ABSTRACT: Herbert Spencer and F.A. Hayek developed bodies of liberal political thought that stress the importance of evolutionary social adaptation as a type of spontaneous order. An evolutionary social theory, properly understood, can form part of a liberal theory of politics. Improperly understood, it has been employed to form defenses of the modern nation-state, and nation-states are not products of spontaneous social evolution but rather are destructive of it. Hayek and Spencer both regard social evolution as a process that is progressive in nature, producing large, complex industrial societies. Their emphasis on the importance of our lack of knowledge needed to design social norms and of the destructive nature of efforts to do so generally supports the classical liberal values of limited government and personal freedom for which they argued. On the other hand, their evolutionary arguments to explain the emergence and persistence of freedom and limited government in mass societies are deficient, because the evolutionary arguments on which they rely depend upon social rules being established and sustained in smaller societies than the mass societies they imagine as the end of social evolution. A consistent liberal evolutionist, therefore, would not also defend the nation-state.
Keywords: evolution, spontaneous order, herbert spencer, friedrich hayek, classical liberalism
Dr. Boykin is Associate Professor of Political Science, Georgia Gwinnett College (firstname.lastname@example.org). The author thanks his colleagues Dr. Laura Bourland, Dr. Dovilė Budrytė, and Dr. Laura Young for comments on an early draft of the article.
F. A. Hayek once described himself as a “ghost from the nineteenth century” (Hayek 1982, 287). He does in fact form a bridge in the tradition of classical liberal thought between the nineteenth and twentieth centuries; in particular, his use of evolutionary theory to support political and economic liberalism revives a number of Herbert Spencer’s core ideas. This article compares the main views of Hayek and Herbert Spencer and argues that their evolutionary analyses should not include the nation-state.
Considering Spencer and Hayek together is significant for two reasons. First, as a matter of intellectual history, the likeness of Hayek’s and Spencer’s central ideas is remarkable. Hayek seldom refers to Spencer, and he never acknowledges any acquaintance with Spencer’s antirationalist social theory or psychology, both of which argue along lines quite similar to his own. Hayek’s contribution to political theory lies in his effort to build a system of liberal thought on the ideas of spontaneous order and social evolution, and his writings have played an important role in recent developments in classical liberal thought. Like Hayek, Spencer mounted a systematic defense of classical liberalism on the same fundamental ideas and was an intellectual leader among the liberals of his day. Given the critical notice devoted to Hayek over the last two decades, it is unfortunate that interest in Spencer’s work faded so quickly after his death in 1904.
Second, the measure of agreement between Hayek and Spencer represents the attributes of a distinctive political argument. Spencer’s and Hayek’s defense of classical liberalism rests ultimately on their claim that once evolution gives rise to liberal societies, it becomes irrational to turn back. This is so, they argue, because our inevitable lack of knowledge of complex social phenomena in developed societies and our reliance on tacit or inarticulate knowledge makes market intervention and social planning, which they interpret as a return to the institutions of preliberal societies, impracticable and destructive. This line of thought became more pronounced in Hayek’s later work, but much the same type of argument was clearly spelled out in Spencer’s writings of the nineteenth century. The side-by-side comparison of Hayek and Spencer offered here brings the key elements of this defense of classical liberalism into sharp focus and provides a perspective from which the theoretical possibilities and implications of this type of liberal thought can be fleshed out.
Finally, this article builds to an analytical point of current interest: a coherent liberal evolutionism should not include the nation-state. In broad outline, Spencer’s and Hayek’s contrast of societies established and maintained by consciously designed order and those that emerged through an evolved order of norms has much to offer libertarian political theory. On the other hand, their evolutionary arguments to explain the emergence and persistence of freedom and limited government in mass societies are deficient, because the evolutionary arguments on which they rely depend upon social rules being established and sustained in societies smaller than the mass societies they imagine as the end of social evolution. In fact, we should expect evolved social rules to decay and disappear in mass societies, creating a vacuum that leads to the growing political power that Spencer and Hayek warn against. A consistent liberal evolutionist would not defend the nation-state. A coherent liberal evolutionism does not regard social and cultural development as having a historical direction such as greater group size, and it is the state, not evolution, that establishes the ever-larger societies that Spencer and Hayek envision through force, not voluntary cooperation. A well-developed and properly circumscribed theory of social evolution can help illuminate the emergence of institutions that form the framework of a free society, including private property, markets, and such intermediate social structures as the nuclear family. An evolutionary social theory, properly understood, can form part of a libertarian political theory. It should not be employed to form defenses of the modern nation-states, which are not products of spontaneous social evolution but rather are destructive of it because they create societies too large for evolved cultural rules to persist.
The first two sections discuss and compare Spencer’s and Hayek’s theories of social evolution. The final section shows why the evolutionary processes they describe do not operate successfully in mass societies and thus that a liberal evolutionism should not include the nation-state.
I. SPENCER ON SOCIAL EVOLUTION
Spencer is popularly misconceived as a crude “social Darwinist” who regarded social evolution as a brutal contest among individuals (see, e.g., Hoftstadter 1955, 41–46). His views were not that simple. He argued that social evolution tends toward the development of larger, more complex societies that produce greater freedom and want satisfaction for their members. For Spencer, social evolution produces moral sentiments that support principled behavior based on ideas of just conduct and altruism.
Social evolution proceeds in part by means of group competition and in part by means of individual competition. Group selection involves, in the most visible way, armed conflict among societies in which some displace or subsume others. More generally, social or “superorganic” evolution proceeds by group selection as the cultural institutions and practices of a group that serve group survival functions tend to persist and enable that group’s practices and institutions to displace those which are less effective at promoting group survival in other groups. This is a key locus of social evolution, and it is an aspect of Spencer’s thought that is similar to Hayek’s. Social evolution for Spencer also involves competition among individuals, though his view of individual competition places greater emphasis on individual adaptation than the “social Darwinism” with which he is popularly associated (Taylor 2007, 52–56; Carneiro and Perrin 2002, 233).1 This last dimension of social evolution occurs within a much broader process of social change operating at the level of social groups. A key feature of Spencer’s thought that is of particular importance for the argument in part III below is that he views social evolution as including violent conquest to establish states and as progressive in producing larger societies. He regards regressive change to “militancy” and socialism as a product of bad policy rather than as a by-product of mass societies.
Spencer’s theory of social evolution is grounded initially in his psychology, which sought to harmonize empiricism and a subjectivism derivative of Kant’s epistemology via evolutionary theory. The human mind develops through experience of the world, but our experience is a translation of the environment through our nervous systems, which we have inherited and which develop over our lifetimes in response to our interaction with our environment. We learn through trial and error in our experiences, and learning modifies the connections in the nervous system that classify the phenomena we encounter (Spencer 1896, 1:330–75, 468–96). Individuals’ minds grow more complex through learning and adaptation, and many of the connections of the nervous system are passed on to subsequent generations by means of Lamarckian transmission of acquired characteristics (1:439–67). As a result, the minds of human beings have become more complex over the course of generations.
Spencer describes social evolution in terms of variation and natural selection among both individuals and groups. Human action is goal directed, and “adjusted actions are preceded by unadjusted ones” (Spencer  1982, 1:50). Behavior patterns that bring a greater degree of goal fulfillment are naturally selected over those that are less efficient. Through a testing process of trial and error, both in the individual and the group, we move to “adjusted actions” from “unadjusted ones.” The “evolution of conduct” involves competition among individuals: “[A] successful adjustment by one creature involves an unsuccessful adjustment by another creature, either of the same kind or of a different kind” (1:51–52). Spencer refers to this, however, as “imperfectly evolved conduct”; he argues that the emergence of moral constraints on individual competition increases the likelihood that more individuals in a group will realize their goals (1:59–79). Group survival also supports the emergence of altruistic behavior (1:234–38). The emergence and development of moral sentiments tempers individual competition and promotes group survival:
[C]onduct restrained within the required limits, calling out no antagonistic passions, favors harmonious cooperation, profits the group, and, by implication, profits the average of its individuals. Consequently, there results, other things equal, a tendency for groups formed of members having this adaptation of nature, to survive and spread. (2:43)
Social evolution, which involves both individual competition and group selection, tends toward the development of moral sentiments and social institutions and practices that promote optimal well-being and make “the totality of life greater” (2:52–53).
Spencer thus argues that social evolution has efficiency properties that tend to increase the quantity of want satisfaction in a society. He draws his conclusions from the emergence of animal behavior that has the results he describes, an argument to be found in the work of contemporary biologists analyzing the emergence of cooperative and altruistic behavior among animals (Smith 1978, 1982; Smith and Price 1973, 15–18; Dawkins 1989, 166–88). He uses this argument to explain the emergence of morality among human beings, equating “highly evolved conduct” with “what is called good conduct” among human beings (Spencer  1982, 1:78–79).
The morality and culture of societies evolve as “changing systems of ethics, proper to changing ratios between warlike activities and peaceful activities” under changing circumstances (Spencer  1982, 1:170; Spencer 1901, 1:442). Morals and culture evolve through an unplanned evolutionary process as groups whose practices support larger populations and produce greater want satisfaction for their members emerge and spread. As this process cannot be planned, it is a species of spontaneous social order.
Evolution and Spontaneous Social Order
Societies are analogous to individual organisms in that “both consist of mutually dependent parts. In both cases, the assumption of unlike activities by the component members is possible only on the condition that they severally benefit in due degrees by one another’s activities” (Spencer  1982, 1:175–76). Although some authors have argued that he held an organic view of societies (Offer 2010, 196–222; Paul 1988, 269–70; Paul 1983, 621), Spencer states that his reason for analogizing organisms and societies is to facilitate the depiction of evolved social order, and by no means to argue that the two types of order are the same.2 Taylor (1992, 132) suggests that Spencer’s comparison of society and organism is intended to engender “the view that society was not an object of conscious human design” and that this limits the ability of government to rationally direct social processes (see also Gray 1985, 246–53; Simon 1960, 294–99; Elwick 2003, 35–72). For Spencer, social evolution is the emergence and development of spontaneous social order, and the ability to control spontaneous social processes in a beneficial manner is highly limited. The complexity of societies greatly limits the ability to intervene rationally on behalf of some particular goal, because the unintended consequences of such intervention may be destructive: “[T]his spontaneously-formed social organization is so bound together that you cannot act on one part without acting more or less on all parts” (Spencer  1982, 392).
Spencer argues that evolving social orders tend to grow larger, and that as they grow larger, they also grow more complex. A social order grows more complex, because as it grows larger there emerges a “progressive differentiation of structures” which “is accompanied by progressive differentiation of functions” (Spencer  1901, 1:450). As a society grows more complex, it also grows more abstract. This occurs as individuals come to rely more and more on the emerging complex relations of the social order; it is those relations which become the more permanent elements of the order, as the individuals who compose it change over time (Spencer  1901, 1:452–60; Spencer  1982, 392). Spencer identifies two ways in which a society may become larger and more complex: either a discrete social group may proliferate, or different groups may be joined together. Spencer refers to the latter as a “compound” group. A larger social order clearly requires a greater productive capacity to support it. This is made possible by structural and functional differentiation. As a society grows larger, its productive and political order becomes more complex, both in terms of structure and function (Spencer  1901, 1:463–89; Spencer  1971). It is this greater complexity that permits the support of a larger population: “But along with advance of organization, every part, more limited in its office, performs its office better; the means of exchanging benefits becomes greater; each aids all, and all aid each with increasing efficiency; and the total activity we call life, individual or national, augments” (Spencer  1901, 1:489). Social evolution thus tends toward a larger and more complex social order.
Spencer classifies societies into “militant” and “industrial” types. One way that groups grow is through conquest of one group by another or through the formation of military alliances for common defense or imperialism (Spencer  1901, 1:519–22). In such a society, military preparedness is likely to be one of the group’s main purposes. Spencer argues that if the group’s main purpose is military readiness, that purpose entails the concentration of power, which extends over the society’s productive activities as well as its relations with other groups (1:523–47). Spencer describes the “militant” society as one in which “compulsory co-operation” (coercion as opposed to voluntary cooperation) predominates (1:564) (emphasis in original).
Industrial societies evolve from militant societies, because the origin of large societies is initially in conquest or defense-oriented military alliances (Spencer  1901, 1:565). Industrial society, as a sociological type, is a construct rather than an empirically observable social order, because real societies only approximate it in some respects. In the industrial society, production is carried on by exchange rather than by government direction, and individuals in it have rights to property. Disagreements are resolved by bargaining or impartial adjudication rather than by arbitrary decisions of the state. Persons in such a society view their actions toward others as being constrained by rules and view their government as being limited in its authority over them ( 1:564–69). “Voluntary co-operation” predominates in an industrial society (1:569, emphasis in original). An industrial society may revert to the militant form through foreign threat or imperialism, and Spencer saw this as happening in the Europe of his time (Spencer  1901, 1:579–86). Spencer also saw the emergence of socialism as another means by which this transformation might occur. Socialism requires massive intervention and direction by the state in a society’s economic life, which requires the concentration of power (Spencer  1982, 498–518).
Evolution and Ethics
For Spencer, the good associated with social evolution does not rest upon the fallacious notion that whatever evolves is good in a moral sense simply because it has evolved. G. E. Moore famously attributed the naturalistic fallacy to evolutionary ethics, but he observed correctly that “[t]he view, which, as I have said, seems to be Mr. Spencer’s main view, may also be held without fallacy. It may be held that the more evolved, though not itself the better, is a criterion, because a concomitant, of the better” (Moore 1903, 54). In Spencer’s theory, social evolution inclines toward greater degrees of complexity and, in consequence, greater productivity. In industrial society, cooperative rather than coerced interaction is possible. Cooperation requires that individuals keep agreements. If they do this, they will be better off, both individually and collectively, because voluntary social relations permit the greater complexity of social relations which makes a society more productive. Voluntary social relations are also nonviolent, so that no one is actively harmed by the actions of others (Spencer  1982, 1:170–78). Nonviolent, cooperative social relations are good, because they make those who live in such an environment better off.
For beyond so behaving that each achieves his ends without preventing others from achieving their ends, the members of a society may give mutual help in the achievement of ends. And if, either indirectly by industrial cooperation, or directly by volunteered aid, fellow citizens can make easier for one another the adjustment of acts to ends, then their conduct assumes a still higher phase of evolution; since whatever facilitates the making of adjustments by each, increases the totality of the adjustments made, and serves to render the lives of all more complete ( 1:53–54).
Spencer’s argument for the good of social evolution is utilitarian. Social evolution inclines toward a larger, more complex social order that generates greater amounts of want satisfaction due to the greater productivity of the society. Also, more people may experience life itself, since the evolving social order grows larger. Life is good, Spencer argues, because life makes pleasure possible, and “the good is universally the pleasurable” (1:66).
Although Spencer’s ethics are utilitarian, and he argues that acts are good if they generate utility, he is critical of the “empirical” utilitarianism that advocates choosing particular acts or rules based upon their expected social utility. Spencer argues that interpersonal comparisons of utility are too likely to be wrong to be useful, and that
Making general happiness the immediate object of pursuit, implies numerous and complicated instrumentalities officered by thousands of unseen and unlike persons, and working upon millions of other persons unseen and unlike. Even the few factors in this immense aggregate of appliances which are known, are very imperfectly known; and the great mass of them are unknown. (Spencer  1982, 1:187)
He argues that individual rights are the means to generate the greatest amount of utility in a social environment about which we have limited knowledge. The reason individuals should be treated as bearers of rights is that these rights establish an industrial society, which can support a greater number of persons enjoying higher levels of well-being (1:236–38).
As products of social evolution, Spencerian moral rights arise in particular societies as customary or conventional rules, even though Spencer, when engaging in moral philosophy, defends them as consistent with his law of equal freedom and on utilitarian grounds (Spencer  1982, 141–58; Weinstein 1998, 72–74, 156–62). Rights are “conditions to the achievement of happiness,” and “our rational course is to bring existing intelligence to bear on these products of past intelligence, with the expectation that it will verify the substance of them while possibly correcting the form” (Spencer  1982, 1:199). Spencer understood rights to liberty and the protection of private property as means to social cooperation and adaptation, “[s]o that utility, not as empirically estimated but as rationally determined, enjoins this maintenance of individual rights; and, by implication, negatives any course which traverses them” (Spencer  1982, 163–64).3 Spencer, then, may be interpreted as an indirect utilitarian in that his utilitarianism relies, not on calculations of the utility of particular rules or acts, but instead on the efficiency of spontaneous ordering processes, including social evolution (Gray 1986, 107).4 He calls his indirect utilitarianism “rational utilitarianism,” meaning by the qualifier “rational” that it takes account of the constraints on our knowledge of individual actions and preferences (Spencer  1982, 2:260).
Spencer contends that social evolution tends toward an ideal society that he calls the “social state.” The most general rules of the social state, he says, can be rationally discovered and used to evaluate real societies (Spencer  1982, 1:303–05). The “social state” combines maximum individual freedom with maximum utility for each member of the society (Spencer,  1970, 62). The “social state” is based solely on voluntary agreement and is a stateless society. The stateless society in which maximum freedom is combined with maximum well-being is for Spencer the end toward which social evolution is oriented.
Spencer’s treatment of the “social state” changed over time. In his first book, Social Statics, Spencer argues that “government is essentially immoral” (Spencer,  1970, 186) because its coercive basis is so at odds with the voluntary nature of the “social state.” In the 1877 second preface to Social Statics, Spencer presents the disclaimer that in writing the book he had not sufficiently recognized “the transitional nature of all political institutions, and the consequent relative goodness of some arrangements which have no claims to absolute goodness” (Spencer,  1970, xi). He goes on to attribute his newfound relativism to the anthropological information he acquired in preparing The Principles of Sociology. In his last major work, The Principles of Ethics, Spencer places a good deal of weight on his distinction between “absolute and relative ethics.” The “adaptation of humanity to the social state” (Spencer  1982, 1:303) has not yet been completed, so humanity is not ready for its great freedom. In the meantime people will have to submit to the state and its commands because they are still not good enough to be free (Spencer  1982, 1:287–308). Spencer’s acceptance of “relative ethics,” of things as they are, follows from his claim that social evolution is moving in the direction of the “social state.” For Spencer our destiny is perfect liberty, but we must be willing to wait for some unspecified period of time to claim that liberty. His view leads to a conservative quietism that has been noted by others (Taylor 1992, 167–75; Burrow 1966, 184–87).
II. HAYEK ON SOCIAL EVOLUTION AND ETHICS
Hayek’s theory of social evolution is similar in several ways to Spencer’s. Like Spencer, he relies on an evolutionary psychology and a theory of group selection and individual feedback to explain social evolution. Further, he distinguishes taxis, or societies in which planned order predominates, from cosmos, or societies in which unplanned or spontaneous order predominates. Like Spencer, he regards the growth of government in the advanced industrial states as a threat to spontaneous order and to personal freedom, and to social evolution as a type of spontaneous order. Key features of his thought that are shared with Spencer (discussed in part III below) are his view that social evolution is progressive in generating larger and growing societies and that socialism, as the transformation of cosmos into taxis, is a result of bad ideas and policies rather than the mass societies that he sees as the products of social evolution.
Hayek’s psychology is important to understanding his social theory, which places great emphasis on the limited powers of human reason. In a manner surprisingly similar to Spencer’s, Hayek’s psychology is a kind of Kantian subjectivism with an evolutionary twist. Mental phenomena are “representations of the external environment” (Hayek 1952, 121). Mental experiences are distinguished and related to one another by the classification of stimuli through connections between neurons, and mental experiences derive their attributes from the “following” through which they travel in the nervous system (Hayek 1952, 61–64). Objective differences in the physical world are translated into subjective differences through the physiological following established by classification (Hayek 1952, 119). The classificatory structure evolves as the “system of connexions is acquired in the course of the development of the species and the individual by a kind of ‘experience’ or ‘learning’” (Hayek 1952, 53). The system contains perceptual and behavioral patterns that are tested through action, and the results “select and confirm those which are useful as adaptations to typical characteristics of its environment” (Hayek 1979, 42). It is probably impossible to determine how much of this system is inherited, and how much is the product of individual experience (Hayek 1952, 102). The structure becomes more complex through experience as new classes form and new relations among classes are established. The emerging complexity adds to the cognitive “map” that enables the individual to interact successfully with the environment (Hayek 1952, 109–12).
Hayek maintains that the mind’s mechanism for adaptive learning places insuperable limitations on the human capacity for knowledge. The central nervous system is a hierarchical structure in which the formation of new connections on lower levels is governed by those already present on higher ones, so the system of classification must contain “a part of our knowledge which, although it is the result of experience, cannot be controlled by experience” (Hayek 1952, 169). These include “supra-conscious” connections that guide all mental activity and of which the individual is never aware (Hayek 1967, 61; 1979, 45). Thought operates through connections that are beyond awareness and cannot be altered because they are the basis of the classificatory scheme that makes thought possible; evolutionary psychology thus reveals “the limited powers of our reason” (Hayek 1973, 33).
It is because of our lack of knowledge, Hayek argues, that we rely upon rules. For Hayek, all human action is “rule-guided.” We need not “know” a rule, in the sense that we are able to explain it, but we may nevertheless “know” the rule in the sense that we can follow it. We can be said to know a rule, furthermore, in that we can recognize whether the conduct of others conforms to the rule (Hayek 1967, 43–45; 1960, 22–25). We can therefore learn to obey such rules without having them explained to us because we can learn them by imitating the conduct of those around us (Hayek 1973, 17–18; 1988, 21–23). Knowledge of rules of conduct can contain tacit elements, so it is clear why Hayek maintains that such knowledge is acquired by imitation rather than explicit instruction by another: knowledge that is not articulate cannot be imparted through explicit instruction. Our ability to use tacit knowledge enables us to use more total knowledge, since more information may be embodied in that tacit knowledge than in what we can know explicitly.
Rules of conduct serve to adjust our behavior to an environment about which we have imperfect knowledge. Some rules will be better attuned to the environment than others, and the less adapted ones are displaced through a process of natural selection:
[I]n social evolution, the decisive factor is not the selection of the physical or inheritable properties of the individuals but the selection by imitation of successful institutions and habits. Though this operates also through the success of individuals and groups, what emerges is not an inheritable attribute of individuals, but ideas and skills—in short, the whole cultural inheritance which is passed on by learning and imitation. (Hayek 1960, 59)
Hayek’s conception of social evolution thus incorporates the transmission of acquired characteristics—here through imitation—and the Darwinian theory of natural selection (Hayek 1979, 155–58; 1988, 23–28). The process operates as groups following different rules pursue their ends, testing their rules against the environment. Those rules that are better adapted to prevailing conditions will make those who act according to them more productive than those who follow less adapted rules. The rules followed by the more productive group will become the ones most widely followed, as the more productive group will grow larger due to its productivity, and as others seek to emulate the success they observe in them (Hayek 1979, 80; 1988, 70, 122–27).
Even after certain rules emerge and become dominant in an area, those who follow them need not understand the process through which they developed: “The group may have persisted only because its members have developed and transmitted ways of doing things which made the group as a whole more effective than others; but the reason why certain things are done in certain ways no member of the group needs to know” (Hayek 1973, 80). Thus “[c]ulture is neither natural nor artificial, neither genetically transmitted nor rationally designed” (Hayek 1979, 155). This is a clear difference with Spencer, for whom cultural adaptations were genetically transmitted. Hayek maintains that we should “view morals, not as innate instincts but as learnt traditions” (1988, 155), and that “[w]hat has made men good is neither nature nor reason but tradition” (1979, 160). He nonetheless recognizes, like Spencer, that organized violence may play a role in group selection because “the displacement of one group by another, and one set of practices by another, has often been bloody” (Hayek 1988, 121).
Hayek distinguishes orders that are intentionally created, such as organizations, from orders that are emergent or evolved, which are instances of “spontaneous order” (Hayek 1973, 36–38). An intentionally created order must be structured in such a way that its creator or creators can be aware of the activities of all its parts, but a spontaneous order, not the product of intentional design, can be so complex that no one is aware of the activities of all its parts or even of all the rules that guide those activities (Hayek 1967, 66–72). A spontaneous order may also be “abstract.” In a spontaneous market order, for example, the individuals involved may not know most of the relationships that exist among them (Hayek 1973, 38). Spontaneous orders are “purpose independent.” Only orders that have been intentionally created can be said to have a purpose. So far as the spontaneous order is concerned, the aims of its components are irrelevant. The components’ “rule-guided” activity sustains the order, but the order itself is a by-product of the pursuit of the aims of the individuals within it (Hayek 1976, 1–6; 1988, 75–83).
Although spontaneous order is useful because it allows for order beyond the capabilities of design, people have less control over it than over intentionally created orders such as organizations. They cannot deliberately determine what will happen to particular components of a spontaneous order without tampering with and disrupting the order (Hayek 1967, 23–24; 1973, 41–42). Social evolution, as a type of spontaneous order, permits a society to use information efficiently in the emergence of the social rules (Hayek 1960, 56–67). Hayek regards social evolution as being itself a spontaneous process, referring to “our traditional, spontaneously evolved morals” (Hayek 1988, 134). Spontaneous order, as in the market, efficiently transmits information about the present and expected future, while social evolution transmits information about what rules have been tried and naturally selected over others in the past.
Hayek contrasts spontaneous order with planned order, and he uses the terms cosmos and taxis to refer to each, respectively. A planned order, or taxis, is governed by rules that are consciously designed to serve a specific purpose, and as a result there are limits on the kinds of things it can do. A designed order can achieve only a limited level of complexity, because the activities of each individual in the organization must be known to and regulated by some other individual. Consequently, a made order can make use of only a limited amount of information in coordinating the activities of its members. Since it is designed for some specific purpose, the number of ends that can be pursued by it are limited, and the goals of each member of the order are subordinate to the overall goals of the organization (Hayek 1952, 141–52; 1978, 77–106; 1988, 75–88).
Cosmos, or spontaneous order, on the other hand, is abstract, complex, and purpose independent. The emergence and reproduction of an order of social norms makes it possible for individuals to pursue ends that they have chosen, because the order is produced and preserved by behavior which everyone can to some extent predict. Everyone can plan for themselves based on their justified expectations about the behavior of others. Expectations about others’ conduct can be justified insofar as a society is governed by an effective order of social norms, one function of which is to lend stability and predictability to society, which enables individuals to make plans based in part on what they know about others around them. The purpose-independent nature of a spontaneous order means that it can become complex in that it can include an unlimited number of individual ends and plans that can be based on an unlimited amount of information. As far as a society’s general social order is concerned, cosmos is superior to taxis because of the amount of information that can be utilized in it. The greater the amount of usable information in a society, the greater the amount of want satisfaction can be generated by that society (Hayek 1960, 22–27; 1976, 107–32).
The state has a limited role in spontaneous social order. Its function is to “provide an effective external framework within which self-generating orders can form” (Hayek 1979, 140). Hayek’s criteria for the rule of law, that laws be general, abstract rules that are equally applied (Hayek 1960, 149–54), is intended to limit government’s ability to interfere with spontaneous social processes. Hayek has been interpreted as holding a Burkean conception of government as being itself the result of a kind of spontaneous evolution (Buchanan 1975, 183n13; Brennan and Buchanan 1985, 9–10). The limited basis for such a view is found in Hayek’s support for the common law, through which he seeks to show how the body of a society’s law may be viewed as being, in part, a product of evolution. In fact, however, Hayek’s own work on constitutional design, which includes an “ideal constitution,” makes clear that he views theory as helpful to understanding how consciously-designed governmental structures may facilitate the emergence of spontaneous order and cultural evolution (Hayek 1979).
Hayek’s theory of cultural evolution might be described in part as the transformation of taxis into cosmos. Hayek notes that primitive groups resemble taxis in that the rules of such societies are oriented toward the achievement of a particular goal and might strictly regiment the members: “In its primitive form the little band did indeed possess what is still attractive to so many people: a unitary purpose, or a common hierarchy of ends, and a deliberate sharing of means according to a common view of individual merits” (Hayek 1978, 59). The emergence of private property and market exchange in some groups made these groups more productive, and these groups superseded those following older rules. As institutions supporting commerce grew, taxis was displaced by cosmos (Hayek 1988, 29–47). The driving force behind evolution in Hayek’s account is rules that conduce to the production of wealth, and, in turn, the size of the population that the group can sustain.
For Hayek, socialism is the epitome of taxis, in which the state is responsible for planning production and distribution. It is not only central planning but calls for “distributive justice” that pose a grave threat to spontaneous order and individual freedom. If government is to achieve a pattern of distribution, it must intervene in persons’ plans and lives to obtain the desired result. Government direction of the economic life of societies is destructive of personal freedom and spontaneous social order (Hayek 1976, 80–84). Further, the redistribution of wealth to promote material equality runs counter to the individual feedback requirements of social evolution, since the natural selection of practices presupposes differential success, so “the imposition of egalitarianism must stop further evolution” (Hayek 1979, 172). As Spencer saw in socialism the reemergence of the imposed order characteristic of militant societies in late nineteenth-century Europe, so also Hayek saw the threatened destruction of spontaneous social order and its replacement by the planned order of taxis as the end result of government growth throughout the twentieth century.
Social Evolution and Ethics
Hayek is an indirect utilitarian (Gray 1986, 59–60; Hardin 1988, 14, 78), as was Spencer. He argues that utilitarianism that calls for the deliberate choice of actions or rules on the basis of their purported utility-generating properties is a form of “constructivism” and has the same problem as others, namely that it requires knowledge which cannot be collected by anyone (Hayek 1976, 115–18). His argument is directed against “rule” and “act” versions of utilitarianism. Rules operate as part of a system of rules, and therefore generate utility only as parts of a system of rules. For that reason it is a mistake to evaluate a discrete rule on the basis of its utilitarian qualities (Hayek 1976, 18–20). In a large and complex society it is not practically possible to acquire the information needed to assess the outcomes resulting from the use of a specific rule: “Each person has his own peculiar order for ranking the ends that he pursues. These individual rankings can be known to few, if any, others, and are hardly known fully even by the person himself” (Hayek 1988, 95). Hayek’s critique of utilitarianism may be summarized as follows: “The trouble with the whole utilitarian approach is that, as a theory professing to account for a phenomenon which consists of a body of rules, it completely eliminates the factor which makes rules necessary, namely our ignorance” (Hayek 1976, 20).
Hayek nonetheless defends social evolution on utilitarian grounds. Social evolution, according to Hayek, is a process through which the rules of a society adapt to accommodate increasing population size and complexity of social relationships. Changes in rules that expand the scope of the division of labor and market exchange bring about a more efficient use of information concerning the employment of productive resources. As a result of increased productivity, the society produces a greater amount of wealth. As a greater amount of wealth is produced, the society is able to maintain a larger population. The larger population means more complexity and the further development of the division of labor, market exchange, productivity, and wealth (Hayek 1988, 126–34). Thus, for Hayek, social evolution generates increasing size and complexity in a society. He is not concerned solely with population size, as Gray (1986, 141) believes. Hayek contends that the increasingly efficient use of information brought about by social evolution ensures that:
the material equivalent of any given individual share will be as large as it can possibly be made. In other words, while the share of each … will be determined partly by skill and partly by chance, the content of the share which is allocated to him by that mixed game of chance and skill will be a true maximum. (Hayek 1976, 119)
In Hayek’s account of social evolution, then, societies advance in terms of population size, and the average level of want satisfaction will be as great as can be at a given moment. For Hayek, social evolution generates institutions that lead to increasing aggregate utility in a society, with the average level as high as it can be at a given time.
III. THE NATION-STATE AND SOCIAL EVOLUTION
Spencer and Hayek conceive of social evolution as a progressive process that produces larger societies. There are good reasons to disagree with them. Evolution is a type of explanation that simply refers to the unplanned emergence of order suited to its environment and has no inherent direction (Mises  1985, 367–70). While some of its nineteenth-century adherents, such as Spencer, Benjamin Kidd (1906), and Walter Bagehot (1906), regarded social evolution as progressive, albeit in very different ways,5 Darwin ( 2013) himself and other theorists such as William Graham Sumner ([ca. 1900–06] 1992) and T. H. Huxley ( 1997) absolutely denied that there was any association between evolution and progress. Spencer and Hayek are wrong to identify social evolution as progressive, and they are wrong to include the nation-state in their understanding of social evolution. In this section, I show why social norms that promote voluntary cooperation are likely to break down in the mass societies created by states, resulting in the very government growth Spencer and Hayek warned against.
Evolution Is Adaptation, Not “Progress”
A “liberal” evolutionism is a theory of the emergence of voluntary cooperation but has no internal criterion of “progress.” Darwin’s ( 1968, 115) theory of biological natural selection is elegant: where there is scarcity, there is a “struggle for existence.” Organisms that produce the largest number of offspring that reach adulthood and reproduce carry those traits that enable them to survive in their environment. Those that fail to reproduce occasion the gradual disappearance of their unique traits. The traits represented in the larger number become the dominant traits. Darwin ( 2013, 56) argues that among social animals, which include human beings, there is a “moral sense” that is oriented toward group survival. In animals other than human beings, it is instinctive. In human beings, that moral sense involves restraints on our conduct toward others that tend toward group survival, even if it is not thought of that way. Darwin argues that as human societies grow larger (beyond the small group), humans extend these moral sentiments to a larger sphere of persons. His thinking here is definitely influenced by the Scottish Enlightenment theorists, such as David Hume and Adam Smith, who maintained that this moral sense gives rise to conventions that promote cooperation among people.6 These nevertheless remain distinct. That is, biological evolution, where it does not involve social behavior, is indeed based upon the “struggle for existence.” Social evolution, on the other hand, tends to reduce that struggle within the group. Thus, as “Darwin’s Bulldog” T. H. Huxley maintained, social evolution works against the “struggle for existence” ( 1997, 299–300).
What can be distilled from theorists of social evolution such as Darwin, Spencer, Hayek, and others, is that social evolution has the following characteristics. First, conventions emerge through individual adaptation and spread through a group through learning and imitation. This does not entail material benefits for each individual following a social rule. In some instances this will clearly be the case, though in some types of learned behavior, such as altruistic behavior and adherence to nonrational rules of conduct that involve a sense of rightness and self-restraint, there is no discernable material payoff to the individual following the rule. Second, rules that have spread through a group persist because they enhance the survival chances of the group following them and thus continue to carry the social rule forward. This kind of logic offers a functional account of the persistence of social rules among groups but does not explain their emergence in the first place.
Where individual conduct is selected because it enhances individual welfare, individual learning and imitation are straightforward: people follow a practice because it is rational to do so to attain their goals. This selection does not require that the behavior enhance group survival chances or increase the attainment of ends by the group as a whole, nor does it imply that the behavior will do these things. If, for example, conditions make dishonesty, theft, promise breaking, and the like rational in terms of individual welfare, those behaviors can be expected to become dominant among the group, because they have a better payoff for the individual than do behaviors that enhance group welfare.
Consequently, the real problem is to explain the emergence of behaviors that enhance group welfare or, barring that, the conditions under which such rules are more likely to emerge. We can easily imagine conditions that are inimical to the establishment and persistence of rules that tend to promote group welfare. The familiar prisoner’s dilemma game (see, e.g., McCarty and Meirowitz, 2007, 87–88) offers a useful illustration. Let Row and Column be individuals with ordinal preference rankings A > B > C > D. Each has choices available to follow a rule of cooperative self-restraint (cooperate) or one of advantage taking (defect), with payoffs as follows:
There is a unique Nash equilibrium where both players defect because neither player can improve his payoff by changing strategy when the other player defects. If the players are rational, they reach this suboptimal outcome (C, C), but both would be better off, realizing the optimal outcome (B, B), if they chose to cooperate.
The theory of social evolution predicts that rules facilitating cooperation that emerge among a group will persist because they enhance group welfare. Thus, social evolution hypothesizes that a rule of cooperation will persist if it emerges among the players. We do not know why such a rule would emerge. Perhaps they have adopted a philosophical or religious conviction that it is right to cooperate when others are willing to do so. What matters for the theory of social evolution is that if they adopt such a rule, the rule will persist among them where they interact with each other because it enhances group welfare.
Key to this result, though, is that the individuals are able to identify one another as persons who follow the beneficial practice. Otherwise, they should expect other rational persons they encounter to defect and do likewise themselves, leading to the predicted and suboptimal equilibrium outcome. Where the group is small enough that the individuals composing it interact repeatedly and can identify cooperators, a rule of cooperation may persist. James Buchanan (1978) and Robert Axelrod (1984) argue precisely this. Axelrod showed in computer simulations of the iterated prisoner’s dilemma game that a mixed strategy of “tit for tat,” which is to cooperate on the first interaction with other players and thereafter follow the other player’s choice on the last interaction, emerged as an “evolutionary stable strategy” in competition with strategies of always cooperating or always defecting. Buchanan, Axelrod, and others demonstrate that when group size exceeds that which enables individuals to identify fellow cooperators, a pure strategy of defection becomes dominant.
Cultural Evolution and Group Size
Spencer and Hayek rely upon group selection with individual feedback mechanisms to explain social evolution. But as shown in the preceding section, the emergence and stability of the kinds of cooperative practices (i.e., cultural or moral rules) that Spencer and Hayek have in mind require that individuals interact repeatedly with either the same individuals or others who share adherence to the same cultural or moral rules. This, in turn, requires that groups be relatively small in size (and, as discussed below, be partially isolated from each other in territory). Viktor Vanberg (2014, 42–46) offers a useful distinction between “conditioned” and “unconditioned” spontaneous order. “Conditioned” spontaneous order emerges in the context of a framework of rules that imposes some constraints on individual choice and conduct, such as the market. “Unconditioned” spontaneous order emerges in the absence of such constraints through a “blind” process of variation and selection; biological and cultural evolution serve as examples.7 Since the market requires known conditions to flourish as a spontaneous order, Vanberg rightly asks about cultural evolution whether “‘suitable’ framing conditions are required for its ‘beneficial’ operation” (Vanberg 2014, 48). While it is not my task here to fully specify the conditions for social evolution that would have the efficiency properties that Spencer and Hayek have in mind, theoretical work to identify such conditions is needed to determine whether the “evolutionary” liberalism they championed is a viable enterprise. Indeed, it is a common criticism of Hayek that he fails to specify such conditions (See, e.g., Witt 1994, 178–89). I argue in the following that one such condition—group size—has important implications for institutional design if cultural evolution is to generate the socially useful results that Spencer and Hayek believe it does.
Rule following that promotes cooperation demands self-restrained behavior toward others. If such self-restrained behavior is to persist, individuals must be able to avoid being taken advantage of by uncooperative individuals. Cooperative individuals can avoid being taken advantage of if they know who is unlikely to cooperate, and they are more likely to be able to make accurate predictions about others’ behavior in a smaller group than a larger one (Taylor 1987; Axelrod 1984). When the group is too large for individuals to be confident in their expectations of spontaneous cooperation from others, coercion becomes necessary to ensure rule following (Olson, 1971, 45; Hardin 1982, 185–200; Ullman-Margalit 1977, 22–47). A significant implication of this analysis is that “the spontaneous order which we call society,” as Hayek puts it, will not spontaneously “form” beyond the size that permits individuals to know what to expect from each other. In other words, a society that is a spontaneous order will be a relatively small one.
Much research on cultural evolution has focused on the emergence of altruistic behavior, which presents the question of how individuals who practice altruism avoid being exploited by others who do not. Spencer himself analyzed the emergence of altruism in evolutionary terms (Spencer  1982, 1:271–85). Altruism is useful for analytic purposes, because it involves a self-restraint that characterizes adherence to moral rules that regulate our interaction with others by imposing constraints on our alternative courses of action toward them. For this reason, analysis of how altruism may emerge and remain stable helps explain how social rules of the kind Hayek and Spencer have in mind could emerge and persist.
Group size and partial isolation by territory are key variables in the emergence of altruism by group selection, because individuals must interact predominantly with others who share the same behavior in order for altruism to become stable (Smith 1964, 1145–47). More recent research reaffirms that small group size and isolation are key conditions in the evolution of altruistic behavior (Cooper and Wallace 2004, 316). Research employing individual selection in the form of behavior responding to collective action problems likewise concludes that group size is a key factor in the emergence and persistence of cooperative behavior. In a large group in which individuals do not know the strategies employed by others (cooperation or egoism), egoism will emerge as the dominant strategy and drive out cooperation (Ostrom 2000, 145).
Hayek and Spencer contend that social evolution is a process of increasing group size and complexity to sustain a larger population and a higher standard of living. Rules of conduct are adaptations that facilitate the increase in population and complexity of relations within a group. Clearly, if group selection requires a small group size to enable members to avoid the problem of free ridership that exploits constrained choice or altruistic behavior, group selection cannot explain the emergence of new practices that benefit the group as it grows to an unlimited size. In a group that is large enough, new rules will cease to emerge, and traditional rules will decay and disappear.
The technological and economic advance affected by social modernization in mass societies leads to the gradual disappearance of easily identifiable social subgroups (Almond and Powell 1966, 24–25; Huntington 1968, 30–35). Further, the emergence of mass societies produced “the anonymous individual—less powerfully socialized for moral response, and inhabiting new social contexts in which the old moral and personal prescriptions failed to apply, and liberated in considerable measure from self-control of the old kind” (Wilson 1985, 324). These phenomena make the theory of rule selection by group less plausible as our shared expectations about others’ conduct diminish and as partially isolated groups following different rules disappear. As a group becomes larger and more complex, the evolutionary process that Spencer and Hayek envision should begin to lose the efficiency properties that they attribute to it. Thus, while Spencer and Hayek argue that social evolution tends toward the development of larger, more complex societies, it appears instead that social evolution with the efficiency properties they predict will work among smaller groups rather than increasingly large societies.
The Extended Order of the Market Is Not a Group
When Spencer and Hayek discuss the increasing size and complexity of social order generated by social evolution, they naturally have in mind the modernization that emerged in Europe over centuries and took off rapidly with the Industrial Revolution. As this process unfolded, productive capacity and population grew. Population pressure required that there be more trade beyond the boundaries of small communities, and the emergence of larger urban centers supported the development of manufacturing and finance. Population pressure spurred on innovative production techniques, making possible the support of larger populations at a higher standard of living than ever before in human history. When Spencer and Hayek wrote of the development of modern market societies, they clearly visualized the historical process as it unfolded in the nation-states that were present in Europe from the seventeenth century on, and the benefits they saw in this process involve the increasing standard of living and political liberalization that have characterized that period.
Since what Hayek calls the “extended order” (1988, 6) of market cooperation grew up in the environment of nation-states, it is unsurprising that observers would regard the larger populations that the emergence of modern economies supported as groups which are specifically the populations of nation-states. David Rose (2011) views the cultural and moral background of market economies as a means by which free ridership on the trustworthiness of others in the large group context is overcome. This cultural and moral background, he argues, enables large groups with greater productive capacity than smaller groups to sustain institutions that support private property and markets. Rose is right about the necessity for a cultural and moral background of keeping agreements, trustworthiness, and the like as essential to the preservation of market economies. But the extended order of market cooperation is not a group. In fact, it is the willingness of persons to engage in economic transactions with people who are not members of their own communities that gives rise to the extended order and makes it possible. The increasing population supported by market relations of increasing complexity is not a group, but many groups, and it is also not a state. We should not conceive of the extended order of market cooperation as in any sense coterminous with the territorial jurisdiction of states. When Hayek and Spencer refer to the increasing size and complexity of society as a concomitant of social evolution, they do in fact seem to be thinking in terms of modern states. Insofar as they are thinking of modern states, one reason they are wrong is that they refer to the extension of the market order, which does not necessarily involve recognizable social groups.
Persons who engage in trade with one another do not, by virtue of that fact alone, form a group. A social group will include economic interaction, but it includes much more besides. The predominant theories through which sociologists understand the concept of a “group” are social identity theory and self-categorization theory, both of which maintain that a group of people is an aggregation who think of themselves as a group or are characterized by others as a group (Turner 1999). Market relations do not generally involve groups in this sense. If I am a vendor of flooring material, for example, I do not think of myself and my wholesalers, distributors, and customers as a “group,” because while there is a functional linkage that connects me with the others, we never form anything like a group. I might think of a trade association to which I and other vendors of flooring material belong as a group, but that is not a group engaged in market activity in any ordinary sense. In fact, a trade association is more likely to be engaged in political actions than in economic transactions as its principal activity. It is in part because of the impersonal nature of market transactions that they are able to assume the level of complexity and abstractness that Hayek and Spencer correctly attribute to them.
Identity and other commonalities in groups do not preclude the increasing size and complexity of the extended order of cooperation. People who have nothing else in common with one another can engage in cooperation on a limited scope for the sake of mutual benefit such as voluntary exchange. Thus, for example, a business engages in importing goods from another country for resale, and they purchase goods from trading partners with whom they have no other social relationship than that specific, narrow relationship of exchange. If either of them decides that they do not trust the other as a trading partner, they can quit doing business with them and they can turn to someone else. In this way, the discipline of the market can operate in the extended order among people who do not share values with each other beyond rational self-interest. This extended order is different than a group, because these persons do not do anything other than engage in limited, narrow types of voluntary exchange for their own mutual benefit. They are not (or need not be) neighbors, do not engage in more intimate forms of social relations beyond trade, and do not engage in collective action with one another. If they have any of the foregoing in common it may be entirely accidental, and mutual benefit rather than group identity or self-categorization is the basis of the market relationship. The market order offers the advantages that it does precisely because the emergence of its spontaneous order of trade relations is not dependent on the existence of a group, and that is why it can grow increasingly large and complex, as Hayek and Spencer observe.
The extended order of cooperation can be facilitated by small groups with a culture that promotes cooperation within the group and which incidentally gives rise to the kinds of moral values that make persons trustworthy in market transactions with people outside their communities (Rose 2011 refers to these values). Such communities are, or can be, the building blocks of the extended order of cooperation. Large empires are not only unnecessary, but, as I argue in the following sections, they undermine the moral culture that can facilitate the extended order of market cooperation.
Social practices that support trade among persons who have little in common with one another culturally emerge to support mutually beneficial exchange relationships. Peter Leeson (2014) notes that the familiar example of the Law Merchant facilitated international trade beginning in the Middle Ages and enabled persons who differed culturally and in other ways to engage in trade without government regulation. Leeson also describes interesting and often colorful examples of how persons with facially antagonistic interests, such as herders and reivers along the Anglo-Scottish border from the late Middle Ages through the early modern period, who developed a system of customary law to resolve disputes.8 This shows how a regime of evolved rules that perform a narrow and specific kind of social function can emerge without governmental involvement, and it also shows that social norms can emerge and persist among networks that do not constitute geographically defined societies. He also describes how European traders in nineteenth-century Africa were able to develop conventions that facilitated mutually beneficial exchange such as, for example, the emergence of credit transactions to make trade more profitable than banditry for indigenous peoples. Mutually beneficial trade does not require the existence of a group with a common culture. Trade relations are of a narrow and specific type. They do not include something like self-governance among a community for an indefinite period of time. Communities do that, but the extended order of market cooperation does not require much depth of social relations, which is what makes the extended order possible. It is the great achievement of markets that they extend across groups and do not require a large group in the sociological sense to operate.
The extended order of market cooperation is not a group, so it would be a mistake to conflate the development of the extended order of the market with the growth in size of states. States seek to preserve themselves by increasing their assets of power, which they can do by a variety of means, some of which involve increasing their territory to obtain control of human and material resources (Morgenthau 1985, 122–40). A principal factor in the increasing size of states historically has been advances in military technology; whether the increasing growth of states contributes to economic development depends upon factors other than sheer size, such as how the state protects property rights or seeks to redistribute wealth (North 1981, 64, 96–97, 143–57). States increase their size by different means and for different reasons than the expansion of the spontaneous emergence of cooperation among people for purposes such as trade that extends across different groups. Influencing trade or controlling trade routes have often been means by which states seek to increase national power, including by controlling territory through imperialism, but this kind of state action is the antithesis of spontaneous market order.
The State and Cultural Disintegration
The optimism about progress in the nineteenth century was replaced by a sociological pessimism in the twentieth, and I tie this narrative together with the string of evolutionary theory, which hypothesizes that voluntary cooperation will diminish among groups that are too large for evolved rules that support such cooperation to persist. Into this vacuum steps the coercive power of the state, and there is a general tendency for that power to grow and become more centralized. There is a direct line from Oppenheimer (1922) to Nisbet ( 2010) showing how the modern state causes intermediate social structures to deteriorate. Oppenheimer (1922, v) is clear that the state, which exists to expropriate and control wealth, deliberately seeks to dominate such structures in a manner that will weaken them. Nisbet emphasizes how the state supplants intermediate social structures, such as the family and religious bodies, in providing for order through centralized coercion rather than decentralized socialization. The aristocratic warrior class that created the nation-state through war and conquest (Oppenheimer 1922) established the large territories that were a fertile ground for the emergence of large industrial concerns in the first century of the Industrial Revolution. It is not principally the extension of the division of labor and development of the money economy, which Weber ( 2002;  1947) and Simmel ( 1990), for example, argued were responsible for the emergence of larger industries and growing territories, but the state in the first instance that preceded these large production units and territories. The advent of these large enterprises with bureaucratized, routinized innovation and production undermined the social esteem of entrepreneurship and the value of freedom of contract and private property (Schumpeter  2008). Schumpeter argued that capitalism undermines its own foundations in favor of socialism as it had previously undermined those of feudalism.9 But capitalism does not do this alone. The large nation-state created by the precapitalist aristocracy set the stage for the social and economic conditions that favor the emergence of socialism.10
The state changes the environment in which people live, and people adapt to that environment in a way that erodes traditional, evolved constraints on behavior. It establishes a territory determined by political power rather than by voluntary cooperation. While formal analysis of social evolution demonstrates that voluntary cooperation emerges only where groups are small enough that individuals can identify those who will cooperate (and those who will not) and refuse to cooperate with those who will not cooperate with them, the scope of the territory the state creates is as large as the state can hold. Where this territory is larger than the scope that voluntary cooperation can support, voluntary cooperation becomes less viable as a self-perpetuating means by which individuals can survive and achieve their goals and by which societies can sustain themselves. Norms of conduct once enforced by communities through noncooperation with those who do not observe evolved norms weaken and disappear, and these are replaced by coercive rules enforced by the state. Where voluntary cooperation diminishes, political power and coercion become more necessary to provide order in society. The state, initially established as an instrument of domination and exploitation, makes itself “necessary” to ensure order. Thus modernity is the experience of diminishing shared norms and increasing state power. The state, established everywhere to dominate and exploit persons, thus makes itself necessary to the perpetuation of a society it has created by means of power.
Intermediate social structures such as family, religious bodies, civic groups, fraternal organizations, and less formal social groups play an important role in preserving and transmitting values and social norms, and such structures are in themselves important sources of social order. Tocqueville recognized in American society of the early nineteenth century how civic associations accomplished all manner of collective purposes without governmental involvement: “Everywhere that, at the head of a new undertaking, you see the government in France and a great lord in England, count on it that you will perceive an association in the United States.” (Tocqueville  1951, 2:106). But as societies grow larger and more complex, the capacity of voluntary association diminishes and the state progressively replaces voluntary association with political power:
It is easy to foresee that the time is approaching when a man by himself alone will be less and less in a state to produce the things that are the most common and the most necessary to his life. The task of the social power will therefore constantly increase, and its very efforts will make it vaster each day. The more it puts itself in place of associations, the more particular persons, losing the idea of associating with each other, will need it to come to their aid: these are causes and effects that generate each other without rest. (Tocqueville  1951, 2:108)
Voluntary associations weaken and decay with the rise of mass societies created by nation-states. As these social structures deteriorate, the function they perform ceases to operate as before. This means that they cannot reliably perform their former role as preservers and transmitters of values, and the anonymity of mass society becomes more and more the predominant mode of life, particularly in urban areas. As voluntarily association weakens, the state steps in, as Tocqueville, Oppenheimer, Nisbet, and others have expected, and seizes control of functions once performed by smaller groups.
These intermediate social structures represent and constitute “social capital” in that they are repositories of values and evolved social rules (Putnam 2000, 19–26). As social capital diminishes, behaviors such as altruism and attitudes such as trust diminish as well. Shared values can mediate and prevent social conflict, but the disappearing social capital of mass society means that one should expect more conflict among individuals and groups competing for political and other advantages over each other. Altruism that promotes cooperation diminishes, and trust that inhibits conflict diminishes, and the result is more conflict and less cooperation.
These are called “intermediate” social structures, because they create a buffer between the individual and the state (Nisbet 1986, 21–22). As these weaken and disappear, with the resulting increase in social conflict and diminishing voluntary cooperation, the state exerts more power directly over individuals to produce the order lacking from evolved and natural structures. The state, by creating an environment in which intermediate social structures decay, makes itself more “necessary” in that the state must impose order where the emerged order supported by intermediate social structures deteriorates along with those structures.
The specific functions performed by intermediate social structures are to transmit social rules and to establish networks among individuals in which relationships involving reciprocity and trust can emerge and persist. It is in such networks that social rules are enforced through cooperation and noncooperation. Traditional rules that facilitate cooperation are not, as Hayek recognized, followed because individuals realize that they conduce to greater aggregate productivity and wealth. People follow such rules because they have been taught them by example and through training by the intermediate social groups in which they grow up and reach maturity. Families, religious bodies, and formal and informal social groups are the schools of morality in which traditional rules are transmitted through generations, and the enforcement of traditional rules through cooperation and noncooperation is a part of their transmission and persistence.
States create nationalism. They do this in part by reducing or eliminating the social functions performed by smaller, more local groups and communities. A national identity based largely on myth and abstract symbols replaces more concrete identities and attachments to intermediate social structures that individuals knew and interacted with. The extension of the territory of the effective social unit beyond that in which people can know those they interact with is a reason this happens. Nisbet ( 2010) argued that nationalism is what emerged as a substitute for the identity, attachments, and social functions performed by local communities, families, and churches before the advent and rapid growth of the nation-state. Nations are constructs. They are constructs because they are (poor) substitutes for the foregoing. Nation-states have supplanted the economic and security functions performed by those social structures without performing the essential social functions of transmitting values and providing identity, roles, and meaning to the lives of persons. The result is the mass society that Durkheim ( 1951) analyzed, in which anomic individuals are connected to nothing and lack a rational basis for conforming to social rules. This means that the state makes itself more “necessary” to provide the order that the civil society that it has destroyed once provided.
The evolutionary social theories of Hayek and Spencer are accounts of large-scale social change. Each is principally concerned with types of societies and is intent on showing why market-oriented societies displace alternative modes of social organization due to their capacity to promote economic development and sustain larger populations with better standards of living. Each is also really a sketch of how social evolution works, because both fail to devote attention to how social norms preserve and replicate themselves within societies. Hayek (1989, 136–37) mentions religion as a factor but does not regard religious institutions as important in social evolution. He does discuss how the nuclear family is important regarding the accumulation of property (1989, 137) but does not discuss its place in transmitting values. He recognizes that the growth of the “extended order” brings on the disappearance of traditional social groups, and he wants government to occupy much of the social space left by this disappearance. He says, for example, that social welfare programs are necessary because: “as a result of the dissolution of the ties of the local community, and of the development of a highly mobile open society, an increasing number of people are no longer closely associated with particular groups whose help and support they can count upon in the case of misfortune (1979, 54). Spencer, very much the nineteenth-century utilitarian, places no emphasis on religion or family, nor does he view intermediate social structures of any kind as important in preserving or transmitting values and norms to future generations. Both have seen the forest but missed the trees. Neither theorist has a theory of social evolution that explains how social norms are preserved and transmitted over time. Because of this, their theories of social evolution are incomplete. A theory of social evolution that explains how the free societies they have in mind will replicate the norms that make them possible must explain how these norms preserve and replicate themselves among people. This Spencer and Hayek fail to do.
There are striking similarities between Spencer’s and Hayek’s theories of social evolution. Comparing their work side by side, we can see the outlines of a type of defense of classical liberalism that incorporates social evolution. There is a significant flaw in Spencer and Hayek’s view that social evolution produces ever-larger societies, because the evolutionary process they describe is not likely to function in the large, mass societies which they envision as its product. In fact, that process is likely to break down, precipitating the growth of government that further erodes and destroys evolved norms and social institutions. A “conditioned” evolution consistent with thought will not include the nation-state. Left open here is the question of whether a libertarian society that will persist must be stateless or whether small states on the model of classical liberals from Montesquieu to the Antifederalists could remain free societies in the modern world, because that requires addressing additional ethical and political issues beyond the scope of this paper. One thing is clear, though: a libertarian society that remains free in the modern world will be a small society.
1. “Social Darwinism” is a concept invented by Richard Hoftstadter ( 1983) and never described a real movement in sociology or political thought. Hofstadter created the philosophical bogeyman of a laissez-faire capitalism grounded in the idea that individual competition would eliminate the weak and unfit and produce a superior species, but the authors to whom he applied the term, particularly Spencer and William Graham Sumner, did not really employ such an idea, so “social Darwinism” was for Hofstadter “an epithet to discredit views he opposed” (Leonard 2009, 41).
2. Spencer states that his comparison of organic and social order is only for the purpose of “illustrations of structures and functions in general.” “Here let it once more be distinctly asserted that there exist no analogies between the body politic and a living body, save those that are necessitated by that mutual dependence of parts which they display in common” (Spencer  1901, 1:592).
3. In Social Statics, Spencer argued that private ownership of land reduced the land available to be owned by others, so the right to property drawn from the law of equal freedom did not apply in this context. See Spencer ( 1970, 103-22). Spencer thus at one time argued for a form of voluntarism or individualist anarchism, at least for the ideal society, combined with land nationalization. Whether these ideas cohere is questionable, but Spencer later vociferously rejected his early land reform idea. See Spencer ( 1982, 2:455–58;  1982, 52, 116; 1904, 2:536–37).
4. Indirect utilitarianism is the idea that we should select for their utility-generating consequences social systems or broad processes as opposed to specific rules (as in rule utilitarianism) or actions (as in act utilitarianism). Spencer is an indirect utilitarian in his advocacy of evolutionary social development to facilitate utility-generating consequences.
5. Kidd argued that social evolution involved the spread of religiously inspired altruism, and Bagehot argued that evolution produced rational deliberation over moral and political matters.
6. Darwin, as well as Hayek, was influenced by the Scottish Enlightenment theorists, especially their conception of a moral sense. See Marciano (2007). Hayek sees the “emergence of the twin concepts of the formations of spontaneous order and of selective evolution” (Hayek 1988, 146) in the social theory of David Hume (Hayek 1967, 106–21) and Bernard Mandeville (Hayek 1978, 249–66). For the influence of Hume, Smith, and Ferguson on the development of the theories of spontaneous order and social evolution, see also Hamowy (1987).
7. The idea of conditions for spontaneous order is also reflected in Jasay’s distinction between “first-order” and “second-order” orders, where the “second-order” consists of principles or rules, which may or may not be themselves products of an evolutionary development, that require enforcement for the emergence of a “first-order” spontaneous order. (Jasay 1994, 51–52). Hayek’s lack of clarity about what can be properly designed in an otherwise spontaneous order is a common objection to his work. See, e.g., Kukathas (1989, 103–04).
8. Leeson (2014, 48–50) notes that a moral sense of honor was important to upholding the customary Law of the Marches. It is noteworthy in connection with Leeson’s observation about the efficacy of a sense of honor in upholding the system of customary law that a common “culture of honor” in which one’s own reputation was highly valued was present among the herdsmen of the border region between northern England and Scotland (Brown and Osterman, 2012, 222–24).
9. Schumpeter attributes this in part to capitalism’s creation of an intellectual class that has social criticism as its occupation. Hayek (1949) presents a similar argument. Analytically, we can distinguish between social norms that facilitate cooperation and norms that are not in some obvious way linked to cooperative behavior. The latter often seem to be linked instead to group identity or cohesion, or to be linked instead with different social phenomena such as religious faith. Social norms that do not have some obvious beneficial purpose are most vulnerable to rationalist criticism. Because social norms exist as elements of cultural complexes of norms, it is not clear that we can either distinguish in practice between cooperation-generating norms and other categories of norms, and it is not clear that one could, as a practical matter, abstract only those norms that facilitate cooperation from a cultural complex and shed the rest to produce a new and improved social order. It is in part for these reasons that traditional conservatives (e.g., Kirk 1986, 8–9) counsel against such efforts. There is a tension in Hayek’s thought between his admiration for the socially liberalizing tendency of the “Great Society,” in which social norms diminish in scope (1979, 54–55), and his emphasis on the tacit and nonrational nature of social rules. It is arguable, even in Hayek’s antirationalistic approach to understanding social norms, that those which clearly have as their function to facilitate cooperation may be the tip of an iceberg of underlying social rules, practices, and values, and that removing some of the underlying pieces may cause the whole to collapse like a house of cards. The impact of rationalist criticism is not the principal focus of this essay, but clearly, as Weber and others have suggested, its impact has been great. It is relevant here because it is a part of the process set into motion by large societies that make traditional social rules vulnerable.
10. “Nation-state” is a somewhat inconvenient term used here to describe the very large territorial jurisdictions we observe throughout the world today, and in which governments seek to create large groups that identify with the nation-state, follow (to a large extent) one legal regime of rules, and engage in collective action as a national group. “Nation-state” is an inconvenient term, because there are, in fact, nation-states with relatively small populations, such as the states in the Balkans and the Baltic states. One cannot identify with mathematical precision when a society is too large for social conventions that facilitate cooperation to persist, so one cannot identify with such precision when a state is too large in that sense. There is, in fact, a tendency for states to occupy as much territory and control as much population as they can, so states tend to grow as large as circumstances and military power permit. Nevertheless, the argument presented in this essay is not of itself a normative argument against the state full stop. It could readily form part of such an argument, but it could also form part of an argument for small states, confederations of small states, or the like. That is an issue to address on another occasion.