1. Marginal Notes on Reading Political Philosophy
by James M. Buchanan
Neither of the authors of this book is a full-fledged political scientist by disciplinary specialization and training. Moreover, even within the ranks of the acknowledged professionals, political theory and political philosophy constitute subdisciplines of substantial independence. It would, therefore, be presumptuous in the extreme for us to claim here that we have mastered even the accepted "classics" of political philosophy sufficiently to measure our own preliminary investigations and analysis against some wider criteria than our own subjective standards.
We are well aware, however, that the problems of social organization discussed in this book are among the most important that learned philosophers have debated throughout recorded history. Our work could, quite properly, be charged with serious omission if we should fail to include what must be, at best, relatively uninformed commentary on the classical treatment of some of these problems. Therefore, it appears useful in this Appendix to offer some marginal comments that have been prompted by a reading of some of the selected works in political philosophy. We hope that these notes will be helpful in relating our analysis to what has gone before and in pointing up the differences which, in our view, make the analysis contained in the main text of this book essentially unique.
Politics, Morals, and the Methodology of Political Science
"What ought to be" is the primary normative question. "What is" remains the basic positive one. The distinction has separated the moral philosopher on the one hand from the scientist on the other, but the dichotomy so achieved is too simple in relation to the problems that arise in political theory and philosophy. At the beginning of the text of this book, we stated that political philosophy has been concerned with what the State ought to be while political theory has been concerned with what the State is. Note that, even in such a purely introductory and nonrigorous statement, it was necessary to move beyond the simple form of the normative-positive dichotomy. A subject for the "ought" and "is," the State, was introduced; and this apparently slight change gives rise to a whole set of particularly difficult problems.
The State, or the polity, may be conceived as a set of rules or institutions through which individual human beings act collectively rather than individually or privately. That is to say, we may best describe what is normally called "the State" in terms that specify such rules and institutions. As we have previously emphasized, all attempts to make the State into more than this are regarded as falling entirely outside our frame of reference in this book. It seems unnecessary for us to compare our constructions with those of scholars who, at base, have adopted organic conceptions of collective life. A given set of rules describes a social organization, a political order. In discussing this order a useful, indeed an essential, line may be drawn between positive and normative theory. A positive science of politics should analyze the operation of an existing, or a postulated, set of rules for collective decision-making quite independently of the efficacy of this set in furthering or in promoting certain "social goals." A normative theory of politics should, by contrast, array the alternative sets of rules in accordance with their predicted efficiency in producing certain ends or goals which should be, if possible, made quite explicit. Normative theory must be erected upon and must draw its strength from the propositions of positive science, but it is only when this extension of normative theory is made that "reform" in existing institutions can be expected to emerge from specialized scholarship. Indeed the only purpose of science is its ultimate assistance in the development of normative propositions. We seek to learn how the world works in order to make it work "better," to "improve" things: this is as true for physical science as it is for social science.
Political science, normative or positive, is a science of human action; or, to adopt modern terminology, it is a behavioral science. The social order which is its subject matter consists, finally, in a network of human actions, human relationships. Moreover, individual behavior is not wholly predictable or predetermined, even when some allowance is made for stochastic variation. Individual human beings can make errors, and they can deliberately choose differently from the ways which, in fact, they do choose. Saying this, of course, commits us to a definite, but still debatable, philosophical position. However, the validity or the invalidity of presuming individual freedom of will is only indirectly relevant to the main point to be made here. This is that once individual behavior is introduced as a variable in the study of the social order under analysis, a second whole area of normative theorizing is opened. Moreover, as the whole history of political philosophy so amply demonstrates, it becomes very difficult to separate norms for the organizational structure—for the rules within which individual actions take place—from norms for regulating individual behavior itself.
The introduction of an analogy with the science of political economy may be useful in clarifying the distinction that is of central importance here. Here, as in politics, the study involves a social organization, the social order that relates the separate economic activities of individuals to each other. Here, also, a set of positive propositions may be derived, and, on the basis of these, normative propositions aimed at "improving" the working of the economy may be developed. However, students and scholars alike have continued to confuse this essentially appropriate normative theorizing with a second sort which relates to "improving" individual achievement in the operation of any specific economic setting. Many ill-informed scholars and students, especially those who work on the fringes of the discipline, conceive the study of economics to be aimed primarily at establishing norms for the earning of higher incomes by individuals and higher profits by business firms. The normative statements of economics are conceived to take the form of demonstrating to the individual what he should do (how he should behave) in order to further his own position in the economy vis-à-vis that of his fellows. Properly understood, this is not at all the subject matter of political economy: the latter is concerned with the norms for individual behavior only insofar as these norms determine individual action which, in turn, becomes data to the analysis of social organization.
Having generated considerable confusion in economics, where by comparison the distinction is relatively straightforward, it is not surprising that the fully analogous but far more subtle distinction has been blurred in political theory. To compare the study of business administration with that of political obligation may appear ridiculous at first glance, but a moment's reflection will reveal that methodologically the two are precisely analogous in their relation to economics on the one hand and to politics on the other. The science of politics, normative and positive, should be confined to the study of the political order. The positive aspects of this science should include the derivation of propositions that are conceptually refutable. The normative aspects should involve the construction of proposals aimed at securing "improvement" in the social organization (in the political order of affairs)—"improvement" being measured against some postulated set of goals derived, finally, from a fundamental ethical position. As with the science of economics, the behavior of the human actors in the process should be incorporated as data in the underlying positive analysis. There should be a sharp distinction made between the norms for ordering this individual behavior and those for improving or reforming the social order itself.
This basic distinction has never been made sufficiently clear. As a result the history of political theory-philosophy has been one of "politics and morals." Few modern theorists who discuss the underlying conceptual basis of polity have been able to free themselves of the compulsion to discuss political obligation. The obligation or the duty of the individual citizen to obey the law, to abide by the will of the majority, to act collectively in the "public" rather than in the "private" interest: these have occupied center stage in much of modern political philosophy. These are, of course, vital and significant issues, but it should be recognized that they raise questions of personal morality. As problems, they do not belong properly in political theory. Political obligations, as accepted by the average citizen and as revealed in his political behavior, become (or should become) data to the political theorist. The task of the theorist here does not include the derivation of normative propositions relating to these duties of citizenship or these responsibilities of rulers. Along with the economist and other social scientists, the political theorist should take his human actors as he finds them.
It should once more be emphasized that this proposed separation of politics from morals does not suggest that the political theorist remain purely positivist. There remain normative aspects of political theory, quite apart from morals. These aspects relate to proposed "improvements" in the political order, in the institutions of politics, and not to improvements in individual behavior. Problems of social organization need not be moral problems. The separation called for is not, however, so straightforward as the above paragraphs might have initially suggested. The methodological breakthrough is complicated by the fact that questions or issues of political obligation (of personal morality) arise in two distinct places. First, there is the question concerned with individual obedience to or acquiescence to the sovereign will, independent of the manner in which this will is itself determined. Secondly, there are those questions concerned with the moral precepts to be employed in determining this will (in the making of the law): that is, the obligation, duty, or responsibility of the prince, the bureaucrat, the cabinet minister, the legislator, or even the ordinary elector, to act in a certain way in his capacity as decision-maker, law-giver, for the collectivity. Both of these obligations—that of citizen and that of ruler—involve moral issues, and both require the introduction of norms for individual behavior. It is relatively easy to see that if the State is conceived at base to be nothing other than a continuing embodiment of the sovereign will, and, further, if it is assumed that this will is the only meaning of law, any "improvement" or "reform" can only come through some change in the behavior of individuals. Under this conception of political order it becomes impossible to separate politics from morals.
The point to be made can be most clearly illustrated with reference to the genuinely absolutist ruler. Under such a regime, making the State what it "ought to be" in terms of any postulated ethical standard reduces quite simply and quickly to making the prince "behave" differently from the way that he does, in fact, behave. In this model it is not possible to separate the moral choices of the prince from the institutional setting within which these choices are made and implemented. There would be little point in making any attempt at separation in any case since, by hypothesis, the institutional setting itself can only be modified by the action of the prince himself.
The divorce of politics, as a science, from political obligation, as a moral problem, can only be accomplished if the institutions through which collective decisions are made are themselves subject to variation (to change) only as a result of a second or "higher order" kind of collective decision-making process. Only if the "constitutional" decision, as we have called it, can be separated from operational collective decisions (that is, from those decisions that are taken within predefined constitutional rules) can political science emerge independently from the rather murky discussions of political obligation. The achievement of this independence seems to have been one of the essential logical purposes or aims of the contractarian approach to political philosophy, despite its obvious shortcomings and despite the confusion that has served to obscure this aspect in modern discussion of contract theory.
As the political order is shifted away from absolutism and toward democracy, the distinction called for here can be more clearly made. A genuine "science of politics" can be developed that is almost wholly independent of moral philosophy. This "social" science can include both positive and normative elements, but the variables with which it deals are social institutions, rules of the political game, not human motives. Insofar as this science becomes normative, ethical questions must remain, but these do not pertain to precepts for ordering individual behavior in acceding to or participating in collective decision-making. Within "political science," so limited, the scholar who proposes to answer the question "What ought the State to be?" must first make an explicit ethical choice. The information that he provides to the external observer then becomes as follows: "Given these ends for society, the set of rules describing the political order that would come closest to achieving these ends is as follows...." In this process the political scientist may specify the goals of social organization as broadly or as narrowly as he chooses. He may restrict himself to presenting his own personalized view of the "good" political society, always constructed from the behavior of real human beings rather than that of idealized "good" men. Or, by contrast, the political scientist may try, as best he can, to develop the normative implications of a set of ethical standards that he thinks should command widespread acceptance among all members of the social group. The point to be made is that, in either case, he must take men as they are, not as he would like them to be.
The normative aspects of the theory developed in this book are more restricted than either of those mentioned. We have tried to develop a "theory" of the political constitution. This theory is based on an analysis of specific rules for collective choice-making, given certain well-defined assumptions about human behavior in political action. On the basis of this analysis, we have then tried to answer the question: What set of rules should the fully rational individual, motivated primarily by his own self-interest, seek to achieve if he recognizes that the approval of such rules must embody mutual agreement among his fellows? Stated somewhat differently: What is the structure of the political constitution that will maximize "efficiency," in the broadest sense, for all individuals in the group, independently considered? As we have suggested earlier, the approach taken requires a minimum of ethical premises. We assume only that individuals are the relevant philosophical entities to be considered and that all individuals are to be considered equally capable of choosing. We have been concerned primarily with demonstrating the calculus through which constitutional decisions might be made, not with the precise configurations of the political institutions that might result from the calculus.
We have assumed that the individual whose calculus we have analyzed (the "representative" or the "average" individual) is motivated by self-interest, that his fellows in the constitutional decision are similarly motivated, and that, within the chosen set of rules for collective choice, individual participants are likewise directed. As we have suggested, this assumption about human motivation is perhaps the most controversial part of our analysis. It seems useful to repeat, in this methodological context, that, by making this assumption, we are not proposing the pursuit of self-interest as a norm for individual behavior in political process or for political obligation. The self-interest assumption, for our construction, serves an empirical function. As such, it may or may not be "realistic": this can only be determined by a comparison of some of the positive analytical implications with observable real-world facts.
From this rather elliptic discussion of the relation of political science to moral philosophy and the place of our own construction in this respect, we may now try to suggest some of those "classics" which seem congenial. It is perhaps clear that most of the so-called idealist theory-philosophy of political order is quite foreign to our approach. Writers in this tradition concern themselves more or less directly with questions of political duty or obligation. By our suggested classification these works belong in moral philosophy, and we should look to these, not for help in devising reforms in political institutions, but for guidelines of an individual ethic. It should not be surprising, therefore, that the most "sympathetic" or "congenial" works are to be found among the "realists" in the history of political doctrine. Initially we look to Glaucon in Plato's Republic, to Thomas Hobbes, and to Benedict Spinoza. Of these, and others within this tradition, only Spinoza's work seems to have much in common with our own, and only his seems deserving of special comment.
In his Tractatus Politicus, published posthumously in 1677, Spinoza approaches the whole study of political organization in a way that seems surprisingly modern by our standards. First of all, men are assumed to be motivated solely by considerations of interest. This is an underlying assumption of the models through which Spinoza examines alternative organizational arrangements. He states, quite specifically, that human behavior is taken as an empirical fact and that he makes no attempt to attach either praise or condemnation to the behavior that he observes. Spinoza examines the various political institutions in terms of their efficacy in producing results which he holds to be desirable. To him, political institutions are variables subject to change and perfection, and he conceives the primary task of the political scientist to be that of analyzing the workings of alternative organizational structures and of making such recommendations for change as seem indicated. His work on the political order anticipates, in many respects, that of David Hume and that of Adam Smith on the economic order. Spinoza deliberately sets out to construct political institutions in such a fashion that individuals acting in pursuit of their own interests will be led, by the institutional structure within which such action takes place, to further the interests of their fellow members in the political group.
The constitutional and the operational levels of collective decision-making are clearly separated in Spinoza's work. For the latter, at least in his aristocracy model (his discussion of democracy was not completed), simple majority rule is acknowledged as appropriate for reaching decisions in legislative assemblies. For changes in the constitution, in the basic laws, "common consent" or relative unanimity is suggested. Spinoza's work, in many respects therefore, may be taken as the most appropriately chosen classical precursor to that of this book. It should be stated, however, that Spinoza's influence on our own ideas has been limited to his general and indirect effects on the Western intellectual tradition. In a specific sense, we have carefully reviewed Spinoza only after the completion of an initial draft of the main body of this book.
Although Spinoza is often described as a follower of Hobbes, we do not find Hobbes' work at all similar to Spinoza's in relation to our own construction. As we have suggested above, it seems essential that some separation of the constitutional and the operational level of decision be made before politics, as a social science, can be satisfactorily divorced from moral philosophy. If sovereignty is conceived as being necessarily undivided and indivisible, this essential separation cannot be made readily. The contractual apparatus, to Hobbes, becomes an excuse or a justification for political obedience of the individual and little more. Hobbes' construction is aimed at establishing a basis for political obligation, whereas Spinoza's construction becomes a genuine theory of political order that, more than most others, is largely divorced from all issues of obligation.
At this point, as elsewhere in this Appendix, it is necessary to refer to the work of David Hume. As we shall suggest in the following section, Hume did discuss issues of political obligation, and he made notable advances over the contractarian theorists in this respect. Hume recognized quite clearly, however, that the question of the obligation of the individual to obey the law was conceptually distinct from those questions that arise when alternative political orders are considered. He specifically divorces "political science" from "moral philosophy": indeed the title of one of his essays is "That Politics May Be Reduced to a Science." In this essay he states that the purpose or aim of the checks and controls provided by the political constitution should be that of making it "the interest, even of bad men, to act for the public good."
Rational Choice of Restrictive Rules
As we have suggested, most of the important political philosophers have been concerned with the question of political obligation. In their discussions of this subject we may find points of departure that are helpful to an explanation of our work. John Locke and all of the writers who were responsible for developing the conception of "natural rights" made much of the distinction between the constitutional decision, which determines the rules for collective action, and the operational decision, which determines the shape of collective action within previously chosen rules. The individual, possessing certain inherent or natural rights, enters into a contractual relationship with his fellows, a relationship that is expressed in a constitution. The subsequent obligation of the individual to abide by the decisions made by the collectivity, so long as these are reached constitutionally, lies in his obligation to fulfill the contract once made. This basis of political obligation runs into immediate difficulty as soon as constitutional rules are made to apply to individuals other than those who might have been party to the original contract.
It is in this respect that the conceptions of David Hume appear most helpful, and they seem to have much in common with our own. Our basic analysis of the individual calculus that is involved in choosing among alternative organizational rules, in selecting a political constitution, has demonstrated that it will often be to the rational self-interest of the individual to select a particular rule that can be predicted to produce results on occasion that run counter to the self-interest of the individual calculated within a shorter time span. By shifting the choice backward from the stage of the specific collective decision to the stage of the constitutional decision, we have been able to incorporate the acquiescence of the individual to adverse collective action into a calculus that retains an economic dimension and that can still be analyzed in nonmoral terms. In this respect our immediate precursor is Hume, who quite successfully was able to ground political obligation, neither on moral principle nor on contract, but on self-interest. Hume did this by resorting to the idea that the self-interest of each individual in the community dictates the observance of conventional rules of conduct. These rules, which may or may not have been formalized in contract, are necessary for the orderly conduct of social affairs. This argument, which does not base political obligation on contractual obligation, allows the primary difficulty of the contract theorists to be neatly surmounted. Not only is it to the initial interest of parties to agree on conventional rules if such rules do not exist, but it is also to the continuing interest of individuals to abide by the conventional rules in existence. Hume recognized, of course, that, were it possible, the individual's own interest would best be served by the adhering to the conventional rules of all other persons but himself while remaining free to violate these rules. However, precisely because such rules are socially derived, they must apply generally. Hence each individual must recognize that, were he to be free to violate convention, others must be similarly free; and, as compared to this chaotic state of affairs, he will rationally choose to accept restrictions on his own behavior.
Individualism as an Analytical Method and as a System of Social Order
Many political philosophers, and especially those who have been concerned with the history of political doctrine, have not recognized the dual sense in which "individualism" may be employed as a descriptive noun identifying a theoretical-philosophical system. In the interest of clarity in discussion it seems useful to distinguish individualism as a method of analysis and individualism as a norm for organizing society. The fact that, in the development of political theory, those who have adopted the individualistic methodology have tended for the most part to adopt individualistic norms for social organization has served only to compound this particular confusion.
Individualism as an analytical method suggests simply that all theorizing, all analysis, is resolved finally into considerations faced by the individual person as decision-maker. Regardless of the role of the individual in the actual social-choice structure—whether he be ruler or ruled—analysis reduces to an examination of his choice problem and of his means or opportunities for solving this problem. To this approach is opposed that which starts from the presumption that some unit larger than the single person, some group of persons that includes two or more members, is the entity whose choice problems are to be examined. In this approach the individual member becomes an integral and inseparable part of the larger entity, and an independent-choice calculus for the separate parts is presumed meaningless. The individualistic method of analyzing political and social action is contrasted with the organic method, and these methodological differences need not, indeed should not, necessarily carry particular implications concerning the normative rules for organizing society.
One of the primary purposes of the contract theorists of political order seems to have been that of reducing the logic of collective organization to a logic of individual calculus, or, stated somewhat differently, of deriving a logic, an "idea of reason," as it was called by Kant, for collective organization from the individual-choice situation. These theorists asked the question: Can the existing organization of the State be "explained" as an outgrowth of a rational calculation made by individual human beings? In large part, the success or failure of the contract theorists should be assessed in these terms against their attempts to answer this question.
The individualist approach or method tends to obliterate any logical distinction or difference between the "public" and the "private" sectors of human activity. Collective action, along with private action, is motivated by individually conceived ends, and all action proceeds only after a mental calculus is performed by some individual or individuals. As decision-making or choosing bodies, individual human beings remain fundamentally invariant over the range of both private and public activity. All attempts of the political philosophers to distinguish sharply between "public right" and "private right" seem foreign to this approach.
Our theory of constitutional choice is avowedly individualistic in this analytical-methodological sense. Therefore, we react sympathetically to the works of those political theorists who have most clearly discussed the logic of collective organization in terms of an individual calculus and who have specifically rejected the conceptual demarcation between public and private sectors of human activity in the analysis of this choice problem. Johannes Althusius, who wrote very early in the seventeenth century, must be noted especially in this respect, for he seems to have been the first scholar who attempted to derive a logical basis for collective organization from contractual principles that were held to be applicable to all forms of human association. Later writers of the seventeenth and the eighteenth centuries, within the general contractarian tradition, followed Althusius on this point, although the emphasis on the common logical basis for public and private association tends to become less pronounced in their works than it is in Althusius'.
We find a similar emphasis in the writings of Christian Wolff in 1750. Wolff's work is also noteworthy because of his clear conception of the collective organization as a set of rules or institutions that are subject to analysis, to modification, and to reform. His method, like that of Spinoza, was that of examining alternative political institutions on which members of the community of rational individuals might agree jointly.
The individualist methodology found another staunch defender more than a century later in A. Fouillée, and his work is important for our purposes because he recognized, more clearly than most other writers, the distinction discussed in this section: that between individualism as a method of analysis and individualism as a norm for social organization. He recognized that there exists no logical inconsistency between individualism as a method of deriving principles of social organization and collectivism as a descriptive characteristic of this organization. As we have suggested, an individualistic approach is contrasted methodologically with an organic one. Either approach may be employed conceptually as a means of presenting either individualistic or collectivistic ideas for social reform. Given certain underlying assumptions about human-behavior patterns, along with a specific ethical position, a collectivist political-economic order may be rationalized from a calculus of individual choice. Fouillée understood this, and he argued correctly that there was nothing internally contradictory in Fichte's position which tended to be both individualist (contractarian) and socialist (collectivist). In our terminology Fichte's position could be described as methodologically individualistic, up to a point, and normatively collectivistic. Among political thinkers Burke comes perhaps closest to representing the reverse position. As regards alternative systems of social order, Burke was anticollectivist. On the other hand, methodologically he was clearly anti-individualist, and he vigorously rejected all attempts to explain collective activity on the basis of rational individual choice.
It is perhaps not surprising that proponents of methodological individualism are to be found among French political theorists as a part of the reaction against the excesses committed in the name of Rousseau's conception of the "general will." Somewhat later than Fouillée we find the work of Leon Duguit. He rejected categorically the conception of "national sovereignty" as the foundation for a system of public law, and he attempted to construct an alternative system on the basis of the public-service State. Duguit saw the State, not as an organ of command exerting power over its subjects, but instead as a means through which public services may be provided to individuals. These services were said to be required because of the fact of social interdependence. To this point Duguit's approach is similar to our own, which, at base, defines the political relationship in terms of co-operation. Duguit failed, however, to recognize that different individuals and different groups may desire different "public services" from the collectivity. To him, "public utilities" assume an objective character which, presumably, reasonable men can discover without great difficulty. He did not consider, therefore, the problem of the proper extension of public services. As a result his conception of the "public-service" State can easily be employed to provide a theoretical foundation for the growth of what is sometimes called the "welfare State."
Realism and Relevance in Contract
The contract theory of the State can be interpreted as representing both an attempt to divorce political theory from moral philosophy and as an attempt to derive a logic of collective action from an analysis of individual choice. Since our own efforts embody both of these elements, it follows that our work falls within the broadly defined limits of the contractarian tradition. It seems useful, therefore, to discuss some of the criticisms that have been advanced to this conception and to try to relate these to our analysis.
Both the contractarians and their critics have been too much concerned with the origins of government. The contractarians have discussed the original formation of government out of the voluntary consent of rational, previously "free" men. Their critics seem to have considered the contractarians demolished when they showed that such an original contract was, for all intents and purposes, a purely intellectual construction with little or no basis in reality. The relevance of the contract theory must lie, however, not in its explanation of the origin of government, but in its potential aid in perfecting existing institutions of government. Moreover, viewed in this light, some version of contractarian theory must be accepted in discussion about matters politic.
The origin of civil government and the major influences in its development may be almost wholly nonrational in the sense that explanation on a contractual basis is possible. Societies form governments and change governments for a variety of reasons, many of which remain mysterious and far below the level of objective, scientific analysis. Political institutions, like languages, get changed, almost beyond recognition, by the gradual and largely unconscious modification imposed on them by the movement through time. In this sense political society can be said to develop and to grow organically; and, if the purpose of investigation is solely that of explaining such growth, there is perhaps little purpose in inventing anything like the contractual apparatus.
It is clear, however, that the uncontrolled and the uncontrollable process of historical development is rarely called on to explain all changes in political society. If all change, from some origin to the present and beyond, is presumed to take place independently of conscious direction, political science, as a positive-normative discipline, loses its purpose. If, in fact, political institutions are not considered to be subject to rationally chosen modification and change, it is surely wasted effort to try to explain uncontrollable change. On the other hand, if it is accepted that political society is "perfectible," that political institutions are subject to designed "improvement," the analysis of alternative possible changes and the selection of criteria through which actual or potential changes may be judged become highly important tasks. At this level the explanations of the origin of civil government and the reasons for the major nonrational developments of this government are almost wholly irrelevant. Discussion must be concentrated on the "margins" of variation in political institutions, not on the "totality" of such institutions, and the relevant question becomes one of criteria through which the several possible marginal adjustments may be arrayed.
The contract theory, in this context, may be interpreted as providing one such criterion. Adopting the criterion implicit in the contract theory, the analysis of political institutions asks: On what changes in the existing set of rules defining the political order can all citizens agree? This embodiment of the unanimity rule for all basic, structural reforms in political institutions, in the constitution, reflects the individualistic ethic in its broadest sense. Other criteria for judging changes in the political constitution may, of course, be advanced. These may range from the purely personal criterion of the scholar who asks: What changes in the existing set of political rules do I think should be made?—to the more complex criterion introduced by the scholar who asks: What changes in the existing set of political rules would be "best" for the "greatest number" of individuals in the group, as I interpret their interests? Note, however, that such criteria as these, and any others that might be employed, must introduce a stronger ethical postulate than the individualistic criterion that the contract theory embodies.
In this interpretation the contract theory of the State in political theory occupies a position that is analogous to the Pareto rule for assessing changes in the more technical discipline of modern welfare economics. It may be useful to recall the discussion of Chapter 12 in the text. To define a position as Pareto-efficient or Pareto optimal does not suggest that all changes that have moved the group to that position were themselves Pareto optimal. On the other hand, to define a position as nonoptimal does suggest that there exists a means of moving to an optimal position in a Pareto-optimal manner. Applying this fully analogous reasoning to the contractarian terminology, we may say that the definition of an existing set of political rules (the constitution) as reflecting consensus implies only that there exist no particular changes on which all citizens can agree. Analogous to the Pareto-optimality surface, which contains an infinity of points, the fact that an existing set of political institutions reflects consensus, so defined, does not in any way imply that this set, and this set only, is the only "optimal" or "efficient" government. There must exist also an infinite number (conceptually) of other institutional arrangements which would similarly embody consensus. By contrast, the definition of an existing set of institutions as nonoptimal in the sense that it does not reflect consensus means strictly that changes are possible on which all members of the group may agree.
This interpretation of the contract theory, which divorces the existence of consensus from the means through which the existing situation has been produced, allows Hume's criticism of the contractarians to be fully accepted without seriously weakening the usefulness of the construction itself in its provision of a meaningful criterion against which changes in political constitutions may be judged. Having advanced this "marginalist" interpretation of the contract theory, we do not suggest that an explicit statement of this interpretation is to be found in the writings of the contractarians. To our knowledge they did not make the essential distinction between the "total" and the "marginal" explanation of political constitutions. Strictly interpreted, therefore, their "theory" of government cannot be accepted. However, when an attempt is made to advance an alternative "theory," one which will provide a useful criterion for evaluating constitutional change that is, in fact, controllable, some modified "marginalist" version of the contract approach seems essential. It is in this latter sense that the constructions of this book may be classified as falling within the contractarian tradition.
The Economic Approach to a Theory of Politics
As we have suggested in Chapter 5, the relatively recent work of William J. Baumol represents almost the only attempt to develop a theory of collective activity from the economic calculus of the individual citizen. Baumol's work, in one sense, developed the political implications of modern welfare economics, grounding the logic of State activity squarely and quite properly on the existence of external effects resulting from the private behavior of individuals. We believe that our work extends that of Baumol in two essential respects. First, as we have noted, the generalized-externality argument is applied to the constitutional problem, the choosing of decision-making rules. Secondly, the essentially economic approach embodied in the concentration on alternatives open for choice is more fully analyzed. The existence of external effects from private behavior has been shown to be neither a necessary nor a sufficient condition for collective action. A theory of collective action has been developed only after a careful consideration of the costs and the benefits expected to result from alternative organizational structures (alternative sets of rules).
In the literature of political theory-philosophy, a partial reading of which prompts this Appendix, it is not surprising that no narrowly conceived precursors to our work in this particular sense are to be found. The doctrinal developments in economics, on which our constructions are based, at least to some degree, have taken place during the period in which economics has existed independently of politics as a discipline. We do find, however, one rather neglected work in political theory that may be appropriately classified as being closely related to our own. It is again not surprising to discover that this work was written by one of the important figures of the Enlightenment and that it was completed during the last decade of the eighteenth century, although it was not published until a half century later. We refer to Wilhelm von Humboldt's Ideen zu einem Versuch die Gränzen der Wirksamkeit des Staats zu bestimmen. Humboldt argued that the only legitimate sphere of collective action was that which included the provision of security to the individual against external attack and against the encroachment of his rights by his fellow men. The role of the State was that of removing or reducing the external costs of private action. As might be expected, Humboldt conceived the externality problem too narrowly. He rejected all efforts of collective action toward promoting the positive welfare of individuals. In so doing, he failed to recognize that, in an opportunity cost sense, the failure to take co-operative action when such is actually more "efficient" is precisely equivalent to the taking of positive private action that is detrimental to over-all "efficiency."
It is in his careful discussion of the logic of State action in those cases of demonstrable externality, however, that Humboldt reveals clearly what was, at base, an economic approach. He recognized that the mere existence of spillover or external effects resulting from private action did not justify State action: the decision must rest on a comparison of the costs, in terms of the greater limitation on individual freedom, and the benefits, in terms of the greater security provided by some collective limitations placed on private behavior. He recognized that the function of theory in such cases cannot be that of laying down general rules; rather, this function must be that of "pointing out these moments of deliberation," that is to say, to outline the mental processes or calculus through which such decisions must be reached.
Humboldt seems almost alone in his very clear discussion of the voluntary arrangements that would tend to emerge to remove the external effects of private action—arrangements that we have discussed at some length in Chapter 5. He argued that, where possible, such arrangements are to be preferred to State action because of the unanimity that is implicit in all voluntary arrangements. At the outset of his work Humboldt criticized other thinkers for their excessive concentration on the question concerning who should govern and their insufficient attention to the question concerning the proper sphere of government. He recognized clearly that these questions were closely related and that the first question was rather empty until and unless the second one was resolved. This criticism seems to hold with almost equal force against most of the modern works in political theory.
The Classical Conception of Collective Choice
To our knowledge no political philosopher has approached the question of choosing among alternative decision-making rules in a manner that is similar to that which we have attempted to develop in this book. A partial explanation for this may lie in the very fact that those scholars who have been interested in political theory have been philosophers. As such, they have tended to think of collective decisions in terms of "will." If individuals differ in their desires for collective action, the decision-making rule must in some way determine whose "will" is to prevail. Individual or group interests, viewed in this way, tend to be treated as being mutually exclusive. Clearly the "will" of the majority and that of the minority cannot at the same time be prevalent. This whole approach to political process ignores or overlooks the possibility of quantifying individual or group interests. "Will" and "power" are terms that do not lend themselves readily to quantification.
By contrast to this "classical" approach, our approach is essentially economic, in that political decision-making is viewed, in the limit, as analogous to the determination of the terms of trade in an exchange. When individuals engage in trade, interests differ. Each individual desires to secure the most favorable terms of trade. However, no one draws from this the conclusion that the separate interests are mutually exclusive and that one must prevail over the other. Shall the "will" of the seller or the buyer prevail in a particular exchange? To the economist such a question is empty because "will" is meaningless unless specified more carefully. If it is defined as some maximum advantage from trade, the answer to the question must normally be that neither the "will" of the buyer or the seller prevails, although trade is observed to take place. On the other hand, if the term is defined as some improvement over an initial, before-trade position, the answer must be that both the "will" of the buyer and that of the seller prevail as a result of free exchange.
The point to be made here is that the very "vocabulary of politics" tends to focus attention too quickly on the particular problems presented under the existence of sharply defined and mutually exclusive alternatives. Choices at the ultimate constitutional level are interpreted as being of the "either-or" type. This is not to deny that such mutually exclusive choices do arise, and that when they do, decisions must be made. However, central to an economic approach to choice problems is the possibility of variation at the margin. If such variation is possible, choices become "either-or" only for small incremental changes; and, considering a total complex, some of all alternatives may be chosen. Interest in, or a desire for, a particular alternative becomes a function of its cost or price relative to the other alternatives available for choice. It is this functionally variable aspect that seems to have been almost wholly absent from the "classical" analysis of political decision-making. Practical politics has been traditionally recognized as consisting of the art of the possible—of the art of compromise. However, to our knowledge few political philosophers have recognized that once the necessity of compromise is acknowledged, alternatives are no longer considered as mutually exclusive, and the discussion that proceeds as if they were becomes largely irrelevant.