order. I do not contend that the following abstract principles of decentralized jurisdiction can be used to determine specific allocations of rights to particular individuals. [FN54] Instead, these principles serve as functional criteria for evaluating the conventional rules that are needed to make specific allocations. In other words, these general principles cannot take the place of laws to govern the specific allocation of resources, but any such laws should be critically assessed to determine if they function consistently with these principles.

With this caveat in mind, we can characterize a strategy of decentralized jurisdiction by four desiderata. (1) Jurisdiction or discretionary control over resources must be delegated to identifiable individuals and groups. If decisions concerning resource use are to be knowledgeable, decision-making authority concerning resource usemust belong to the persons and associations with access to such knowledge. Conversely, those who lack the requisite knowledge of resource use should lack the authority to interfere with the decisions made by those with knowledge--at least as a general matter. All else being equal, the distribution of jurisdiction over physical resources should mirror as closely as possible the distribution of access to knowledge in society. Once again, I am speaking now of the types of jurisdiction that a legal system should acknowledge, not any specific allocation of jurisdiction.

(2) The allocation of jurisdiction should reflect an assessment of who is in the best position to have personal and local knowledge of the resources in question. Insuperable knowledge problems prevent us from allocating jurisdiction on the basis of which particular person or group of persons is actually in the best position to know how certain resources may be used. If a centralized institution charged with allocating jurisdictions knew what it needed to know to make such allocations, a decentralized jurisdictional strategy would be unnecessary. The most we can hope for is to determine the general characteristics of those who are in the best position to have knowledge of potential *843 resource uses, regardless of whether they in fact always have the best knowledge. In sum, we rely on these general characteristics to establish a presumption of competence in favor of individuals and groups who have access to the personal and local knowledge pertaining to their own situation.

To avoid any confusion, I again wish to emphasize what I am not claiming. I am not claiming that each individual jurisdiction should originally be allocated according to some criterion of competence or that changes in jurisdiction should be made in this manner. For example, I am not claiming that Ann should be given jurisdiction over a berry patch instead of Ben because we independently determine that Ann either knows better or is in a better position to know how the berry patch should be used. This sort of evaluation of the relative competence of particular persons and groups would be nearly as difficult as the substantive decisions that are part of the first-order problem of knowledge. Rather, I am asserting that this desideratum argues in favor of allocating jurisdiction at the level of individuals and associations who as a general matter have access to this information, rather than at the level of more global, overarching institutions that lack access to the relevant knowledge. I have as yet said nothing about how jurisdiction should be allocated among competitors for resources each of whom are on the appropriate level.

(3) The domain accorded any particular individual or group must be bounded. If the distribution of jurisdiction over physical resources mirrors as closely as possible the distribution of knowledge in society, then this also means that such jurisdiction must be limited or bounded. Because access to personal and local knowledge is limited, no one has access to all such knowledge. Consequently, to take the extreme case, no person or group should have jurisdiction over all physical resources.

(4) Because the knowledge of individuals and associations is dynamic, not static, jurisdictional boundaries must be subject to revision. Jurisdiction cannot be allocated once and for all. Knowledge of how resources may be used is constantly changing. Absent the need to continually adjust the jurisdiction of individuals and associations, we might imagine a centralized regime being able to allocate jurisdiction once and for all. A centralized regime, however, would be overwhelmed by the need to constantly readjust jurisdictional boundaries according to shifting personal and local knowledge. The restricted *844 access to such knowledge renders such an approach not merely impossible, but inconceivable.

Although the dynamic nature of the first-order problem of knowledge makes changes in jurisdiction necessary, allowing jurisdictional boundaries to change gives rise to a very ticklish knowledge problem. If a potential user were permitted to displace the present user simply on the basis of a mere assertion that he has knowledge of how resources may "best" be used, this would provide no way of assuring that the prospective user is really in any better position to use the resources than the present user. Interpersonal comparisons of knowledge (or interests [FN55]) cannot reliably be made by the parties themselves or by third parties. [FN56] What is crucial to understand is that some systemic means of transferring jurisdiction must exist that reflects the knowledge (and interests) of both parties. As I explain in the next Section, contractual consent performs this vital function.

b. How Consent Addresses the First-Order Problem of Knowledge

Recall the two dimensions or aspects of the first-order problem of knowledge discussed above: (1) one must be able to act on the basis of one's own personal knowledge or the local knowledge one has access to as a member of an association; (2) when so acting, one must somehow take into account the knowledge of others of which each person is hopelessly ignorant. The concept of consensual transfers addresses both dimensions of the first-order problem of knowledge. First, permitting consensual transfers of jurisdiction enables persons to act on the basis of their personal and local knowledge by authorizing them to exchange jurisdictions they currently have for jurisdictions they believe they can put to better use. In this way, a transfer of a person's jurisdiction reflects her local and personal knowledge.

*845 Second, requiring that all transfers of jurisdiction be by consent addresses the second dimension of this knowledge problem by enabling--indeed forcing [FN57]--persons to take into account the knowledge of others when making their decisions. For changes in boundaries to reflect the knowledge of all affected parties, such revisions must be based on the manifested consent of the individuals or associations whose boundaries are changed. By requiring consent, the new claimant is compelled to take the knowledge of the present "jurisdiction-holder" into account--including the present holder's knowledge of her own perceptions, preferences, opportunities, etc.

For example, if Ann, based on her knowledge of her situation, would prefer to maintain jurisdiction than see it transferred to Ben, then to obtain Ann's consent to a transfer, Ben must offer Ann something he thinks she would value more. In other words, the onus falls upon Ben to provide Ann with jurisdiction over some other resource that she could put to better use than the jurisdiction she currently holds. So, for example, Ben could offer Ann jurisdiction over a book she has yet to read in exchange for her jurisdiction over a book she has already read. Only if Ben must obtain Ann's consent is there any assurance that his claim to jurisdiction will take her knowledge into account.

But the requirement of consensual transfers affects our ability to take into account the knowledge of others far more profoundly than this simple "micro" example suggests. Such a requirement also makes possible the evolution of a powerful "macro" institution that enables personal and local knowledge to be "encoded" and transmitted worldwide in a form that can be easily understood by others and incorporated into their decisions without centralized direction. In short, the requirement of consent permits the evolution of a set of resource prices.

*846 Prices are by far the most neglected form of knowledge we have. [FN58] The reason for this is that the knowledge embedded in prices is not explicit; we are never conscious of it as knowledge. It is encoded knowledge, and we are conscious only of the code. Prices reflect the vast personal and local knowledge of the many competing uses to which any physical resource may be put. My computer is constructed of plastic, glass, various metals, and other resources. My desk is made of wood. These resources could have been used in a variety of other ways by people throughout the globe. I have not the slightest way of knowing even a small fraction of the specific alternative uses that others might find for these resources. And yet without a comprehensive knowledge of all the alternative uses of these resources, how can a knowledgeable decision be made on how these resources should be used?

I have already explained how, in light of the dispersed nature of personal and local knowledge, the problem of knowing alternative uses of resources is immense. It would require the compilation of all persons' personal knowledge of perceptions, interests, and opportunities and all local knowledge of associations as to their shared interests and opportunities, the integration of this knowledge into a coherent plan, and the communication of everyone's allocated role. This is a knowledge problem of such enormous proportions that less information is preferable to more. That is, even if we could have direct access to all the knowledge we require, the sheer volume of such knowledge would prevent us from putting it to use. We need somehow to condense this knowledge into a usable form. We need to convert it to a form of local knowledge that can itself be integrated into each person's personal knowledge. And this process of condensation need not be perfect to be superior to the only alternative: near-total ignorance that results from the general inaccessibility of personal and local knowledge. This vital function is performed by the device of resource prices.

Resource prices condense the personal and local knowledge of each one of us into a form of local knowledge that can be integrated into the personal knowledge of all of us. Resource prices are local knowledge insofar as they are communicated from one person to another in *847 an intelligible form. Once communicated, they may be integrated into the personal knowledge of individuals concerning their available opportunities. For example, a trip to Aix-en- Provence has a resource price attached to it. When I consider this choice, I must consider the subjective cost to me of paying this price. This cost is the most highly valued set of opportunities that I will forgo by choosing to go to Aix. [FN59] Less formally, I must consider what I will have to sacrifice to make the trip. Of course, even with a market price of zero, there is no such thing as a truly cost-free trip to Aix, because such a trip will require me to forgo other potential uses of my time. But the monetary price to travel to Aix will strongly influence the cost to me of such a trip. And the monetary price reflects the uses to which others may put the resources that it would take to get me to Aix. [FN60]

Prices are able to communicate this information, however, only because the consent of those with jurisdiction over particular resources is required before jurisdiction may be transferred to another. None of this calculation would have been performed had I not been required to obtain the airline's consent to fly me to France and had the airline not been required to obtain the consent of all those whose cooperation is needed to make the flight possible. The need of others to obtain the consent of a jurisdiction-holder means that anyone wishing to obtain a transfer of jurisdiction must offer the present jurisdiction-holder jurisdiction over other resources that the present holder believes he or she would put to better use. The types of offers, as well as the number of persons offering to make exchanges, educate the holder of the value that others place on the resources. When this *848 value reaches a certain level, the holder is induced to make an exchange, thereby revealing that the value she placed on the resource was less than the value to her of the resources offered. Without the requirement of consent, this information would never be revealed and meaningful prices could not arise.

With a set of resource prices, a person is able to--indeed must [FN61]-- decide whether to use a resource, save it for later use, or exchange it for another resource by comparing her knowledge of the different uses she has with the knowledge and preferences of countless others that are encoded in the market price for the good. If the market price is higher than the value she places on the resource then she will be induced to exchange it. If the market price is lower, she will either use the resource or conserve it for later use or exchange.

The process is dynamic in that the holder of jurisdiction is incorporating price signals--a form of local knowledge--into the personal knowledge on which she bases her decision. In turn, her decision (to hold or sell) will influence the price signals received by others and will then be incorporated into their personal knowledge. For example, my ongoing decision not to sell my house both influences the market price of housing and, simultaneously, is influenced by the market price of my house and by the market price of alternative housing. True, the effect of my decision alone is unlikely to "move the market," but, in the aggregate, the current market price is a product of everyone's decision either to sell or not to sell. This process of knowledge generation and transmittal could not occur if the rights to resources could be transferred without the consent of the rights-holder.

c. Summary

The concepts of several property and freedom of contract both help address the first-order problem of knowledge. By delegating discretion to make choices concerning the uses of resources, several property enables persons and associations to act on the basis of their personal and local knowledge without outside interference. Freedom to contract enables persons to exchange their rights on the basis of their knowledge that other rights would better serve their purposes. *849 It also enables them to make gifts of their rights on the basis of their knowledge that others could make better use of these rights. Freedom from contract protects the expectations of current rights-holders, permitting them to put their knowledge into effect over a period of time free from the interference of others.

Moreover, without adherence to the principle of freedom from contract, resource prices would not arise. In making our personal or local decisions about resource use, each person or association needs "input" or knowledge about the potential resource use of others relative to the supply of resources. Resource prices provide this knowledge in a usable form and in the only manner that such knowledge could ever be provided systemically. Such a knowledge- conveying mechanism would cease to exist without the requirement that one's jurisdiction over resources cannot be displaced without one's consent. Nonconsensual transfers of rights "short-circuit" the price system of knowledge transmittal and make it impossible for individuals and associations to take the knowledge of others into account when putting their own knowledge into action.

B. Two Problems of Interest [FN62]

the fact that I have so extensively discussed how consent addresses the first-order problem of knowledge might suggest that I view this as its only important social function. To the contrary, even if there were no knowledge problem, we would still face a serious social problem of interest. The ability of the liberal conception of justice, with its principle of freedom of contract, to handle this pervasive problem provides an independent and reinforcing reason for adhering to the requirement of consensual transfers. [FN63]

The problems of interest take many forms, but they all spring from the common tendency of persons to make judgments or choose actions that they believe will serve their subjective preferences (although these preferences may not always be self-regarding). This tendency is not, by itself, a problem. Acting out of interest can be considered a problem only against some normative background that *850 distinguishes objectionable from unobjectionable actions. In this Section, I consider two distinct problems of interest: the partiality problem and the incentive problem. [FN64] These problems of interest would require some solution even if we faced no problems of knowledge. Those who urge that contractual consent be abandoned or highly qualified must explain how these problems can be handled successfully in some other manner.

1. The Partiality Problem

The partiality problem arises from the fact that people tend to make judgments that are partial to their own interests or the interests of those who are close to them at the expense of others. The word "partial" reflects both the cause and consequence of this problem. One meaning of the term is "affecting only a part; not complete or total." [FN65] In this sense, it is inevitable that individuals can have only a partial or incomplete view of the facts that go into reaching any decision. It is very hard to avoid seeing the world from one's own particular, and therefore partial, vantage point. Like other interested action, the existence of partial judgment is not itself a problem. The term "partial" merely denotes an incomplete, rather than an incorrect point of view.

But this partiality or incompleteness of vision also leads to the other meaning of the term partial: "favoring one person, faction, etc. more than another; biased; prejudiced." [FN66] Partiality, in this sense, is judgment affected by interest. Once again, this is not in itself a problem. Just as most of our actions are motivated by interest, much of our judgment is to some degree partial towards our own interests and the interests of those whom we care about.

A partiality problem arises when persons whose viewpoints are influenced by their own interests are called upon to make judgments *851 that are supposed to take into account the interests of persons remote to them. This type of impartial or objective assessment is required when systems of resource allocation require some person or persons to make a general or society-wide determination of how resources are to be used. Yet it is simply very difficult for persons charged with making such a decision to set their own interests in proper perspective in order to make an impartial assessment. In sum, the partiality problem refers to the difficulty of making judgments concerning resource use that take into account all available personal and local knowledge without succumbing to the tendency of persons to give priority to their own knowledge and interests.

Even if, contrary to my thesis (but as many believe), persons with centralized jurisdiction over resources could gain sufficient access to the personal and local knowledge of others to address the knowledge problem, we would still need to confront the problem of partiality. Assuming that these persons have access to the local and personal knowledge of others, what assurance do we have that their decisions concerning resource use will be based impartially on this knowledge, rather than on a partial judgment of what is in their own interest? [FN67] Although several property plays an important role in addressing the partiality problem, [FN68] I shall confine myself here to explaining how this problem is addressed by the principles of freedom of contract.

*852 Even if a prospective rights-holder had reliable access to the knowledge of a current rights-holder, when seeking to obtain the rights held by another, we cannot expect the prospective rights-holder to be impartial in assessing whether he can make better use of the resources than the present rights-holder. By making consent of a rights-holder a necessary condition of rights transfers, the principle of freedom from contract forces prospective rights-holders to take the knowledge of current holders into account when deciding whether to effectuate a transfer.

The fact that property rights may not transfer without the consent of the current owner means that, to acquire the right to use these resources, any prospective owner is compelled to induce the current rights-holder to consent to a transfer. The amount and kind of this inducement reflects the personal and local knowledge of the current rights-holder as to how these resources may be used. By becoming part of the prospective owner's cost of obtaining control over the resource in this manner, the knowledge of the current rights-holder is brought to bear on the allocational decision. A prospective owner is compelled to take the current rights-holder's knowledge into account without requiring that the prospective owner have direct access to the personal or local knowledge of the current rights-holder. [FN69]

It is now apparent how the system of resource prices that arises from adhering to the principle of freedom from contract addresses not only the first-order problem of knowledge but also the partiality problem. The requirement that everyone pay the price for resources held by others does more than enable persons to take into account the knowledge of others. It also forces them to take the knowledge and interests of others into account, even when they would not otherwise find it in their interest to do so. In this manner, the concept of freedom of contract functions with the concept of several property to ameliorate the partiality problem.

*853 2. The Incentive Problem

Solving the first-order problem of knowledge by adhering to the concept of several property assumes that people will have sufficient interest to actually use their access to knowledge of how resources may be used. It assumes that the prospective benefits of acquiring knowledge and putting it to use will outweigh the subjective costs of such action. Suppose, however, that rights-holders lack adequate incentives to use their access to knowledge when deciding how to use their resources. This incentive problem concerns choices among actions that justice permits. It refers to the need to close the gap between the conduct that the decentralized allocation of resources permits, and what persons subjectively perceive to be in their interest to do. [FN70]

The incentive problem arises most graphically when the benefits of exercising knowledgeable control over resources do not accrue to the person or persons exercising such control. To appreciate the nature of the incentive problem, let us imagine a world of several property where control over resources was decentralized in much the same manner as in western countries. Those generally in the best position to have beneficial knowledge of resource use would be those who had legal control as well. In other words, the allocation of legal control to individuals and associations would closely reflect the distribution of personal and local knowledge.

Now imagine that all the benefits accruing from a knowledgeable exercise of control were routinely siphoned off and given to others--for example, via a steeply progressive income tax or a confiscatory wealth tax. The inability to reap the benefits from using one's knowledge to control resources would greatly reduce the incentive to exercise knowledgeable control in the future. Some incentive to act productively might still exist if exercising control were for some reason intrinsically rewarding or if one had a special affinity for the person receiving the benefits. As the inherent interest in doing a job declines or as the recipient of the benefits becomes increasingly removed, however, even this residual incentive to act knowledgeably would decline.

*854 This analysis suggests that, just as the distribution of control over resources should correspond to the distribution of knowledge in society, the distribution of benefits should closely correspond to the distribution of control. Although the concept of several property disperses the control of resources throughout society in a way that tends to match the distribution of knowledge, an important function of the two principles of freedom of contract is to address the incentive problem.

The principle of freedom from contract ensures that changes in control of resources reflect the knowledge of the original rights-holder. Only if the rights-holder consents to a transfer will it be recognized as valid. Consent will not be given unless the rights-holder subjectively values the distribution of rights resulting from the transfer more highly than the original distribution of rights. Without a requirement of consent, the incentive to use one's knowledge to improve the property within one's rightful jurisdiction would be undercut by the prospect that others could dispossess the rights- holder without his or her consent.

The principle of freedom to contract provides incentives for beneficial transactions by enforcing agreements motivated by the prospect of receiving a benefit or "profit." This prospect creates powerful incentives to investigate and discover previously unknown opportunities for beneficial transfers. [FN71] Entrepreneurship is the ability to identify previously unknown or neglected opportunities for beneficial transactions. [FN72] If contracts producing so-called "speculative" gains were unenforceable, then the incentive for such entrepreneurial activity would be eliminated.

Conversely, the prospect of incurring a "loss" induces a level of caution in persons' actions. One has an incentive to be careful about putting one's knowledge into action if one incurs the full cost of any mistake. Moreover, the only way to eliminate losses is to transfer *855 resources to the actor who has made the bad bargain from others who have not. In the absence of consent by the person to whom the loss is shifted--for example, a consent to a risk- pooling or insurance scheme--such a policy of coerced loss spreading will have adverse incentive effects on those from whom this compensation is confiscated.

Freedom of contract, then, both inhibits transfers adversely affecting interest and encourages beneficial transfers. The principle of freedom from contract--that is, no transfers without consent--ensures that rights transfers will not create negative incentives. The principle of freedom to contract--that is, consensual transfers are valid--makes entrepreneurship possible by ensuring that positive incentives exist for beneficial rights transfers. [FN73] In these ways, freedom of contract addresses the problems of interest as well as the first-order problem of knowledge.

C. Communicating Justice: The Second-Order Problem of Knowledge

In the functional analysis just presented, the consent that addresses the problems of knowledge and interest is subjective--what some view as the only "real" assent. [FN74] But reliance on subjective assent encounters the following difficulty: because every person's intentions are part of his or her personal knowledge, and because these intentions are inaccessible to observers, they provide an extremely problematic basis for determining rights transfers. Lacking any direct access to these intentions, recipients of transfers can never be entirely sure ex ante that subjective assent is present. Third parties seeking to adjudicate conflicting claims of rights also have difficulty making such an assessment ex post.

More generally, even if a consensus existed that adhering to the liberal conception of justice was the best way to address the first-order problem of knowledge, this strategy would fail if no one in the world *856 had access to its requirements. Without this knowledge no one's conduct could be influenced by the dictates of justice, an order of actions would not be achieved, and the first-order problem of knowledge would go unaddressed. Unless acting consistently with precepts of justice is entirely instinctive to human beings, we need a way to disseminate knowledge of justice in such a manner as to make its requirements accessible to everyone in a society.

The second-order problem of knowledge, then, is the need to communicate knowledge of justice in a manner that makes the actions it requires generally accessible to all. It is "second-order" because it arises only once the liberal conception of justice is used to address the first-order problem of knowledge.

1. The Second-Order Problem of Knowledge and the Rule of Law

The problem of communicating the requirements of justice is handled by the formal requirements of legality associated with the liberal conception of the rule of law. A well-known summary of these elements of legality was provided by Lon Fuller:

[T]he attempt to create and maintain a system of legal rules may miscarry in at least eight ways; there are in this enterprise, if you will, eight distinct routes to disaster. The first and most obvious lies in a failure to achieve rules at all, so that every issue must be decided on an ad hoc basis. The other routes are: (2) a failure to publicize, or at least to make available to the affected party, the rules he is expected to observe; (3) the abuse of retroactive legislation, which not only cannot itself guide action, but undercuts the integrity of rules prospective in effect, since it puts them under the threat of retrospective change; (4) a failure to make rules understandable; (5) the enactment of contradictory rules or (6) rules that require conduct beyond the powers of the affected party; (7) introducing such frequent changes in the rules that the subject cannot orient his actions by them; and, finally, (8) a failure of congruence between the rules as announce