The Limits of Liberty: Between Anarchy and Leviathan
By James M. Buchanan
Publisher
none
- Foreword
- Ch. 1, Commencement
- Ch. 2, The Bases for Freedom in Society
- Ch. 3, Postconstitutional Contract
- Ch. 4, Constitutional Contract
- Ch. 5, Continuing Contract and the Status Quo
- Ch. 6, The Paradox of Being Governed
- Ch. 7, Law as Public Capital
- Ch. 8, The Punishment Dilemma
- Ch. 9, The Threat of Leviathan
- Ch. 10, Beyond Pragmatism
- Selected Bibliography
Law as Public Capital
Man makes laws; in this respect he differs from other animals. He chooses deliberately to impose constraints on his own behavior; he distinguishes between rational planning and response to stimuli. Self-imposed law in an isolated individual setting is wholly distinct, however, from agreed-on law in a social setting, in an interaction with other persons. In the latter, the individual accepts defined constraints on his own behavior, not because his own well-being would be privately enhanced by such constraints but in exchange for favors that take the form of acceptance of like behavioral constraints by other parties in contract. That is to say, the individual does not enter into social contract for the purpose of imposing constraints on himself; he could always accomplish this by more effective means. He enters into agreements with others to secure the benefits of behavioral limitation on their part. The individual will find his own adherence to law unprofitable except insofar as this is directly related to others’ behavior. This constraint on his own liberty is the cost side of the contract. Rationality precepts, strictly interpreted, suggest efforts toward maximizing “law-abiding” by others and toward minimizing “law-abiding” by the party in question. Law is simply a reciprocal relationship among parties, an embodiment of contract. It is voluntaristic in a sense analogous to any contractual relationship; parties agree on the whole set of terms. This does not imply that unilateral adherence to these terms in the absence of effective enforcement is utility-maximizing. Each party has an incentive to violate the contract, to violate law, if he can predict that his own behavior will exert no influence on the behavior of others. A person has little private-personal incentive to adhere to the terms of a pure “external” contract, except insofar as this represents required “payment” for reciprocal action offered by contractual partners. Law-abiding by a single party, treated independently, exerts a pure external economy on other parties.
In the discussion contained in Chapters 2 through 4, the constraints on individual behavior embodied in law were conceived to emerge from basic constitutional contract. Until and unless some such constraints are in existence, individuals are not themselves “defined” with sufficient precision to allow postconstitutional trades or contracts to be implemented. As there noted, both the theory of private goods and the theory of public goods are normally based on the presumption that a well-defined set of individual rights and claims exists, as imbedded in a functioning legal structure. Although not often made explicit, there is also a presumption that there exists a clear demarcation between those interactions in which behavior is limited by formal law and those which continue to utilize anarchy as the organizing principle. It is this latter presumption which must be relaxed at this point. The formation of constitutional contract is continuous. “Law” is in a continual process of change and modification; communities are observed to be adopting new constraints on individual behavior, to be shifting additional areas of human activity from anarchy to law. At the same time, communities are observed, or should be observed, to be shifting yet other activities from law to anarchy. The effective constitutional status quo is dynamic. The trade-off between the individual’s liberty of person that is present only in the complete absence of law and the order that is present only with formalized legal constraints on behavior is subject to change as tastes, technology, and resources change.
The purpose of this chapter is first to discuss this trade-off with the concepts of public-goods theory, although the differences as well as the similarities between the two applications of analysis will become apparent. As a second cross-classification, the capital-goods characteristic of legal structure is examined.
Law and Public Goods
To the extent that law embodies the contractual origins discussed above, or that law may be conceptually explained on the “as if” presumption of such origins, law-abiding exerts a
pure external economy. This feature distinguishes law from the more orthodox public-goods interactions among persons. For comparison here, we may look at one of the classic examples, the provision of lighthouse services for a community of fishermen. In the absence of collective action, a single person may provide at least some of these services; if he does so, he exerts significant external economies on others in the group. But in the process he will also be securing some share of the total benefits. The external economy is not pure; others than the acting person receive less than the totality of benefits that the action produces. In this lighthouse-type case, the presence of significant external benefits allows us to predict that independent, individualistic behavior will result in suboptimal levels of service; too few resources will be invested in the activity under normal conditions. We cannot, however, predict that there will be no services made available in the absence of collective contract. Individuals may well invest some resources under certain cost configurations and certain community sizes.
With “law,” however, no such results emerge, regardless of group size. Precisely because law-abiding is a pure external economy, and as such involves behavior from which the actor secures no private, personal reward, an economic model would predict an absence of all such behavior in the strictly individualistic setting. (I am abstracting here from any moral or ethical precepts that may lead persons to act as if some contractual basis for reciprocally advantageous behavior exists. Abstraction from such considerations for the purpose of analysis is not, of course, equivalent to asserting that such precepts are nonexistent, or, if existent, that they are empirically unimportant.) In somewhat more technical language, “law” of the sort analyzed here qualifies as a pure collective-consumption or public good,
*89 and one for which independent adjustment yields corner solutions for each person. No person will provide, by his own restricted behavior, the benefits of law-abiding to others. I define “law-abiding” here to mean a generalized respect to the defined rights of all others in the community, as opposed to a particularized or directed result. Probabilistically, if a single person, A, chooses to respect the rights of
all others in the group of size
n, each person in the
n – 1 set becomes more secure in the possession of his rights.
A second possible difference between “law” and the more familiar lighthouse example of public good may be mentioned. In the latter, even if cost configurations suggest that no person independently invests resources in generating the external economy, cooperative, club-like arrangements may be made among two or more persons but substantially fewer than the whole community membership. Two or more fishermen may find joint provision of lighthouse services advantageous, and they may carry out such a venture even when they recognize that they will confer significant external benefits on remaining members of the community, benefits for which no payment can be exacted. Again “law” may be quite different. Privately arranged contracts that take the form of cooperative, club-like arrangements among relatively small numbers of persons may emerge, but those who remain outside the contract but within the larger community need not secure spillover benefits comparable to those received in the familiar public-goods examples. Indeed, the result may be just the opposite. Small groups may be motivated to form coalitions, but the law-abiding that is embodied in agreement may be strictly
internal to members of the coalition. The law that emerges may well be selective and discriminatory, with a sharp difference between the specified behavior of a person toward members of the coalition and that toward outsiders. The primal disarmament contract may be limited; predatory behavior vis-à-vis persons and groups that remain outside the boundaries of agreement may continue and perhaps accelerate. (This setting is, of course, descriptive of the world when we look on nation-states as the effective coalitions of persons.) Those persons who do not form a part of the internal agreement may secure no spillover benefits from the behavioral constraints that are accepted by members. For the ideal-type public good, exemplified by the lighthouse, technological attributes of the good itself make it impossible or inefficient to exclude noncontracting persons from benefits. The good, as produced, is necessarily nonexcludable. With law, conceived in this publicness context, that which is produced tends by its nature to be selective. Exclusion tends to be efficient in this case, and self-interest will insure its presence.
If, from an initial anarchistic equilibrium, a preliminary coalition forms, members of this coalition secure differential advantages over persons who remain outside the subgroup. The latter persons suffer costs because of the increased productivity of predatory efforts that may be directed against them by members of the coalition, who have by agreement ceased internal predation. This effect will be present even if there are no scale advantages in either defense or predation; if scale economies exist, the differentials will be larger. This situation becomes almost the converse of the more familiar free-rider problem when nonexcludability of public-goods benefits is present. Unorganized persons are confronted with a choice set that places greater incentives for (1) joining the initial coalition, or (2) forming another coalition. As some of the persons unorganized take either of these steps, those left still unorganized find their own situations further worsened. Their incentives to join existing coalitions or to form new ones increase further.
If several independently organized coalitions emerge, each of which embodies internal agreements on law, movement toward an all-inclusive legal contract for the whole community may take the form of bargaining among the separate coalitions. In either case, there should be potential gains to all parties from contractual arrangements that extend the legal structure to the appropriate limits of community membership. The existence of major discontinuities in the interaction sequence, such as those produced by defined spatial boundaries, may reduce the potential gains. The many complexities introduced in attempting to construct a plausible conjectural history for the emergence of a general legal system, with law being made coextensive with membership in the community, need not occupy too much discussion. The point of emphasis is the “publicness” of law itself, the attribute which allows us to use the tools of modern public-finance theory to shed some light on pressing social issues of our time.
*90
The Benefits and Costs of Law
Law-abiding on the part of an individual is the cost that he pays as a part of the overall legal-social contract between himself and others in the community, treated as a unit. In a private, personal utility sense, any limits on individual behavior are “bads.” But rational persons accept such limits in exchange or trade for the “goods” which law-abiding on the part of others represents. This behavior on the part of others creates “goods” because of the predictable order, security, or stability that it generates in the individual’s choice set. To the extent that other persons respect the limits laid down in law or rules, and to the extent that these are known by everyone, the individual can make his own private decisions in a reasonably predictable and stable social environment. This feature is applicable for many rules that might be called “laws,” whether these be simple rules of the road, such as right-hand driving, or complex arrangements, such as those required among different owners of condominium units. For purposes of illustrating the analysis, we may think of law as embodying general agreement to respect a set of rights or disposition over physical property. Individuals in the community, having agreed on or having accepted this assignment of rights, are said to abide by law if they do not attempt to secure others’ assigned rights without consent. It is not difficult to see that a legal structure confers benefits on all members of the community in most cases, benefits that stem from the order that is introduced. The appropriate comparison is with the anarchistic alternative which is, in the limit, characterized by a total absence of behavioral constraints.
These benefits are achievable only at the cost of limiting individual freedom, only at the sacrifice of personal utility which the spontaneous adjustment to changing circumstances would offer. To an economist, this benefit-cost framework suggests that there is some optimal or efficient amount of law for any person, some level of generalized behavioral restriction that is preferred to alternative levels.
*91 Each of us would probably agree that some traffic laws are beneficial, whether these specify right-hand driving, as in the United States, or left-hand driving, as in Britain. There is probably not a single person who values his own freedom of choice so highly that he would prefer a nation without traffic rules. On the other hand, for each of us there are probably some laws or rules in existence that yield benefits insufficient to offset the sacrifice of personal freedom involved. The Virginia Pollution Control Board, acting as empowered by the legislature, prohibits open brush- and leaf-burning, clearly a restriction on my own freedom of action, and one that is not offset in value to me by the prediction that other persons will also follow the regulation.
These examples point up a third difference between law and the standard public good, a difference that is of major importance. The achievement of “efficiency” in the overall quantity of law becomes extremely difficult, and, furthermore, the normative properties of any “efficiency” criterion become much less persuasive. This difference involves the adjustment in relative cost-shares among persons. In the standard public-goods model, all persons in the community share in the benefits of a uniform quantity of the nonexcludable good or service, despite their possibly differing marginal evaluations of this quantity. Separate persons may, however, be brought into agreement, or closer to agreement, on this common quantity by appropriately made adjustments in marginal cost-shares or tax-prices. (These tax or cost-share adjustments among individuals substitute for the more familiar adjustments in individually demanded quantities in markets for private or partitionable goods and services, where prices are uniform over separate buyers.) The aggregate cost of a public good or service, that which must be shared among beneficiaries in some fashion, is determined exogenously, and this cost is based, ultimately, on the alternative product values that might be produced by the resources used. This may be contrasted with the adoption of a general law or rule that does qualify under the “publicness” rubric and that constrains the behavior of all members of the community. As noted, the existence of and adherence to this law introduces greater stability into the behavioral environment for all persons and in this way qualifies as a purely public good. The costs of this rule or law are measured in the losses of utility suffered by each person due to the restrictions imposed on his range of choice options. These utility losses may indirectly reflect subjective estimates for objectively measurable opportunity losses. But the relevant consideration here is that these costs are not exogenous to the choice of the “good” itself. That is to say, the mere adoption of a law or rule carries with it a specific cost-sharing scheme; the law or rule must be generally applied to all citizens. Each person is subjected to a “liberty tax,” to use an appropriate term coined by Thomas Ireland,
*92 a tax that need not be, and normally will not be, valued uniformly by all persons, even if values should be reduced to commodity or money dimensions. It is as if, in the standard public-goods paradigm, some arbitrarily determined tax-sharing scheme should exist which is independent of the relative evaluations of individuals, a scheme that must remain inviolate.
Under these conditions, one dimension of adjustment, one degree of freedom is lost. In the provision of orthodox public goods, this dimension is used to produce more widespread agreement among persons on quantity and/or to accomplish specific by-product distributional objectives. Each of these uses of the flexible tax system in financing orthodox public goods may be briefly discussed for purposes of comparison here. If the utility losses inherent in the adoption of a law or rule, applied uniformly to all persons in the community, vary among individuals, and/or if the evaluations placed on the benefits vary, the preferred quantity of law will be different for different persons. (For analytical clarity we assume that the restrictiveness of law or rules is continuously variable rather than discrete. This allows us to treat the collective decision calculus in the familiar economic model. All-or-none choices can be introduced, but the analysis becomes more complex.) There is no way that the structure of cost-shares can be modified so as to produce more agreement among separate persons. From this it follows directly that, unless side payments or compensations separate from the liberty tax are somehow introduced,
any collectively chosen quantity of law will leave large numbers of persons in nonpreferred positions, either with overly restrictive law or with law that allows for too much liberty. Without compensations, there is no way that political entrepreneurs, even in some proximate sense, can move toward consensus, toward something that might approach Wicksellian unanimity. “Tax reform” is simply unavailable as an instrument of adjustment or compromise. To the extent that “tax adjustment” in real-world political structures is or may be successfully used to attain greater consensus on the level and mix of public spending on orthodox public goods, we must conclude that, by comparison, individuals will be somewhat less frustrated with budgetary outcomes than they will be with laws, rules, and regulations that directly affect their behavior.
*93 The individual loss-function aspects of the paradox discussed in Chapter 6 are necessarily present. Groups may divide sharply between those persons who consider existing law to be overly repressive and those persons who would sacrifice further behavioral options for greater social order and stability.
A variation on this sort of adjustment that may often be available in the provision and financing of standard public goods, represented by components in a governmental budget, involves logrolling among intense supporters of separate items. But this avenue of adjustment may also be closed with respect to collective decisions on laws or regulations of behavior. The direct or indirect exchange of votes on separate budgetary items allows individuals and groups to express intensity as well as the direction of preference. For example, even under a predetermined tax-sharing scheme, those groups that intensely desire some expansion on outlay in, say, space exploration may secure this objective, at least to some partial extent, by agreeing to support expanded outlay, on, say, higher education, which may be intensely desired by a different pressure group, one that is not particularly interested in space. Such logrolling is possible, however, only to the extent that the relative intensities of different groups are differently directed.
*94 With the imposition of laws or rules that restrict individual behavior directly, the scope for logrolling or vote trading may be very limited. Those persons who place relatively higher value on the maintenance and expansion of their own choice options are likely to be in agreement over most, if not all, of the proposed limits. The same applies to those who place relatively little value on freedom of choice. “Law and order” advocates are likely to desire more restrictive regulations on pornography
and on drugs.
We may now discuss the second use of the adjustment in tax-shares in the financing of orthodox public goods and demonstrate that this use, like the one of attaining agreement, is not available with respect to public restrictions on behavior. Even if no attempt is made to adjust taxes so as to secure more agreement on budgetary outlays, the possibility of manipulating relative tax-share offers a means of accomplishing by-product objectives through the fiscal structure. This adjustment is best discussed with the aid of a political model which assumes that a benevolent and omniscient despot makes decisions for the whole community. He can determine the quantity and mix of public goods to be provided, and we may assume that he chooses in accordance with efficiency criteria, derived from the individual evaluations of citizens, evaluations which the despot is assumed to be able to determine. This allows the financing or cost-sharing side of any fiscal decision to be divorced from the expenditure side of the budget. Through changes in relative cost-shares, a certain amount of income-wealth redistribution may be achieved, while maintaining overall efficiency in provision. In the real world, some elements of this political-decision model are always mixed with the alternative democratic-decision model which implicitly informs most of the analysis in this book.
Any observed public-goods provision and financing can be interpreted as embodying some mixture of efficiency and equity results.
*95 If, for example, some redistribution of real income and wealth from the relatively rich to the relatively poor is the acknowledged distributional objective, the variability in tax-shares, in total and at the margin, allows this to be achieved while retaining tolerable efficiency in the levels and mix of provision. But no such by-product redistribution can be accomplished by the enactment of laws or rules that are generally applied to all persons, even though these behavioral restrictions fully qualify as “public goods” under the standard definitions. The “liberty tax” structure is endogenous to the efficiency calculation, and it is not possible to vary the cost or tax-shares among persons. If the benevolent despot is fully informed about individual utility functions, he can determine the “efficient” level of behavioral restriction or regulation. He cannot, however, vary the relative cost-shares which this efficient level of regulation involves so as to promote a secondary purpose. Under normal configurations of preference functions, the efficient level of law may be unique, and this will, at the same time, require a unique distribution of costs among persons. There is a specific distributional result, but this result need not be, and normally will not be, related to plausibly acceptable equity criteria. Those persons who must, willy-nilly, bear the lion’s share of the costs generally are those who place the highest value on freedom and liberty, relative to order and stability in the sociobehavioral environment, particularly those who place the highest value on the actions that law inhibits. Those who bear relatively little costs are those who do not generally value freedom of choice and action highly, particularly those who place little value on the freedom to carry out the activities that the law prevents.
Both sides of the individual’s benefit-cost calculus must be included, and no generalized conclusions about distributional or efficiency results can be reached from cost-side or benefit-side comparisons in isolation. From benefit-side comparisons alone, it seems reasonable to suggest that persons in possession or command of relatively large quantities of private assets, the relatively rich, would place relatively higher values on the general application of behavioral restrictions. Those persons need not, however, optimally prefer more behavioral restrictiveness than those who command smaller asset endowments, provided that the restrictions are constitutionally required to be generally applicable to all members of the community. The utility losses suffered by the relatively rich in having their own choice options closed by law may also be highly valued in terms of a numeraire or common denominator. The relatively poor man may have relatively little to be protected by the behavioral limits to be placed on others, but neither does he need to place great value on his own freedom of choice and action. It is quite possible that the persons who command relatively little in private assets will optimally prefer more restrictive general laws than their more fortunately endowed compatriots. Consider auto theft as an example. A more restrictive law may take the form of intensified vehicle checks for registration and title. The poor central-city resident may consider the interferences with his personal freedom which such procedures involve to be a relatively small price to pay for the greater stability in possession that is promised.
*96 The rich suburbanite, by contrast, may think that the constraints on his own freedom which general application of the proposed law requires are not worth the increased order that is anticipated to be produced. This possible difference in attitude toward law may become even more pronounced when collective decisions concerning severity of punishments are considered, the general subject to be examined in Chapter 8.
Agreement on Constitutional Change
The earlier analysis (in Chapter 4) of the conceptual origins of the constitutional order demonstrated the general advantages of defined behavioral limits, and, through this, the possibility of securing unanimous agreement on rules or law. No attention was given to nonunanimous and imposed changes in constitutional structure, although departures from unanimity were discussed in connection with postconstitutional collective decisions on the provision and financing of public goods. These operational departures from unanimity were, however, themselves conceptually derived from general, and presumably, unanimous agreement in constitutional contract, which specifies rules for reaching operational collective decisions in addition to defining individual behavioral limits. In this chapter, we have introduced the public-goods framework to discuss behavioral limits, or laws, which emerge as part of the basic constitutional contract, and the analysis has indicated that individuals are quite likely to disagree on the extent of restrictiveness preferred and that means of compromising individual differences available in standard public-goods decisions are likely to be of little assistance.
The undiscussed question remains that of how changes in the set of rules defining behavioral limits, the changes in law, are themselves to be made, and on what criteria. Note that this is not the same question analyzed in Chapter 5, which involved agreement on changes in the distribution of individual rights, best conceived in property terms. For those aspects of the basic constitutional contract, the analysis suggested that agreed-on changes from the status-quo distribution of rights were possible, provided that the status quo is conceived in a dynamic rather than a static setting. The question that arises from the analysis in this chapter is quite different, since we are now concerned with a different aspect of basic constitutional contract, a different aspect of general legal structure, that which draws the line between those activities which are to be subjected to behavioral restrictions and those which are not. Changes here are not strictly distributional; they are organizational. The issue concerns individual comparison between anarchy and law; and, as suggested, individuals may differ. The problem of securing general agreement on behavioral limits in this setting was discussed in Chapter 4 in the particular geometrical variation of figure 4.1 where the origin lay outside the lozenge enclosed by the indifference contours through the position of anarchistic equilibrium. As suggested at that point, agreement on rights, or limits, in this case requires something beyond disarmament, something beyond mutual agreement to respect limits to behavior. Compensatory payments in commodities or in a numeraire good may be required to secure general willingness to accept new laws, new rules or restrictions on behavior, or to relax or repeal existing laws.
The analysis suggests that there will be major difficulties in securing agreement, but, conceptually, unanimous agreement on basic structural or constitutional changes remains possible, even when individual preferences for behavioral restrictions differ. And there is no other criterion than agreement with which we might evaluate the overall efficiency of such rules, in the absence of the omniscient despot who is so often conveniently invoked. As noted in Chapter 5, however, constitutional changes can be imposed without agreement. Political constitutions which are at all explicit normally require more inclusive rules for changes in the constitution than for ordinary collective decisions. In practical fact, basic legal rules are modified through long-observed but condoned departures from explicit rules, through judicial fiat, through legal precedent, through encroachment by the legislature on what should be the separate function of constitution-making, and through numerous other instruments. As stated earlier, there is no suggestion that imposed nonagreed changes in legal structure do not occur. The suggestion or implication to be drawn is only that such nonagreed changes as do occur cannot be logically derived from individual evaluations, and hence at this level they have little claim to be called “legitimate.”
Formal and Informal Law: The Role of Ethics
To this point the analysis has been based in the implicit assumption that formal, codified rules and regulations, requiring explicit collective implementation, make up the primary if not the sole content of “law.” Before we consider further causes of observed breakdown in law-abiding, it is essential to incorporate some treatment of the role that ethical precepts play in maintaining social stability. First of all, as noted in earlier chapters, if there is no conflict among separate persons, there is no basis for social contract; there is no need for law, as such. By the same token, however, there is no need for ethics; there is no function of a moral code. In the strict no-conflict setting, pure anarchy remains ideal without tempering. When conflict does emerge, however, anarchy in its pure form fails, and the value of order suggests either some social contract, some system of formal law,
or some generally accepted set of ethical-moral precepts. It is important to recognize that these are alternative means of securing order. To the extent that ethical precepts are widely shared, and influence individual behavior, there is less need for the more formal restrictiveness of legally imposed standards. And vice versa, although the superiority of securing tolerable behavioral order and predictability through ethical standards should be apparent. To the extent that the trade-off between narrowly defined self-interest and the putative general interest is internalized, and made to take place among the arguments within the individual’s own preference or utility function, the resort to external coercive force is minimized. Ordered anarchy, organized voluntarily through privately imposed constraints on behavior, through adherence to basic norms of mutual tolerance and mutual respect for acknowledged rights, is surely preferable even to an idealized formal constitutional structure that might generate a like degree of order, along with a comparable degree of efficiency. And, of course, such a relative valuation would be enhanced when it is recognized that any practicable legal structure must diverge sharply from an idealized one. Significant differentials in levels of predictable behavior may be required before formalized law would be chosen over anarchy, although individuals may vary in their trade-off positions just as in their evaluations of the degree of restrictiveness of law, once chosen. Both anarchy and formalized constitutional structure must be distinguished from a setting in which individuals behave strictly in accordance with customary or traditional modes of conduct, with little or no connection with rationally selected norms. This alternative is likely to be grossly inefficient, and it must be placed beyond the extreme limits of formalized legal structure in its coerciveness. Under such a regime, order is present in the predictability sense, but this order need bear no relationship to the “publicness” of the rules or customs that are being followed.
Fortunately, perhaps, communities do not face either-or questions with respect to basic organizing principles. The status quo rarely presents the stark alternatives: (1) anarchy with full dependence on internal ethical constraints as the only means of resolving conflicts outside war itself, or (2) inclusive and rigid formal laws with a complete absence of internal ethical constraints. At any point in time, behavior that may be observed in a community reflects individual utility maximization, and the pattern of behavior is influenced by the arguments in the preference functions of participants, along with the internal and subjective trade-offs among these arguments,
and by the formally imposed legal constraints on behavior that participants confront, constraints that are nominally embodied in formal rules accompanied by enforcement instruments,
and by constraints imposed by custom and tradition which, although nonformalized, remain external to the individual and carry their own enforcement processes.
As mentioned in Chapter 1, many aspects of social intercourse are organized anarchistically, which means that the observed orderly behavior depends critically on mutual acceptance of certain informal precepts by all parties. Life in society, as we know it, would probably be intolerable if formal rules should be required for each and every area where interpersonal conflict might arise. An indirect test of the cohesiveness of a society may be offered in the range of activities that are left open to informal rather than formal control.
The individual who restricts his own behavior voluntarily, who limits his own freedom of choice because of built-in ethical standards, who acts in accordance with something like a Kantian generalization principle, is continually confronted with a dilemma of sorts. His narrowly defined self-interest may dictate departures from that pattern of behavior required by adherence to his ethical-moral standards. He is in a position analogous to that faced by the potential free rider in public-goods theory, as discussed earlier in this chapter. So long as a significantly large proportion of the community’s total membership abides by the same standards, the temptation placed on any individual, although always present, may not be sufficiently great to cause him to modify his cooperative behavior. However, if and when some persons, or a critically large minority of persons, are observed to violate ethical precepts that previously have been accepted by almost everyone, and to act on self-interest grounds, those who might continue to adhere to the precepts find themselves subjected to what may seem to be exploitation. Once a critical limit is passed here, the standards may erode rapidly as more and more individuals revert to narrowly defined self-interest. The situation is in most respects similar to that in which individual behavior is restricted out of some precept of obedience to formal laws which carry with them no enforcement procedures.
This suggests that one of the most important ethical precepts may well be that of obedience to and respect for formalized law, as such. If individuals place a high value on obedience to law, as laid down through observed political-decision processes, utility-maximizing norms may produce surprising adherence even in the total absence of enforcement and punishment instruments. If individual preference functions embody this principle, it is the announcement and enactment of the rule or regulation that imposes limits to behavior that matters; enforcement and punishment institutions assume secondary importance. Casual empiricism suggests that this attitude may, in fact, explain much of the order that we observe, an order that exists even in those aspects of behavior when most persons recognize that enforcement of formal rules is nonexistent or woefully inadequate.
*97 As with the more general ethical precepts, however, this principle of respect for law, as such, may be subject to rapid erosion once a critical minority is observed to violate the principle. In such instance, unless effective enforcement and punishment are forthcoming, formal rules or laws may become inoperative as means of producing social stability.
*98
The Generation of “Public Bad”
The position of the individual under law, whether this be formal or informal, is comparable to that present in any “publicness” interaction so long as law itself qualifies under this rubric. In the absence of effective enforcement, external or internal, persons are always motivated to violate the standards laid down. This is true quite independently of a person’s preferences with respect to the appropriateness or the inappropriateness of the standards themselves, considered as rational collective institutions generally applied or as viable and widely shared ethical norms. Even the person who places the highest benefit-cost ratio, in total or at the margin, on the extension of behavioral constraints through law may be motivated, in his private, personal capacity, to violate these constraints. He is, as noted several times, in a position akin to that of the potential free rider with ordinary public goods. Economists have adduced the free-rider dilemma to explain the failure of voluntaristic, market-like institutions to supply jointly consumed goods efficiently. A more directly relevant application explains the necessity of coercion in the instruments of taxation. Individuals may not voluntarily pay taxes even if their private-personal benefits from public spending exceed their nominal tax liabilities. Consider the person who has explicitly been party to the putative public-goods contract in which his assigned share of tax is matched against expected public-goods benefits. Suppose that he succeeds in evading his assigned tax obligation; this has the effect of reducing the total revenues available for providing-purchasing the jointly consumed good, the benefits from which are shared by other members of the collectivity.
*99 In evading his tax obligation, which is economically rational for the individual, he creates a “public bad.” The person in question imposes an external diseconomy on all others in the sharing group, all potential beneficiaries of the jointly consumed good or service financed from tax revenues.
This is, of course, nothing more than the converse of the “public good” that is created by law-abiding. Failure to produce “public bad” is equivalent to the creation of “public good.” And the failure to provide “public good” is equivalent to the production of “public bad.” The choice between constructions here depends largely on the purpose to be served by analysis and on the relevance to real-world problems. If, as in traditional public-goods theory, the purpose is to explain why market institutions fail and why governmental action may be necessary, attention should be paid to the “public good” that collective action might generate. Much the same applies to explaining the requirement that law be established collectively. If, by contrast, the purpose is one of trying to explain why long-established institutions of “law and order” break down without effective enforcement, it is best to change the focus of analysis and to concentrate on individual behavior in generating “public bads,” despite the basic equivalence in the underlying models.
There are, of course, many important modern applications of the theory of public bads, notably those which are introduced in analyses of environmental quality. The treatment of law violation in this section is, in almost all respects, identical to that which could be, and has been, applied to explain pollution in basic behavioral terms. To pollute the air or the water or to despoil the natural environment is to create “public bad.” To violate established law, whether this be codified or present in prevailing ethical norms, is formally the same. The whole discussion might be subsumed under the general rubric of environmental quality if we are willing to recognize that the sociobehavioral environment is as important for the quality of personal life as the natural environment. The analysis becomes a theory of behavioral pollution.
*100
Why does an individual pollute? Why does the Los Angeles motorist add his bit to the already smog-laden atmosphere? Why does the family on a picnic dump its litter in the park? If the behavioral bases for pollution in such familiar cases are well understood, the extension to the less familiar terrain of law and order becomes straightforward. The individual pollutes, he creates public bad, because it is in his private, personal interest to do so. In creating public bad, the individual is creating or producing private good. It is through no malevolence or evil intent that the Los Angeles motorist adds to smog. He is not deliberately imposing external harm on others; his behavior produces this harm only as a by-product of his straightforward utility maximization, given the choices that confront him. The individual may recognize full well that there is a conflict between his behavior as a private decision-maker and that behavior which, if generalized to all persons, would produce results more desirable to him. But, in his private capacity through which he must act, there may be no means for the individual to influence the behavior of others, at least directly. Hence, it remains rational for the individual to do the best that he can under the circumstances. And since this is simultaneously true for all persons in the interaction, the aggregate result is pollution, deterioration in environmental quality, a result that may be desired by no one.
*101
The interaction need not reach what we might call full pollution equilibrium, in which each and every participant behaves strictly as directed by narrowly defined self-interest. Saints may continue to exist in every social group. And if ethical standards influence the behavior of some individuals and groups in the community, these may limit their own actions while others are allowed to create the public bads of pollution. If the two sets of actors are heterogeneous, and if the polluting groups remain within certain critical size limits, an equilibrium of sorts may be reached with widely divergent behavior patterns. Even here, however, the situation may be far from optimal, even to those who are the polluters. Despite the possibilities for such a quasi-equilibrium, once pollution on the part of some members of the social group becomes the observable and predictable response pattern, the forces at work tend to shift the system toward some full pollution equilibrium. This conclusion holds even if all parties recognize that they would have been better off had the erosion process never commenced. In the sequence of events, however, each party may have acted rationally, given the choice situations that were presented to him.
As of any moment in time, at any status quo, the sociobehavioral environment embodies some explicit adherence to ethical standards, some implicit obedience to informal rules stemming from custom and tradition, some obedience to formal law simply because it is law, some obedience to law that is due to effective enforcement and punishment expectations. These motivations may be mixed within the behavior pattern of a single person, and they may vary over persons in their relative weighting. From such a status quo, suppose that one person shifts his behavior pattern, and specifically that he departs from that behavior which would reflect acceptance of something like a Kantian generalization principle. He pollutes; he imposes an external diseconomy on all others in the community. By changing his behavior, the single person has modified the environment, he has changed the conditions of choice for others.
Consider a single example, that of auto theft. Suppose that one person who previously refrained from theft changes his behavior and becomes a thief. The precise object of his theft is, of course, under the ownership of a single party, and in this sense the external diseconomy is not general. But in making the behavioral change, which is presumably in his own private interest, the thief imposes diseconomy on all persons in society, over and above the directed harm to the owner of the automobile. Policing services must be increased if the same degree of order is to be maintained; these must be financed from general taxes. Private protection against theft must be increased, and this involves investment by all persons, and not only by those who have their property stolen. Insurance rates go up for everyone who owns an automobile. The predictability under which a person may own and operate an automobile is reduced. The quality of the sociobehavioral environment is reduced by the behavioral pollution that theft represents.
Although it is familiar in a sense, this example may be partially misleading because theft is normally forbidden explicitly in formal law, and enforcement and punishment institutions do exist. The behavioral pollution instanced here can occur only because of some failure of these institutions to accomplish their objectives. That things are not nearly so simple, even here, will become apparent in the discussion of Chapter 8. The pollution of the sociobehavioral environment may, however, be illustrated readily with other examples.
Consider the situation in the orderly anarchy that was the university community in the late 1950s. Although there may have been a few notable exceptions, most university communities were then characterized by relatively pure standards of free expression. Almost any student or faculty group could invite almost any speaker on almost any subject in the assurance that the event would be allowed to take place without disruption. The intellectual environment of the university embodied free expression, and expectations were made on the basis of this fact. In the 1960s, much was changed, and much more than has yet been realized. Certain individuals and groups, acting in accordance with their own privately dictated norms which may or may not have been based on some ultimate ethical values, chose deliberately to preselect speakers and topics of discussion, and to disrupt meetings by speakers and on topics that were beyond the limits. This sort of behavior cannot be generalized to all members of the university community without rapid degeneration into something akin to the pollution equilibrium previously noted. In the 1970s, the student or faculty group that considers extending an invitation to a visiting speaker must make some predictions about potential acceptability to dissident elements. Can anyone seriously dispute the statement that the quality of the intellectual environment was lower in 1970 than it was in 1960? And, once commenced, how can erosion be stopped? How can behavioral standards which allowed the university community to remain an ordered anarchy for so long be recovered once they are lost?
Legal Structure as Public Capital
The question directly suggests a critically important feature of law and law-abiding that has been left aside to this point. The structure of law, whether this be described empirically in formal or informal terms, represents social or public capital stock, the yield from which accrues through a sequence of time periods. The “public good” that law provides is analogous to the lighthouse as initially constructed; it is not analogous to the “public good” that is offered in the municipal fireworks display on July 4, something to be enjoyed jointly but only in the instantaneous time frame. This capital-goods feature is emphasized in the isolated Crusoe example. The very purpose of adopting laws or rules is to restrict behavior in future periods, restrictions that will, in turn, allow planning to incorporate more accurate predictions. The isolated person secures greater efficiency, he accomplishes more good for less bad, if he lays down rules for his own behavior in advance.
This element or feature does not change when the analysis is shifted to the many-person social setting, where laws and rules are designed to restrict behavior in the interest of mutual but not unilateral gains. It would, of course, be possible to analyze the possible emergence of behavioral rules or laws that apply only to a single time-period. Robin Hood and Little John may agree on a “law,” with no implications that this would apply on other subsequent crossings of the footbridge. It seems apparent, however, that the capital-stock characteristic of law has been implicit in almost all discussion and, indeed, the word “law” would scarcely be appropriate in a strict consumption-services setting.
This capital-goods characteristic is important both in the initial formation of law, in constitutional contract, and in the maintenance of existing law. To the extent that an initial agreement on law, or on changes in law, involves major bargaining and transactions costs, the benefits of agreement may be insufficient to warrant enactment of formalized rules at all if the applicability of these is anticipated to extend only over a short time-period. Many rules that might be acknowledged to be mutually beneficial over time might remain nonexistent if negotiation and agreement should be required anew at the onset of each specified point in time. This point is indirectly supported by widespread practice of resorting to generalized rules for interactions that are known to be short-lived but which require some means of ordering the proceedings. Robert’s Rules of Order for single meetings is perhaps the most familiar example. Groups acquiesce in these rules, not because they are necessarily the most efficient for the circumstances of the moment, transactions costs neglected, but because they are in existence and the inefficiencies may be less than the costs of negotiating a particular set of rules for the short-lived interaction.
The limited time-span of decision-makers is important in relation to the capital-goods feature of law. An individual, as he participates in the formation of or change in basic law which restricts his own behavior along with the behavior of others will make his own benefit-cost calculation for an expected life-cycle planning period. From this it follows that the individual’s own expected benefits from the establishment and maintenance of law may be significantly lower than the measure of present-value benefits to an idealized person who expects to live forever. From this it follows that the legal structure which might emerge from the choice behavior of mortal men might be somewhat less restrictive on behavior, somewhat less inclusive, than the structure which might be judged to be ideal by some omniscient being external to the community. The divergence will depend on the subjective discount rate in the utility functions of members of the community, and if these rates are sufficiently low the divergence may be insignificant. An important positive hypothesis can be deduced here; a shift in the subjective discount rate held by members of the community will modify the optimal or efficient levels of behavioral restrictions that they mutually impose on themselves and which are embodied in law.
The most important consequence of the capital-goods feature, however, involves maintenance of the capital stock through time. As we have discussed in earlier chapters, the political-legal structure, the existing “social contract,” need not be based on explicit choices made by those whose behavior it restricts. There need be present no consciousness of individuals having been party to the initial “investment” decisions that the existing constitutional structure describes. In this particular respect, there seems to be no basic difference between public and private capital. The person who inherits an embodied capital stock, measured in claims to private assets, need not feel himself to have been party to the investment decision, need not recall the sacrifice of consumption potential necessary for the initial capital formation. With a private capital stock, however, the legatee acquires full rights of disposition over both the capital asset and the income flow from the asset. He has a private incentive to maintain the source of the income flow that the capital asset represents. With a public capital, by contrast, the individual beneficiary has no such incentive, precisely because of the public-goods feature already discussed. By acting so as to maintain the public capital asset that the existing legal structure describes, the individual confers pure external economy on others in the current time-period as well as on those who live in subsequent time-periods. It may be privately rational for the individual to create “public bad,” to destroy the existing public capital, to convert this asset into privately enjoyed “income.” It is in this context that the term “erosion” properly applies. An individual’s decision may erode the basic structure, reducing the stability of social interaction not only for his fellows but for those who come later.
We may pursue the application of simple investment theory further. By converting capital into income, future incomes are reduced. To restore these to earlier levels, abstention from consumption is required. Consider a simplified numerical example. An existing private capital stock has a present value of one hundred units; it yields an annual income of ten units over and above full maintenance. Let us suppose that, during some period, the owner “eats up” or consumes ten units of stock, along with the yield, generating for himself a consumption flow of twenty units during that period. This action reduces the capital stock to ninety units, and the subsequent yield to nine units per period. For
all periods after this disinvestment, income must remain lower than before. If the owner decides to rebuild his stock, he must abstain from consumption by the full ten units, a feat that may prove difficult since his maximum consumption has already been cut back to nine units per period.
The numerical example is helpful, but it does not fully indicate the results of behavioral changes that might be interpreted as amounting to erosion in the capital value of an existing legal order. In the numerical example, capital yields income immediately on its investment, and at the same rate as that yielded on capital that has been invested for many periods. In application to law, however, it seems clear that benefits are yielded, in enhanced stability of interpersonal relationships, at an increasing rate over many periods of time. That is to say, the benefits from law increase in rate as the investment matures. It is as if, in the numerical example, investment would yield the full 10 percent return only if the asset is maintained for, say, ten years, and that this rate might, say, increase to 20 percent if the asset should be maintained for twenty years.
If the relationship between the yield on the public capital embodied in the legal-constitutional structure and time is such that a constant perpetual stream may be restored, once it has been destroyed, only over a period that exceeds personal planning horizons, the model moves close to one that involves “mining” rather than merely disinvestment. For all practical purposes, public or social capital may be permanently lost once it is destroyed. It may be impossible to secure its replacement, at least on the basis of rational decisions made by individuals.
If a diagnosis of society suggests that individuals, organized groups, and governments are creating “public bads” by departing increasingly from traditionally honored limits for behavior, by increasingly violating both formal and informal law, then a recognition of the capital or investment aspects of the genuine “public goods” that are being destroyed makes corrective action much more urgent than any application of a consumption-goods paradigm might suggest. (Perhaps it really does not matter too much if vandals disrupt the fireworks display; but if they bomb the lighthouse the crisis takes on more serious overtones.)
Law Reform and the Status Quo
Somewhat contradictory implications for reform or change in constitutional contract, for modification in law, emerge from the public-goods and the capital-goods paradigms. It will be useful to discuss these separately. When we examine legal structure in the publicness framework, with no reference to the capital investment feature, the relationship between efficiency and the underlying system parameters becomes apparent. The level of behavioral restrictiveness embodied in formal laws or rules that is optimally preferred by any individual, the level that might be deemed efficient by the omniscient and benevolent despot, the level that might emerge from a plausibly effective collective decision process—any of these will depend on the preferences of persons in the community, on the existing technology, and on the resources available to the community. An exogenous shift in any of these basic parameters will modify the behavioral restrictiveness described by the legal structure considered in its totality. We can best discuss this set of relationships in terms of the choice calculus of a single person as he tries to decide on his own most preferred quantity of general law, applied equally to himself and to others, and treated as a continuous variable. If we interpret ethical norms as affecting behavior through changes in the weighting of arguments in individual preference functions, the interdependence between formal law and preferences is obvious. Consider one individual, and we shall assume that his own preferences do not change. However, for other members of the community, treated as a unit, suppose that ethical norms lose some of their previously effective influence in constraining behavior in potential conflict situations. To the individual in question, the direction of effect on his most preferred quantity of formal law is readily predictable. He will optimally prefer a somewhat more restrictive legal system after the change than he did before the change in the preferences of his fellows. Similar relationships may be traced between technology and the optimally preferred level of law for the individual. These are familiar in modern discussions of the pollution of the physical environment. Until the advent of the internal combustion engine, no person may have rationally desired general laws restricting individual freedom to dispose materials in the air. In the 1970s, by contrast, he may find such generally applied restrictions desirable. The areas for potential conflict among persons, and hence those areas over which some trade-off must be made between unrestricted individual freedom of action and the behavioral limits of laws or rules, critically depend on the technology that is available and in use.
The effects of exogenous changes in the resource base are equally clear. In the most elementary economic sense, an increase in the resources available to the community reduces the potential for conflict. As resources become less scarce, as the economic problem becomes less acute, there should be less need for behavioral restrictions, for laws defining limits on individual rights. The primitive society cannot nearly so well afford the resource wastage involved in pure conflict as the modern, rich society, especially when engaging in certain sophisticated forms of conflict may itself provide utility. This elementary economic principle applies more or less directly to nonhuman resources. But what if the increase in resources takes the form of more human beings, relative to the nonhuman resources available? In this setting, the measure of nonhuman resources per person may well have declined while, overall, the community has become richer. The range of potential interpersonal conflict may have increased substantially with population increase, especially if space itself is treated as a nonaugmentable resource in the strict Ricardian sense. Increasing population concentration will tend to have the effect of increasing the optimally preferred level of formal restrictions on individual behavior. These parametric shifts may themselves be interdependent. An increase in numbers may, in turn, affect the influence of ethical norms on individual behavior. This generates a reinforced relationship between community size and optimally preferred restrictiveness.
If we can make the assumption that, for the individual whose calculus we are examining, the basic trade-off between freedom of choice and behavioral restrictiveness remains unchanged, a diagnosis of modern developments suggests that more rather than less formal law should be preferred in late twentieth century than in the late eighteenth or nineteenth century. Technological developments, increased mobility, and the increased density of population in space—these combine to suggest increasing interdependence and, what is the same thing, increasing areas of potential conflict. But, of course, we need not assume that the fundamental preferences of individuals for freedom and for order have themselves remained stable. Any diagnosis of modern society, especially since the 1960s, must also incorporate the observation that a fundamental shift in preferences has occurred at least for some individuals and groups, a shift toward individual freedom and away from constraints. Although its advocates may have exaggerated its extensiveness or its long-run importance, “the greening of America” is fact, at least for some members of society. The representative young person in the early 1970s does not value order so much as his counterpart in the early 1950s. The legal structure, along with still-influential ethical precepts, may seem repressive, which translates as embodying an excess of order relative to liberty of person. This person may stand willing to change the system in this direction even when he recognizes the opportunity costs as measured in disorder and behavioral unpredictability.
One of the most disturbing characteristics of modern society is the nongenerality of preference changes. If only a subset of the community’s total membership undergoes the preference shift described above, while others undergo no such change, the disappointment or frustration with legal-constitutional structure increases, regardless of which set of preferences dominates the policy choice. If the new-found or rediscovered preferences for individual liberty characterize the politically dominant majority, or the institutions which assume the role of changing basic law, those who suffer from increased disorder will be harmed. On the other hand, if those groups that seek to maintain traditional standards of conduct, expressed in formal and informal laws, exercise sociopolitical dominance, those whose preferences have undergone the libertarian shift will suffer increasing utility loss through time. And, as noted earlier in this chapter, means of compromising divergent interests may be ineffective or nonexistent.
More generally, however, there is nothing in the public-goods paradigm which implies that the legal-constitutional structure should be stable through time. Quite the opposite. The analysis suggests that this structure is a function of the parameters of the society and that changes in these parameters will, and should, change the basic compact. There is nothing which appears to be particularly advantageous about stability in legal order except insofar as individual preferences include some positive evaluation on stability, as such.
This conclusion is, however, contradicted or opposed by an implication that emerges from the complementary capital-goods paradigm applied to law. Here the advantages of maintaining existing legal-constitutional rules, along with ethical-moral rules, because they are existing, because they define the status quo, appear. Law is not only public; law is public
capital. Once this characteristic informs the analysis, proposals for reducing or for increasing the level of restrictiveness in society must surmount barriers that are not suggested in the complementary public-goods paradigm. This point may be demonstrated in an analogy to ordinary physical capital. Consider a person who inherits a capital asset, in some specific physical form, say, a building constructed in 1900. This asset yields an income stream of ten units per year, which, if capitalized, indicates the capital value of the asset to be one hundred. Let us make the further assumption that the owner cannot sell this asset in its current use. The scrap value of the building, which can be sold, is, say, twenty units. Aesthetically, the new owner may not like the architecture, and, economically, there may exist alternative opportunities for investment where one hundred units investment would yield 15 percent rather than 10. The building cannot, however, be converted into this most attractive alternative because of our assumption about its nonmarketability in its current form. Hence, despite the new owner’s dissatisfaction with the asset, both aesthetically and economically, it remains rational for him to keep it rather than scrap it. Destruction would remove the income stream of ten units per year, with only twenty units of present value recovered. Rationality dictates holding on to the physical form of the capital until such time as the return on the scrap value equals or exceeds the return on the old asset.
This analogy is quite helpful when we consider law and the calculus of an individual concerning his optimally preferred level of legal restrictiveness. An individual might prefer that the legal structure, along with the prevailing ethical standards in society, take somewhat different form from that which he observes in the status quo. But he inherits the legal order as it is, not something else. And he cannot “sell” this order to third parties. His choice set reduces to the alternatives that the analogy suggests. He can scrap the structure and start anew, but, in so doing, he must realize that a large share of the income flow from the asset will be destroyed, a flow that may not be fully restored until new investments mature over time. Or, he can hold on to the structure as it now exists, despite his recognition that alternative structures would be more desirable over the long term and that higher yields might have been forthcoming under different historical circumstances. Straightforward economic analysis suggests that large thresholds may exist as between these choice alternatives. This is, of course, recognized in the traditional legal doctrine of stare decisis, and more generally in the basic mystique of orthodox conservatism. The analysis supports these in a framework of rational economic decision-making. The erosion of constitutional-legal order should be recognized for what it is—the destruction of social capital, with all of the consequences therefrom.
This does not, of course, suggest that changes in constitutional order should never take place. The public capital paradigm suggests only that the shifts in basic system parameters must be sufficiently large to surmount the threshold that necessarily exists between “eating up” capital and consuming an income flow.
Is Law Dead? ed. E. V. Rostow (New York: Simon and Schuster, 1971), pp. 379-85.
My use of “law” in this respect is similar to that employed by Rousseau. See Rousseau,
The Social Contract, p. 399.
Alienation and the Soviet Economy (Albuquerque: University of New Mexico Press, 1973), Chapter 3; idem, “An Organizational Model of the Market,”
Public Choice 10 (Spring 1971): 81-92.
The Calculus of Consent, especially Chapter 10.
Econometrica 38 (November 1970): 907-20; M. McGuire and H. Aaron, “Efficiency and Equity in the Optimal Supply of a Public Good,”
Review of Economics and Statistics 51 (February 1969): 31-39. See also William H. Breit, “Income Redistribution and Efficiency Norms” (Paper presented at Urban Institute Conference on Income Redistribution, 1972, forthcoming in conference proceedings volume).
New York Review of Books, 19 April 1973.
Law, Legislation, and Liberty, vol. 1,
Rules and Order. Hayek cites, in elaboration of the specific hypothesis of “man as rule follower,” a book by R. S. Peters,
The Concept of Motivation (London, 1959).
Southern Economic Journal 32 (July 1966): 35-42.
Western Economic Journal 6 (December 1968): 347-58; and my “Public Goods and Public Bads,” in
Financing the Metropolis, ed. John P. Crecine, vol. 4, Urban Affairs Annual Review (Beverly Hills: Sage Publications, 1970), pp. 51-72. Also see James M. Buchanan and Marilyn Flowers, “An Analytical Setting for a Taxpayers’ Revolution,”
Western Economic Journal 7 (December 1969): 349-59.
Public Interest 25 (Fall 1971): 59-98. Also see his “Hockey Helmets, Concealed Weapons, and Daylight Saving,” Discussion Paper No. 9, Public Policy Program (John F. Kennedy School of Government, Harvard University, July 1972). For a discussion applied to ethical standards, see my “Ethical Rules, Expected Values, and Large Numbers.”