Introduction: The Broader View
Introduction: The Broader View
Abstract and Keywords
This introduction to the book discusses the reasons for the expansion of economic thinking into the other social sciences, and argues that an integrated concept of economics and social sciences is not only possible but is indispensable. This is because there is no other way to do justice to the integrated lives of individuals and their choices in trading off objectives in the traditional domains of different disciplines, and no other way to understand all the options for (and implications of) human ambition. Thomas Carlyle's attitude to economics as the ‘dismal science’ is then outlined. The chapter goes on to discuss the content of each of the ten succeeding chapters in detail, under five main headings: broader theories of the firm and the state; the structure of incentives in the modern welfare state; ethnic conflict, discrimination, and coordination in social groups; the state versus the market; and theory, ideology, and cumulative research.
Why did economics expand into the suburbs? One reason becomes evident when we think about the most basic assumptions, explicit and implicit, in economic theorizing. It is not much of an exaggeration to say that most economic thought is constructed out of four “primitives”: four concepts so basic that they are the building blocks with which economic thought is constructed.
One of these primitives is that the individuals in any society have preferences, purposes, or utility functions. Another is that they have various resources or endowments (natural, human, and man‐made) that are useful in producing the goods and services that satisfy these preferences. The third is that there is a technology (susceptible to improvement through research, development, and innovation) that explains how much resources are needed to satisfy given preferences: that is, a stock of knowledge that determines the frontiers of the “production functions” by which resources or inputs are transformed into goods and services. The extent to which the individuals in any society can satisfy their preferences, i.e. their real incomes, are limited by the resources available to it and the level of technology: to increase per capita incomes, a society must either accumulate more capital or other resources or else discover more productive technologies. As Milton Friedman pithily put it, there are no free lunches.
The fourth primitive of economic thought—and of most lay thinking on economics—is so elemental and natural that it is usually not even stated explicitly or introduced as an axiom in formal theorizing. It is the half‐conscious (p.2) assumption that markets are natural entities that emerge spontaneously, not artificial contrivances or creatures of governments. The markets that a society needs, unless prohibited or repressed by government, may be taken for granted. Economists recognize that the transaction costs of some trades would exceed the gains, and that these trades do not take place. But such trades would not be consistent with economic efficiency anyhow. The tacit notion that the markets needed for a thriving economy will, if governments do not block their appearance, automatically be there is fundamental for almost everyone. Nevertheless, as we shall see, this assumption is wrong. Moreover, there was no way that economists could correct the error without going outside the city center.
Though the migration to the suburbs went in all directions, there were two boundaries where the growth of the discipline could be crossed only by moving into fields that other disciplines had already cultivated. One was the boundary that marked goods and services that, though important for well being, are not for sale in the stores. Some goods are not available through the market because they are indivisible: if they are obtained or consumed by any individual in some group or category, they are also available to the other individuals in that group or category. Such goods are, of course, public goods or externalities. If a levee protects anyone in a flood plain from a flood, it protects others in the flood plain as well; if anyone obtains cleaner air when the external diseconomy from air pollution is curbed by an effluent fee, everyone in the airshed can breathe more freely. Economists call the most basic or essential services provided by governments, including law and order, public or collective goods.
For reasons by now well known, large groups, such as the populations of nations, cannot obtain public goods or deal efficiently with externalities through voluntary action in the marketplace. Because the benefits of these goods go to everyone in some area or category—whether an individual has borne any of their costs or not—there is no incentive for individuals voluntarily to purchase or help to pay for them. Thus, it is typically in the government's power to impose compulsory taxation rather than voluntary market mechanisms that normally account for their provision. To deal with goods that are important for income and welfare, but cannot be provided to large groups through voluntary or market mechanisms, economists have had no choice but to extend their theory to cover the exercise of governmental power.
The second boundary that had to be crossed marked the many situations where the incomes of individuals and groups in a society depended not only on their endowments of productive resources and their productivity, but also on the use of power. Goods may be obtained by making them—and by taking them. They may sometimes be taken by individuals acting alone, as when one individual seizes the goods of another in a Hobbesian anarchy or in modern (p.3) street crime. They may also be taken through the same power that typically accounts for the provision of public goods: the power of government.
The more that is taken, the less is the incentive to make. The use of power affects the level and rate of growth of income as well as its distribution. Thus, the economist who wants to explain how much income is produced—or who gets it—has no choice but to take government, law, and politics into account.
2 The Array of Markets
Economists did not usually notice the dramatic differences in the provision of public goods across different types of society until they began to study societies that were very different from those in which most of them did their work. Until recently, almost all of economics was done in Western Europe, the United States, or comparable First World countries. Only after World War II did many economists begin to study what we now call the Third World, and only after the collapse of communism could they study the unprecedented transitions in the societies of the Second World. Though often neglected, the differences between the public good provision in the First World, on the one hand, and in the Second and Third Worlds, on the other, are extraordinarily important. Paradoxically, the magnitude and importance of these differences in public good provision can best be seen by examining markets in different types of countries.
Remarkably, markets are commonplace in poor as well as rich societies. Why? Because many trades—such as those that can be consummated on the spot—are self‐enforcing. The gain that some trades can bring to both parties (and the mother wit of the parties concerned) is all that is needed for the trades to take place. Such self‐enforcing trades give a society some gains from specialization in production and mainly account for such sustenance as the poor societies afford. The standard assumption that the markets needed to realize the gains from trade will spring up automatically is true when transactions are self‐enforcing.
But there are other trades where the quid is provided at one time or place and the quo at another, so the gains from trade will not be realized unless there is third‐party enforcement of agreements. The transactions in insurance and futures markets are generally not self‐enforcing. Neither are those between lenders and borrowers: would‐be borrowers can persuade others to lend them the money only if they can credibly commit to paying it back with interest, and this usually requires third‐party enforcement of the loan contract. Similarly, when equity investors seek the gains from pooling their capital in a joint‐stock (p.4) company, they can generally succeed only if outside enforcement of company laws prohibits corporate management from siphoning off the capital entrusted to them.
The markets required to obtain the gains from the foregoing types of trade do not emerge automatically, but normally are the product of social and especially governmental contrivance. They are typically a product of the legal institutions that enforce contracts and corporate law—and that protect property rights so that borrowers can obtain loans that are secured by the borrower's assets. These institutions are public goods that cannot be provided by market mechanisms, but arise only from what we call “market‐augmenting government.”
In the best‐governed societies, lending even for long terms is commonplace, and widely held joint‐stock companies account for a large part of total production. In the countries without appropriate governance—most of those in the Second and Third Worlds—there is little or no private long‐term lending (except in families and similar social groups), little or no private capitalintensive production, and few if any large corporations.2
So, the familiar tacit assumption—that the range of markets that are needed to reap the gains from trade are (in the absence of government interference) automatically available—is wrong. While self‐enforcing transactions take place spontaneously, there is no automatic process that creates the full range of markets needed for an efficient and prosperous economy. Until recently, many believed that all the communist countries needed to do to obtain a thriving market economy was to “let capitalism happen.” Most economic textbooks (and lay writing on economics) said little or nothing that would lead the reader to expect that a country that repudiated communism and looked forward to capitalism would not naturally or spontaneously obtain many of the markets and gains from trade that are usually taken for granted in the First World.
2.1 The Two‐Edged Sword
Thus, voluntary market mechanisms are not sufficient to provide for the enforcement of contracts, the prevention of anarchy, and the provision of other (p.5) public goods: the coercive power of government is also necessary. But this coercive power is a two‐edged sword. It takes us back to the second basic idea that helps account for the broadening of economics. Goods may be obtained, we recall, not only by making, but also by taking. The same governmental compulsion that is required even to obtain the full range of gains from trade can be—and often is—used simply for taking.
Taking, as we know, reduces the incentive for making, and there is often a lot of taking. Some countries' governments are kleptocracies: the leaders or their corrupt subordinates are mainly in the business of taking. In many developing and once‐communist countries, and often also in the history of the West, this predation has been extraordinary (as parts of this book show), and has generated deadweight losses that are large in relation to the meager production of the societies concerned.
As is evident from some chapters of this book and from other writings, there is a lot of taking even in relatively well‐governed and prosperous countries. Much of it is not recognized for what it is because it involves an implicit rather than an explicit redistribution of income. Typically, when a lobby wins a tariff, a subsidy, a tax loophole, or a regulation that limits the competition it faces, this takes some income from society and shifts it in the direction of the group with the successful lobby. In general, the incentive to produce and to engage in mutually advantageous trade is distorted, and society's income is normally made lower than it would otherwise be. For reasons elaborated elsewhere and not spelled out here, implicit taking often makes societies poorer and more unequal.3 The social losses from special‐interest lobbying and cartelization are sometimes so large that they can account for substantial differences in the growth rates and income levels across countries. For reasons inherent in the logic of collective action, the interests that have the capacity for collective action needed to obtain redistributions through government have, more often than not, above‐average and sometimes very high incomes. This tends to make implicit taking increase inequalities.
There is also some taking that is, on balance, desirable or even essential, even though it interferes with making. We have seen that populations can obtain public goods only through taxation—through taking. When calculating the social cost of public goods, we must therefore think not only of their direct monetary costs, but also of the social costs of any extra taxation needed to raise the money. Taking for redistribution can also be, on balance, desirable. When the non‐poor decide to tax themselves to aid the poor, they presumably do this because they (not unreasonably) believe that the poor need the money more (p.6) than they do, so that the redistribution should increase social welfare. This can be true even though the taking for transfer to the poor also, as is widely known, distorts incentives. The required taxation increases the wedge between the social and private return to work and investment by taxpayers, and the transfers to the poor attenuate recipients' incentives to work.
Since taking—both when appropriate and when not—affects the incentive to make, and thus the level of income, there is no way that economics can leave it out of account. In other words, any logically complete economic analysis of the incentives to produce and engage in mutually beneficial trade cannot ignore the use of power. The use of political and governmental power is, of course, a long‐standing concern of political science and the law. As later chapters of this book show, the generation of power is linked in previously unrecognized ways with social groups and with the ethnic conflicts, social selection, and discrimination that characterizes many social groups.
2.2 Purposeful Life‐Plans and Self‐Interest
Critics may concede that the economist should understand more than the market, but they argue that the methods used to study behavior in the market won't work when behavior in government, politics, and social life is at issue. Standard neoclassical economic theory, according to some critics, necessarily assumes that all behavior is self‐interested and that—though this assumption may not be so far off the mark for the analysis of behavior in business and the marketplace—it is preposterously wrong when behavior in government, politics, and social life generally is at issue. This criticism is twice wrong.
First, nothing in economic theory excludes individual preferences where the individual has a concern for the welfare of others. The economist's type of thinking is not useful when individual preferences have no consistency or stability, but it does not require that individuals care only about themselves. Admittedly, economists (in this book as elsewhere) very often abstract from the extraordinary complexity of human motivation by assuming self‐interest and are thereby able to make problems analytically tractable. The conclusions of their analyses are usually robust—that is, not sensitive to the degree of error in their simplifying assumptions. But to say that an assumption is usually useful is not to say that it is indispensable—or always appropriate.
We economists are, however, guilty of neglecting one important class of cases where this assumption is especially inappropriate. There are some choices that are crucial for a market economy, yet self‐interest plays no role and it is principles and morals that mainly determines outcomes. Disputes about private property rights and about the enforcement of contracts are adjudicated by individuals who have no personal stake in the matters to be decided—and who (p.7) therefore cannot make self‐interested choices. The most prosperous societies, at least, leave these and many other important decisions to judges or jurors who, by virtue of governmental contrivance, can obtain no profit from deciding one way or another. If we did not expect that most people, when they had nothing personally at stake, would decide issues on the basis of principle, we would not want any decisions to be made by judges or juries.
The second problem with the twice‐wrong criticism is that it forgets that it is generally the same people who make decisions in the market and in political and social spheres. Most of these people derive their choices in part from a more‐or‐less integrated life‐plan and make tradeoffs across different spheres of life. Consumers, investors, employers, and employees in the market are also consumers, voters, neighbors, and members of families. Though there are exceptions, such as some directionless young adults that have not yet “found themselves,” most people work out coherent goals for their lives and make integrated decisions about “economic,” “social,” and “political” objectives. The woman who values both a traditional mother's role and a career may have to make anguished choices, but that does not mean she makes them randomly or irrationally. When families choose where to live, they typically consider how good the neighborhood is for educating and bringing up children and also commuting times and housing prices. Sometimes it is different people who are working in different spheres, as with those who are leaders in business and in politics. Yet there is no reason to assume that personal ambition or rational calculation vary much between them.
It follows that an integrated conception of economics and social science is not only possible: it is indispensable. There is no other way to do justice to the integrated lives of individuals and their choices in trading off objectives in the traditional domains of different disciplines. There is no other way to understand all the options for—and implications of—human ambition. A logically complete analysis requires the integrated and conceptually comprehensive type of thinking that is advocated here. Reality cannot be divided into departments the way universities are.
3 Dismal and Not‐So‐Dismal Sciences
It was Thomas Carlyle, in his “Occasional Discourse on the Negro Question” of 1849, who first named economics the “dismal science.”4 He disliked (p.8) Malthus's theory of population, with its pessimistic prediction that population pressure would keep the mass of humanity at the margin of subsistence. In part this was because he believed the population problem could be solved by more European imperialism and overseas settlement.
But Carlyle found political economy (as it was then called) dismal for much more fundamental reasons as well. To him, the defining sin of economics was that it “reduces the duty of human governors to that of letting men alone.” An apologist for slavery, Carlyle found the free black population of the West Indies indolent, and said that a marriage of philanthropy and the “dismal science” was to blame. What men everywhere needed was strong leadership and paternalism. Thus, he disliked the “ballot boxes” of representative government as well as free markets, and he admired feudal lords and British imperialism. A leader of the romantic movement, Carlyle's style of reasoning had nothing in common with the method of economics, and he ridiculed appeals to “statistics” and other “Fool Gospels.”
It is time for disclaimers. First, this focus on Carlyle's writings about economics may create a misleading impression of his work as a whole, and it certainly does not convey the brilliance of his prose style. Second, those who today find economics—or this book—dismal are not in any way guilty by association with those opinions of Carlyle's that are today so extraordinarily offensive (and that offended the economist John Stuart Mill in Carlyle's own time). Carlyle is relevant here because he named economics the “dismal science.”
He is also relevant because his complaint about Malthus's excessively pessimistic predictions applies to some unnecessarily discouraging formulations of modern economics. When our perspective includes the suburbs as well as the city center, we have a brighter as well as a broader view.
The needlessly dismal formulation of modern economics grows out of two of its primitive concepts: that a society has fixed endowments of productive resources, and that the amount of income or preference satisfaction obtainable from these resources is limited by the level of technology or productive knowledge. These primitives are often taken to imply that the only options we have are tradeoffs. A society cannot have more income, with present‐day technology, than can be generated from its endowments of tangible and human resources. So, society cannot have more of this without less of that. Neither can an individual. There is, we recall, no free lunch.
This somewhat dismal conclusion is reinforced by the theory of “efficient markets.” If some corporation has credibly announced that it will introduce a new product that promises it huge profits, it does not follow that we can make exceptional profits by buying its stock. Since the bright prospects of the company are public knowledge, others will have acted on that knowledge, and the price of the company's stock will already have been bid up to a level that (p.9) takes account of the discounted present value of the future prospects. More generally, the efficient markets hypothesis holds that it is not possible to make more than normal returns from publicly available information, and that professionally chosen portfolios will, on average, do no better than randomly chosen investments.
The theory of efficient markets contains a lot of truth; for example, professionally managed mutual funds, on average, fall short of returning as much as market averages and index funds by about the amount of their extra fees and expenses. Thus, most students of finance conclude that no investment formula assures more than normal returns. Just as the wit and self‐regard of those who precede us implies that we cannot expect to find big bills left on the sidewalk, so the enterprise of other investors keeps us from obtaining free returns in the capital markets.
This point holds not only in the financial markets but in the market economy as a whole. Just as any bills left on the sidewalk are picked up very quickly, so the rationality and enterprise of the actors in the economy implies that no industries, occupations, business strategies, or patterns of behavior will earn individuals or firms larger returns in long‐run equilibrium than are normal for the value of the tangible and human capital they possess. If an activity or strategy earns returns out of proportion to the value of the resources employed, more resources will be devoted to that activity or strategy until it offers no more than normal returns. When every kind of human and tangible capital earns its normal return, the economy tends to be efficient.5 In other words, the same elemental force that explains why we don't normally see big bills on the sidewalks also tends to make economies efficient—to ensure that neither individuals nor societies can obtain more of this without less of that. The idea that there are no free lunches has an even stronger justification than it initially appears to have.
Thus, Carlyle's complaint that economics is erroneously pessimistic still applies to many expositions of the subject. Though the dismal logic that there are no free lunches applies everywhere, its application to the poor countries of the world is especially important. It implies that the poor countries are poor because they are poorly endowed: because they lack the natural, human, or man‐made resources needed to produce high incomes. The poor countries are overpopulated, so they have do not have enough land and natural resources for their populations, and they do not have and cannot freely obtain the tangible (p.10) and human capital needed to generate high incomes. So, the abject poverty of many hundreds of millions of people can be overcome only when poor countries accumulate—or, improbably, are given—much more resources. The poor peoples of the world are, as the neoclassical theory of economic growth and standard econometric practice supposes, on the frontiers of their “aggregate production functions.” In short, there are not any big bills left to pick up on the footpaths of poor countries either.
3.1 Broader Is Brighter
This somewhat dismal conclusion depends on three implicit and often unnoticed assumptions. Two of these assumptions become evident when we look back at the logic that forced economics to create suburbs. First, the theory that there are tradeoffs but no free lunches applies only to societies in which there is no socially gratuitous taking. Though some taking is needed to finance public goods and to aid the poor, a very large part of the taking that goes on, as we pointed out earlier, serves no such social purpose. Since taking reduces the incentive to make, a society can increase its output without obtaining any additional resources when it curtails taking.
Second, the no‐free‐lunch theory also overlooks the many missing markets, especially in the developing and the lately communist parts of the world. We know that the countries of the Second and Third Worlds do not usually obtain the gains from trade that require third‐party enforcement. They do not, for example, now usually reap the gains from production with modern capital‐intensive techniques or from mobilizing capital in large private enterprises. If they impartially enforced contracts, they could obtain vast gains from trade from an array of new markets. By continually and impartially enforcing contracts with foreign as well as domestic investors and firms, the developing and transitional countries could tap the trillions of dollars of mobile capital in the developed countries—and continue to do so until the return at the margin to capital became the same as in the First World. (At this point they would have as much capital in relation to their endowments of labor and natural resources as the First World.) If the poor countries were to do this, their gains from trade in capital markets would buy more free lunches than anyone could count.
Third, the no‐free‐lunch theory overlooks the many economic policies that are simply stupid—that is, policies that may have no predatory purpose or involve no missing markets, yet are socially inefficient. If these policies were eliminated, the resulting boost in output would be enough to compensate those who gained from them and still leave something left over for others. Usually, a wider mastery of the economics of the central city is all that is required to remedy these stupidities, so it would take us far afield to go into them here. (p.11) Though the economics of the central city has many sins of omission, it has very few sins of commission. If it were more widely understood, there would not be nearly so many irrational policies, and we would all be better off. Thus, the idea that there are no free lunches is not entirely consistent with the theory from which it is derived: it overlooks the extra lunches societies could buy if they were to wise up.
What is the evidence for the quantitative importance of the foregoing argument? How could we determine whether the societies of the Second and Third Worlds have low incomes mainly because they have poor endowments, or mainly because they suffer more than the rich countries do from socially gratuitous taking, from missing markets, and from stupid economic policies? As it turns out, clear and startling answers to these questions emerge the moment that we look at the borders of countries and at the flows of labor and capital that cross them. The boundaries of countries delineate areas of different types and qualities of governance—of different economic policies and institutions, and thus different structures of incentives. We can learn a lot from the directions and magnitudes of movements of labor and capital across these borders. We can learn a lot from the changes in the productivity of workers that arise when workers migrate from poor to rich countries. And we can learn a lot from looking at the borders where rich and poor countries are adjacent to one another.
That, at least, is what is argued in the next chapter of this book, “Big Bills Left on the Sidewalk: Why Some Nations Are Rich, and Others Poor.” It claims to show that the low‐income countries of the Second and Third Worlds are poor mainly because they are much farther below their potential incomes than the rich countries are. If these countries improved their governance sufficiently, they would obtain colossal gains from foreign investment and advanced technologies. These gains are so large because the low‐income countries of the Second and Third Worlds can enjoy exceptionally rapid “catch up” growth. They can grow, as some low‐income countries have, at more than 7 percent per capita a year, and thus double their per capita incomes in a decade. By keeping this up for three decades, they could obtain an eight‐fold increases in per capita income.
There is no great likelihood that most poor countries will soon come to understand what changes they need to make in their institutions and policies, much less be able to undertake collective action needed to make the appropriate changes. Therefore, our broader perspective does not by any means imply optimistic predictions. Given the ubiquity of bad governance and the tenacity with which even the worst governments hang onto power, ours is not an exceptionally encouraging perspective. But it is not so dismal either, because it does call our attention to a brighter possibility: that any country can become more prosperous by improving its governance, and that the poor countries (p.12) of the world, if they substantially improve their economic policies and institutions, can escape poverty surprisingly quickly. This should be a source of hope for the poor peoples of this world. There are countless free lunches out there, even if misgovernment keeps many of the poor peoples from eating them.
3.2 Resistance to Innovation
The chapter immediately after “Big Bills. . . ” provides additional reasons for concluding that the rate at which a country grows is not pre‐determined by its endowments and depends much more on the extent to which it adopts superior technologies. This chapter, “Innovation and its Enemies: The Economic and Political Roots of Technological Inertia,” is by Joel Mokyr, who is the author of, among other works, The Lever of Riches,6 and thus has studied the Industrial Revolution and the revolutionary implications technical innovations can have.
Sometimes, Mokyr emphasizes, superior technologies are not adopted: “outright resistance to new technology is a widely observed historical phenomenon,” and technological inertia and economic stagnation have been commonplace. This is obviously an issue of momentous importance. In studying it, Mokyr argues—in keeping with the broader approach of all of the essays in this book—that “artificial distinctions between the ‘economic sphere’ and the ‘political sphere’ are doomed,” and that technological inertia is usually the result of rational behavior by utility‐maximizing individuals.
The market, Mokyr points out, provides an aggregation of the gains and losses from an innovation. If the market alone determines whether innovations will be adopted, the innovations that provide a greater balance of market gains than market losses will be adopted. But the market is by no means the only way of aggregating the gains and losses from adopting an innovation: a variety of regulatory or political processes can be used to determine whether an innovation is to be adopted, and each will in general aggregate the gains and losses differently and thus will often come up with different answers about whether an innovation should be adopted.
Though Mokyr is mindful of concerns in many societies about social stability, and aware that the adoption of some innovations might inappropriately disrupt it, what is most striking are his many examples of organizations that have (surely harmfully) resisted the adoption of superior technologies. These include the artisanal guilds of pre‐modern urban Europe, which “enforced and eventually froze the technological status quo,” and similar organizations in China. While emphasizing the importance of other factors as (p.13) well, he argues that differences in the power of guilds was one of the reasons why the Industrial Revolution occurred in Britain rather than on the European continent. He also offers other examples: shopkeepers in Germany in the late nineteenth century persuaded states to impose discriminatory taxes on large department stores; organized workers in Bombay in the 1920s and 1930s resisted technical and administrative rationalization; printers and other workers in London's Fleet Street frequently interrupted production and resisted innovation; unions in the European auto industry resisted flexible practices pioneered by Japanese auto manufacturers. (The social losses from such narrow or special interest groups come up repeatedly this volume.)
By contrast, Mokyr points to the many labor unions in such places as Sweden and Germany that have welcomed innovation, and notes that a union with an “encompassing interest”—one that represented such a large proportion of a nation's income earning capacity that it would obtain a significant share of the gains from a more efficient economy—has an incentive to accept superior technologies. (The argument that encompassing interests, because they have by definition a large stake in the productivity of society, tend to take its interests into account also recurs often in this volume.)
Taking all of Mokyr's examples together, we infer that the aggregation systems that most often resist superior innovations are those in which the separate groups of workers or firms that would lose from a given innovation have substantial influence. If this is true, the extent to which these interests are organized to lobby or to undertake industrial action is an important determinant of economic progress.
4 Broader Theories of the Firm and the State
If the two chapters discussed above are correct, it is the economic policies and organizational arrangements of a society that mainly determine how innovative and prosperous it is. Thus, ideas, such as those in the chapter we consider next, that can help to improve these arrangements are especially important. This chapter is by Oliver Williamson, author of (among many other works) The Economic Institutions of Capitalism.7 In his chapter, “Economic Institutions and Development: A View from the Bottom,” he sets out and summarizes some economic theorizing that is considerably broader than most economics has been. As in most of his other work, he focuses mostly on the firm, and especially the corporate hierarchy.
(p.14) If any kind of organization has always been at the center of economic analysis, it is surely the firm. How, then, can we argue that this book is about the broadening or suburbanization of economics?
Williamson's analysis shows that a broader approach to economics than economists once thought appropriate is needed even for the study of the firm and industrial organization, and that the parallels between hierarchical firms and governments are startling. Williamson builds upon a point that Ronald Coase first made in the 1930s: the existence of the hierarchical firm cannot be explained except in terms of what can be called a “market failure.” Market failure not in the sense that government rather than private enterprises should be producing the goods, but in the sense that, in a competitive economy, the survival of an unsubsidized firm hierarchy can be explained only by the disadvantages or costs of markets. If the time of each worker and the services of each unit of the other factors of production that cooperate in some productive process were always most efficiently coordinated through the market, the costly hierarchies of the typical modern corporate enterprise would not be sustainable. That is, the production that the firm hierarchy organizes would be coordinated at less cost by the market and there would be no reason to bear the costs of the firm hierarchy. As Williamson's chapter points out, markets and hierarchies are alternative ways of organizing production. We can see that in some circumstances firms choose to organize more activity through their hierarchies, as when they integrate vertically and one larger firm coordinates activity previously coordinated by a market relationship among firms; at other times firm hierarchies can coordinate less and the market more, as when a conglomerate breaks up or a firm contracts out for some work it previously did for itself. We cannot, in other words, understand what does and does not go on in the marketplace unless we include firm hierarchies as well as market relationships in our theory.
That Williamson's analysis is broader is also evident from his conception of the firm. Traditionally, economists have conceived of the firm in terms of one of the primitives or building blocks we considered earlier. That is, they have taken the firm to be a “production function”: a relationship, given by the available stock of technological and other knowledge, between the resources or inputs that a firm uses and the goods or outputs that it produces. To Williamson, the firm is more usefully considered a “governance structure,” more an organizational than a technological construction. The general organizational logic that is applicable, for example, to governments is also applicable to firms, and the organizational logic that is evident in the firm is applicable to governments and other non‐business organizations.
For example, Williamson points out that the separation of ownership and control in the corporation has its parallel in government. Berle and Means pointed out long ago that the managers of widely held corporations did not (p.15) have to focus exclusively on the returns to the stockholders for whom they are supposed to work.8 Though neither Berle and Means nor Williamson put it just this way, the many dispersed stockholders have to overcome usually insurmountable difficulties of collective action in order to make their legal control over the management effective. So corporate managers may be able to keep their jobs even when they shirk, or to give themselves excessive perks, or to engage in unprofitable empire building. Even though managers of widely held corporations are also constrained by hostile takeovers and other features of the market for corporate control, these constraints are by no means always sufficient to ensure that widely held corporations are always run solely in the interests of the stockholders.
Williamson shows that the same thing happens in democratic governments. The electorate in a democracy, like the stockholders in a corporation, nominally has complete control. In practice, those who manage the government are often able to indulge their own preferences about government policy and other things. For this and other reasons, governments often fail, sometimes egregiously, to serve the interests of electorates. To put his point in our language, Williamson observes that goods and services can be obtained not only by making, but also by taking, and that taking makes societies less efficient.
Therefore, all mechanisms are imperfect and both markets and governments fail. Economists should not deplore any arrangement as irrational unless there is a remedy that can be implemented—some achievable alternative mechanism that will work better. Thus, Williamson, who has criticized many policies that have emerged from the political process, also counsels us to be “respectful of politics.” This is presumably in part because he emphasizes that appropriate governance is needed to protect property and contract rights. But he also points out that some of the social losses that arise because political power is used for socially costly redistributions in favor of those who wield it are a cost of democracy. There is sometimes no remedy for such losses of efficiency resulting from redistribution—no way that you can always prevent them.
We agree that there is no universal remedy—and sometimes no remedy at all—for the losses that arise from the use of political power for taking. Some constitutional rules are better than others,9 but surely no pages of parchment can hold back all the powerful political forces that can serve their interests (p.16) by obtaining protection against imports, restrictions on competition, tax loopholes, or other types of implicit redistribution. That is one reason why the not‐so‐dismal science is not necessarily optimistic, much less utopian. Nonetheless, most special‐interest lobbying serves the interests of small minorities, so this lobbying will not prevail if the public—or even the intellectuals—understand what is happening. This means that good economic research—and a wider dissemination of the research results and a better climate of opinion—can improve economic performance. Again, while the not‐so‐dismal science can offer no assurances, it does call attention to a brighter possibility.
4.1 A Broader Economic Theory of the State
Whereas Williamson proceeds “bottom up” from the firm to government and society, he agrees that it is also useful to proceed from the “top down”: to look at the incentives facing leaders of governments in various circumstances and the patterns of policies that result from them, and then to note the implications of these policies for firms and for the private sector generally. This is what Chapters 4, 5, and 6 do. The first of these, “Dictatorship, Democracy, and Development,” provides a version of one of Olson's presentations in India that helped give rise to this book. Because some of the subsequent contributions in this volume criticize, build upon, or go beyond it, we must provide a full summary of it.
The chapter is part of a series of writings that began with Olson's essay on “Autocracy, Democracy, and Prosperity,” published in 1991.10 It puts forth, in an intuitive and nontechnical way, a part of the theory that is set out with formal proofs and crucial additional results in Martin McGuire and Mancur Olson's “Economics of Autocracy and Majority Rule.”11 It analyzes the kings or dictators who control autocratic governments—and the oligarchies or majorities or other ruling interests that control other types of government—in just the way that economists analyze the behavior of firms, consumers, and workers. That is, it takes a broader approach to economics by applying the familiar assumption of rational self‐interest to the autocrats or other ruling interests that control a government, and then finds what types of policy will best serve the ruling interest.
When this is done, it quickly becomes evident that a rational autocrat, even if he began as the leader of a gang of roving bandits and is solely interested (p.17) in taking as much as possible from others, will take the interests of those he exploits into account whenever he has secure control that he expects will last for some time. His monopoly over taking, whether we call it theft or taxation, gives him an “encompassing interest”—that is, we recall, a large stake in the productivity of his domain. This makes him moderate his tax theft. He has an incentive to limit the rate of his tax theft because taxation reduces his subject's incentive to produce and thus also his tax base. If he took everything, he would eliminate the incentive to produce and would collect nothing. He maximizes his total tax collections by lowering his tax rate until what he gains at the margin from the resulting increase in the income of his domain, and thus his tax base, just equals what he loses from taking a smaller share of output. For instance, if his tax rate were 50 percent—and at that rate the last dollar he collected in taxes would reduce the national income by two dollars—he would be at his optimum.
An autocrat's encompassing interest also makes him use some resources that he could spend on his own purposes or consumption to provide public goods for those from whom he takes. If his optimal tax rate were 50 percent, he would obviously gain from spending his money on public goods up to the point where the national income increases by two dollars for each dollar spent, because he will receive one of these two dollars.
A majority or other ruling interest made up of people who earn income in the market will necessarily have a more encompassing interest than an autocrat. If such a ruling interest redistributed as much to itself as an optimizing autocrat would redistribute to himself, it would gain from reducing the redistributive tax rate. This increases not only the tax base, just as it does for the autocrat, but also the market incomes of those in the ruling interest. Majorities and other ruling interests that obtain a sufficiently large fraction of a society's income (“super‐encompassing interests”) will, as McGuire and Olson have proven, best serve their interests by redistributing nothing whatever from the minority to themselves. Majorities and oligarchies also provide more public goods than an autocrat would.12
Some aspects of economic policy in most societies are not, in fact, controlled by either an autocrat or a monolithic majority or oligarchy that rationally serves its collective interest. Most protectionism and other types of subsidy favoring individual industries or occupations arise because of pressure from organizations of firms, professions, or workers in that industry or occupation—from the influence of guilds, professional associations, trade associations, unions, or other special‐interest groups. The firms or workers in any single (p.18) industry or occupation are by no means an encompassing interest: they are, on the contrary, a narrow or special interest. If an organized interest in a particular market earned, say, 1 percent of the total income earned in a country or city‐state, it would not have any incentive to cease using its lobbying or cartel power to obtain redistributions for itself until the social income fell at the margin by 100 times as much as it obtained from redistribution; for it would, on average, bear only about 1 percent of the social loss from a less efficient economy. This logic suggests that countries that have a high density of narrow special‐interest organizations will tend to grow less rapidly than otherwise comparable nations.
A country develops a dense network of special‐interest lobbies and cartels only if it has enjoyed a long period of stability. This is because the benefits that lobbying or cartelization bring to the firms or workers in a market automatically go to all the firms or workers in the market, whether or not they have paid dues to the organization that organized the lobbying or cartelization. In other words, they must overcome the great difficulties of collective action by working out the complex agreements and “selective incentives” that make it rational for the firms and workers in a given market to pay the costs of collective action. It takes a long time to overcome these difficulties. That is why only long‐stable societies normally have a high density of special‐interest organizations and suffer the losses in efficiency and dynamism that they bring about.
This, along with international differences in ideology and in economic understanding, explains much of the variation in economic growth across the developed countries. Totalitarianism and foreign occupation destroyed most of the special‐interest organizations in the Axis countries (or replaced them with relatively encompassing structures of allied or other postwar design). Accordingly, in the first quarter‐century after World War II, West Germany, Japan, and Italy enjoyed “economic miracles.” By contrast, the same Great Britain that invented modern economic growth with its Industrial Revolution (see the Mokyr and De Long chapters) suffered from the “British Disease” of slow growth. In long‐stable and undefeated Britain and in the long‐settled and always stable parts of the United States, dense networks of lobbying and cartelistic organizations had emerged. Thus, these places suffered large losses in efficiency and dynamism from narrow or special interests, and this mainly explained their slow growth. By contrast, the more recently settled western and “defeated” southern parts of the United States were not much afflicted by such narrow interests and grew relatively rapidly.
The one country with an exceptionally high level of membership in interest organizations that also had impressive growth in the 1950s and 1960s was Sweden. At that time, with not much more than one big labor union and one employers' federation, it had a uniquely encompassing interest‐group (p.19) structure. Why did this exceptionally encompassing and socially prudent structure emerge in Sweden? Why may it have devolved and deteriorated into a narrower set of special interests with the passage of time? Why ultimately did the Swedish economy head south and discredit the “Swedish model”? As we shall see from Eric Moberg's chapter and later discussion, the answers to these questions raise general theoretical issues and turn out to be important for all countries.
We have argued that autocrats have an encompassing stake in their domains that gives them some interest in their productivity, but that majorities in democracies have more encompassing interests and therefore a greater incentive to take account of the interests of society. Special‐interest organizations, by contrast, face very different incentives. Except in the rare and unsustainable cases where they are encompassing, they represent only the firms or workers in a particular industry or market. These narrow interests give them little or no reason to take account of the social losses that their lobbying and cartelization bring about.13
In the long run, probably the most important difference between autocratic rule and representative government is in the degree to which they protect property rights and enforce contracts. Chapter 4 argues that, when an autocrat, because of uncertainty about his tenure or any other reason, takes a short‐term view, it pays him to take any asset whose tax yield over the short planning horizon is less than its capital value. Thus, any autocrat with a sufficiently short time horizon becomes, in effect, a roving bandit and takes all of the readily confiscable fortunes and assets in his domain. In oligarchies or democracies in which power is shared, no single individual has the power unilaterally to confiscate the assets of others. Those who share power also have an incentive to make sure that no one, including the leader of the government, can become a dictator, so they limit the power of the government and its leader. The resulting limitations also make private property and contract rights more secure. Because of this and the frequency of succession crises in autocracies, property and contract rights are much more secure over the long run in representative governments, whether they have universal suffrage or oligarchic electorates.
4.2 The Broader Approach to the State and the Rise of the West
The importance of the incentives facing the leaders of governments—and the intimate connection between these incentives and property and contract (p.20) rights—is evident from J. Bradford De Long's chapter, “Overstrong Against Thyself: War, the State, and Growth in Europe on the Eve of the Industrial Revolution.” His historical analysis of early modern western Europe demonstrates that it was the interests of kings and the forms of government that mainly determined whether there was economic growth or stagnation—and even partly explained the source of our modern world, the Industrial Revolution. De Long's recent experience in the US government also suggests to him that the same principles that explained the far‐from‐encouraging relationship of political power and economic performance in early modern Europe also apply in late twentieth‐century Washington.
De Long notes that northwestern Europe, over the long sweep of history, has mostly been a relatively backward area, and its status as the cradle of industrial life needs to be explained. He looks first at the cities of Europe and concludes that “city growth had a very strong allergy to the presence of strong, centralizing Princes who called themselves ‘absolute’ in the sense of being not subject to, but creators of the legal order, and a strong attraction to mercantile republics: city‐states governed by representative or not‐so‐representative oligarchies of merchants.” He draws on the quantitative evidence in his important article with Andrei Shleifer on “Princes and Merchants,” which examined urban growth and decline century by century in nine different regions of Europe.14 The De Long—Shleifer statistical analysis suggests that “each century that a west European region . . . was ruled by a strong ‘absolutist’ prince saw its urban population fall by roughly 180,000 people, and its number of cities with more than 30,000 people fall by roughly one and a half, relative to what the experience of that region in that era would have been in the absence of absolutist rule.” After looking at some specific features of European military and political history, De Long can show that there is a causal arrow going from the nature of the rule to the growth of cities. For example, in the year 1000 southern Italy outstripped northern Italy in agricultural productivity, population, and urbanization. But then “Robert I d'Hauteville and his brothers win their wars to bring southern Italy under their control, and its city‐states become part of the ‘prototype absolutism’ that was the Kingdom of Sicily in the first few centuries of the present millennium.” After five centuries of absolutist rule, southern Italy had become quite backward in relation to the more urban and dynamic north.
In pre‐industrial Europe, De Long argues, a city‐state was typically controlled by an oligarchy of merchant burghers. When landlords or burghers with substantial private wealth have an important political role in representative assemblies, they take into account the impact of alternative public policies on their private wealth. When Lorenzo di Medici “the Magnificent” guided the (p.21) government of Florence, his wealth depended as much on the revenues of the Medici bank as on the city's treasury.
De Long's results are completely consistent with his logical analysis and obviously also with the theory in Chapter 4 on “Dictatorship, Democracy, and Development.” An oligarchy, majority, or other ruling interest that not only controls the fisc, but also earns income in the marketplace, is bound to have a more encompassing interest than an autocrat whose income depends only on his control of government. A ruling interest that earns income in the marketplace, we recall, will necessarily best serve its interests by redistributing less to itself than an autocrat who was earning no income in the market would have done.
The evidence gathered by De Long and De Long and Shleifer also shows that absolutist autocracies often lead to short time horizons and thus, in effect, to roving banditry. Taking the kings of England between 1066 and 1715 as an example, De Long and Shleifer show that something went wrong in 18 of the 31 royal successions. They found that “there was only a 13 percent chance that the legitimate heir who was grandson, granddaughter, grandnephew, or other relative of an English monarch would inherit the throne without disturbance in the line of succession.” Though De Long does not go into this, one would suppose that in pre‐industrial Europe the fortunes in cities were more readily confiscable than the widely scattered assets in rural areas. If so, the form of government probably had more impact on economic performance in urban industries than in agriculture.
De Long's accounts of Hapsburg Spain and Bourbon France illustrate how imperial ambition—manifested especially in high military expenditures—implied excessive taxation and economic blight. In defending itself in life‐and‐death wars, the Dutch Republic was driven not only to tax itself heavily but also to borrow huge sums. This implied taxation so oppressive and long‐lasting that the Dutch economy ultimately stagnated. Somewhat later Britain was left with similar military exigencies and the high taxes and borrowing that go with them. De Long argues that an increased population with more taxpayers and other favorable breaks kept the British debt and taxation distortions from reaching ruinous levels, so that Britain was soon host to the Industrial Revolution.
5 The Structure of Incentives in the Modern Welfare State
The logic of narrow and encompassing interests and of different time horizons is no less pertinent to the democratic welfare states of the present day than it (p.22) was to the societies of medieval and early modern Europe. In some ways, the prototypical or most advanced or extreme example of the modern democratic welfare state is Sweden—famous since before World War II for the “middle way” between capitalism and communism and for an exceptionally large and egalitarian welfare state.
Erik Moberg's chapter on Sweden (Chapter 6) partly extends and applies the concepts of encompassing and narrow or special interests set out above and in some of Olson's earlier publications. It also attacks part of Olson's analysis and offers an alternative and novel theory of Swedish developments. Moberg points out that, though the Swedish economy is performing very badly in the 1990s, Sweden was for a considerable period famous for achieving a very high per capita income while having an exceptionally generous welfare state. The high per capita income of Sweden in the 1950s and 1960s needs explanation, as does the more recent severe decline of the Swedish economy.
Moberg reviews the explanation of these developments that Olson, using the theory of encompassing and narrow interests, offered in two books.15 He points out that Olson had, in the 1980s, asked two parallel questions: (1) Why wasn't Sweden's economy doing better? (2) Why wasn't Sweden's economy doing worse? Olson had said the answer to the first question was obvious and well known—it was because of the disincentives and distortions inherent in the exceptionally large Swedish welfare state—and that the real challenge was to answer the second question. Moberg emphasizes that on this formulation of the question the large size of the Swedish welfare state is taken as a given, but he is above all concerned to explain this.
Moberg and Olson agree that Sweden's economic growth has been slowed by the disincentives of exceptionally high welfare spending and taxation—and that it is also a long‐stable country with an exceptionally high density of membership in labor unions and employers' organizations. So why had not Sweden performed worse and got into trouble sooner? As we have already seen, the relatively wealthy and rapidly growing Sweden of the 1950s and 1960s had an exceptionally encompassing structure of interest organizations, and (as the theory of encompassing and narrow interests predicted) the policies of these encompassing organizations showed a concern for the economic efficiency of the society that is not evident in narrow special‐interest organizations. Though he had initially been agnostic about the very long‐run effects of such groups, Olson argued, beginning in 1986,16 that encompassing interest organizations tend to devolve over time into systems of smaller and (p.23) narrower organizations and into interest organizations that, when large, are mainly shells or fronts for increasingly powerful and autonomous branches and other ostensibly subordinate units. The Norwegian rational choice theorist Gudmund Hernes has independently made a similar argument.17
Moberg agrees that Sweden had relatively encompassing interest organizations and also that they tended to display the prudent concern about the impact of their policies on the productivity of the society that the theory predicts. But he emphasizes the need for an adequate explanation of why Sweden came to have encompassing interest organizations in the first place. He argues that Olson's brief and tentative references to Sweden's homogeneity, small size, and special historical circumstances are not by themselves sufficient to explain why Sweden came to have encompassing organizations.
To provide a fuller and better explanation, Moberg notes that Sweden has a parliamentary government with proportional representation and shows that this makes for highly disciplined political parties. In a presidential system of the kind that prevails in the United States, the administration does not require majority support in the Congress for its continuance, so no crisis or change of government need occur when members of the president's party vote contrary to the administration's policy. Such a system can operate with weak and undisciplined political parties. In a parliamentary system, by contrast, the government cannot remain in office unless it continues to have sufficient support in the parliament, so it must have one or more disciplined parties that continue to support it: the government, indeed, is essentially an artifact of the party or parties out of which it is made. With proportional representation, moreover, members of the parliament owe their seats not to any plurality in a district, but to their place on a party list of candidates and the electoral fortunes of that party. The political fortunes of a politician depend on the party's fortunes and how high he or she ranks on the party list. So the politicians stay in line and the political party has discipline and coherence. Moberg also points out that the Social Democratic party that has controlled Sweden for most of the last sixty years is linked institutionally as well as ideologically with one large labor union.
Accordingly, Moberg says it is not so much Sweden's homogeneity, small size, and the other factors Olson had mentioned that explain how it came to have an encompassing interest group structure, but even more the country's constitutional system. His argument certainly applies to the left–labor side of the spectrum, for it generates disciplined political parties, and on the left there is one large party that is institutionally linked with a labor union. Though Moberg's argument may perhaps not help explain the encompassing character of the Swedish Employers' Federation that represents almost all Swedish (p.24) business, it must provide a large part of the explanation of why Sweden came to have one unusually encompassing organization in the form of the Social Democratic party linked with one large labor union. It is important to acknowledge that Moberg here repairs one flaw in some of Olson's writings on encompassing organizations.
Moberg also disagrees at least partly with Olson's contention that Sweden's initial success with encompassing organization helps to explain why it over‐confidently went on to expand its welfare state redistribution beyond sustainable levels. He offers his own novel theory: that the Swedish constitutional structure was the most important factor in explaining the exceptional expansion of Sweden's welfare state. The disciplined political parties in Sweden lowered the transaction costs involved in putting together packages of redistributions that would command a majority. He contends that, by contrast, in presidential systems of the kind in the United States, it is much more difficult to work out the political deals needed to pass a package of redistributive measures. He argues, contrary to common opinion, that the US political system is not especially open to influence by lobbies.
Moberg in his debate with Olson has raised some issues that call for further research. In view of the fundamental importance of these issues, not only for Sweden but also for other countries (and for theoretical reasons as well), we must hope that a number of scholars will investigate these matters. Moberg's examination of the Swedish welfare state also demonstrates that the same tools of thought that are useful in analyzing the market, the corporation, and the kings and oligarchies of early modern times also illuminate the welfare state in a modern democracy. We cannot understand such an important feature of modern society as the welfare state without a broader economics or an integrated social science.
6 Ethnic Conflict, Discrimination, and Coordination in Social Groups
Chapters 7—Edward Montgomery's “Affirmative Action and Reservations in the American and Indian Labor Markets”—and 8—Russell Hardin's “Communities and Development”—turn to a different suburb of economics than those we have so far discussed. These two chapters examine the cultural, ethnic, racial, religious, linguistic, tribal, and caste groupings into which humanity is, to various degrees, divided. The consideration of such groupings raises what may be the best‐known criticism of the economist's paradigm and the theoretically unified approach to social science.
(p.25) This line of criticism correctly emphasizes that economic theory—and the rational‐choice approaches to social science that are integrated with it—take the individual as a fundamental unit of analysis. It is wrong to identify the economic and rational‐choice methodology with “individualism,” much less “rugged individualism,” as a social or political creed. This approach does, however, require methodological individualism. That is, it requires the kind of thinking that De Long brings out at the beginning of Chapter 5, where he reports his reaction to the White House functionary who said that “what you economists don't see, is that you are pushing for the public interest. But there are other interests that can be more important.” De Long's immediate and outraged reaction that the public interest would have to be the sum of private interests, so that no interest could possibly be more important, is one aspect of methodological individualism. But another aspect is evident in De Long's next paragraph, where he takes account of the fact that kings or other political leaders may sometimes place their individual welfare above that of their subjects. If the King says “L'état, c'est moi” and has the power to act accordingly, then a methodologically individualistic account of government must give a great role to the individual who is king.
Critics of the methodological individualism that characterizes the broader economics and the integrated approach to social science emphasize that everyone is born into and is socialized by social groups of the kind the Montgomery and Hardin consider. No man is an island, and individuals are greatly influenced by their different patterns of socialization. Those socialized pick up different cultures with different attitudes toward work, saving, and leisure, different religions and languages, different tolerances and prejudices, and different group loyalties and hatreds. Accordingly, some argue, any method of analysis that begins with individuals is mistaken because it leaves out an even more fundamental reality: the groups and communities that mainly determine what individuals in different groups believe and thus what they choose. Some other holistic, communitarian, nationalistic, or group‐based or class‐centered method should supplant the methodologically individualistic economics and social science on display in this book. Though socialization matters, and there are often important differences in the ways that different groups have been brought up, we shall argue, drawing on the Montgomery and Hardin chapters, that the integrated approach to social science nonetheless generates bigger payoffs than any other single approach.
The case for theory that starts with the group rather than the individual is probably strongest with problems of the kind considered in Montgomery's chapter. He considers race and “affirmative action” in the United States and compares them with caste and the “reservations” of government jobs for “scheduled castes and tribes” in India. In these cases groups are exceptionally important. When there is discrimination against a race or caste—or a (p.26) government program to assist or favor a race or caste—outcomes obviously cannot be explained only in terms of individual attributes: the race or caste an individual is identified with manifestly makes a difference. Yet no one who reads and understands Montgomery's chapter, and the rich literature he draws upon and extends, could deny that the methodologically individualistic method used by him and those he builds upon makes important contributions to our understanding.
To see why, consider a type of discrimination that, though probably not the most important, arises directly from individual choices. Suppose there is an attribute whose average value differs across groups and also among individuals within each group, and that it is impossible or costly to determine the value of this attribute for each individual. Adolescent and young adult males, for example, account for a disproportionate share of violent crimes, yet many of them would not commit a violent crime and many individuals in other demographic categories would. This means that a taxi driver in a high‐crime area may rationally avoid picking up young males at the same time that law‐abiding young males there reasonably resent the unjust discrimination that they suffer. As Montgomery shows, the assumption of rational behavior can not only help us understand the discrimination that arises from stereotyping, but also can illuminate the consequences of different policies for dealing with it.
A “taste” for discrimination (a desire to discriminate), like other tastes, cannot be explained by a theory of rational behavior, but economic theory can say a lot about the costs and consequences of such a taste when there are markets. The employer who indulges such tastes cannot usually obtain the best value when hiring labor and therefore pays a price for acting out of prejudice, as do workers and consumers who make discriminatory choices. Market forces make this type of discrimination less common than it would otherwise be. This logic helps to explain why governments, and extra‐legal sources of coercion, were so heavily implicated in the Jim Crow system that used to prevail in the US South, and why the changes in US legislation, such as the civil rights and voting rights acts of the 1960s, brought about unexpectedly rapid and pacific racial integration of workforces in the South. Though Montgomery does not go into this history, he discusses “taste” models of discrimination and of affirmative action. He shows that government policies that give preferences to specified groups will, not surprisingly, usually be efficiency‐increasing when they offset discrimination, but will usually be efficiency‐reducing when they change what would otherwise have been a nondiscriminatory outcome in a competitive market.
The word “usually” need emphasis. Montgomery's chapter shows the richness and variety of the results that have been achieved in economists' studies of this subject. In reading it, one is struck especially by how sensitive the (p.27) impacts of policies can be to such considerations as the existence of covered and uncovered sectors, and the numbers of individuals with pertinent skills in the groups that are favored or not by affirmative action programs. The economy is a general equilibrium system, and any discrimination or preferential policy in a covered sector of an economy changes prices and quantities of the goods that this sector buys and sells. This in turn makes it profitable for firms and consumers in the uncovered sector to choose different patterns of production and consumption than before, and this generally changes the relative demands in the uncovered sector for different sorts of labor. As a result of such forces, it is easily possible that a public policy can have a net effect that is the opposite of the expected effect. Thus, the methodologically individualistic method of economic theory alerts us, as no other theory has done, to important counter‐intuitive possibilities that policy‐making has to take into account if it is to deal in a socially rational way with legacies of segregation and other “group” problems.
6.1 Social Structure and Collective Action
Hardin's chapter—along with his books, One for All and Collective Action18—also helps us understand the origins of social groups and the patterns of exclusion and discrimination that they sometimes engage in. He emphasizes, building on the two just cited books and on Olson's Logic of Collective Action, that social groups cannot organize or act to achieve any group interest unless they can overcome the difficulties of collective action. If collective action, whether in the political system or in the marketplace, provides some benefit to a group, normally everyone in that group will benefit, whether or not they have borne any of the costs of the collective action. Collective action or organization by a group thus inherently involves an indivisibility of the kind described earlier in this chapter.
We recall that this means that the individuals in any large group do not have an automatic incentive to contribute to or act in the interest of their group, and large groups will be able to organize and act in their common interest only if they are able to work out “selective incentives,” or punishments and rewards, that apply to individuals according to whether they do or do not share in the costs of organization and action on behalf of the group. Thus, the groups that act and play a role in society are not all those with some common interest or need for a collective good, but are those lucky enough to be in circumstances that make it possible for them to organize. Hardin points out that we should not take it for granted that this good fortune gives a group a moral claim on the rest of the society.
(p.28) Hardin has explained another aspect of this matter by considering coordination problems in game theory. Obviously there would be great problems if people in a country did not drive on the same side of the road. But they do, even in the absence of any laws requiring it. People rationally follow whatever convention emerges. Often such coordination leads to outcomes that are far inferior to alternatives that might have been chosen, but still much better than would occur without any coordination.
As Hardin has pointed out, there are many spontaneously successful coordination games, and they are typically the source of conventions followed in each social group. Many of the differences in convention across different social groups are arbitrary artifacts of the outcomes of initial coordination games. This means that some of the distinctive cultural practices of particular groups are often not the result of profound moral choices. Nor are they the only practices that could be workable and congenial for this group. Rather, they are artifacts of unimportant or even accidental factors in initial coordination games. Different outcomes of initial coordination games can also come to play a role in distinguishing and dividing different groups.
Some social groups and cultural practices, such as the Indian sub‐caste or Jati, appear to have mainly different types of origin but are still amenable to a methodologically individualistic explanation. As Montgomery's chapter points out, most of the caste groupings take their names from occupations or crafts. Historically, given sub‐castes had a monopoly of practising a given occupation in a given community. Though the Indian caste groups often go back to far earlier periods of history, they are in this respect like the guilds in Europe and many other parts of the world.
Consider the problem facing a group that has a monopoly over a given occupation and wishes to continue to enjoy the monopoly return and pass it on to their descendants. Though Montgomery and Hardin do not go into this, we can see that, to continue enjoying monopoly returns from a cartel, a group must continue to restrict the supply. If the members of such a group can not only pass membership in that monopoly on to their sons, but also offer dowries for their daughters in the form of rights of membership in the cartel to sons‐in‐law, then the number that practice that occupation in the next generation will double even in the absence of any growth of population.
That won't do for any cartelistic organization that obtains its gains by restricting the supply. Thus, endogamy—the rule that marriage must be only inside the group—is not only a universal feature of the caste system, but also something that enables caste‐type groupings to persist for very long periods: it enables the members of a sub‐caste to preserve the value of their monopoly control and to pass it on to their children.19 In the same way, the traditional (p.29) resistance in European royalty to marriages to commoners was similarly necessary to preserve the full value of royal descent. Such practices as these are examples of the general proposition that those in a group that receive a redistribution, so long as they remain powerful enough to continue to receive the redistribution, will be better off the fewer they have to share the redistribution with. When redistribution is at issue, it is always best to be in a “minimum winning coalition.”
The distinctive social customs of different groups that emerge from coordination games of the kind Hardin has explained can make collective action easier. To the extent that people can feel “at home” only with the customs that they have been acculturated to, they have a preference for companionship, marriage, and social interaction within their own group. They then particularly value the respect of those in their group and suffer social loss if they do not have access to social interaction with them. Thus, respect within and access to a group serves as a selective incentive that can motivate collective action by the group.
Though such collective action can be beneficent, Hardin has shown that it is also often exclusionary and harmful. It is, he points out, the mainspring of ethnic conflicts such as those in Bosnia, Northern Ireland, Rwanda, and Somalia. He also shows how “communitarian” ethical philosophies—which do not recognize the problematic nature of the collective action that creates the communal forces that we observe, or appreciate the arbitrary or accidental origins of communal customs—are erroneous.
When one takes Montgomery's and Hardin's work together and combines them with resonant work by others, we see the not‐so‐dismal science from a new perspective. Some argue, we know, that the methodological individualism that characterizes the not‐so‐dismal science is wrong because individuals are greatly influenced by socialization in the groups into which they are born, so that research should begin with groups that socialize individuals rather than with individuals.
But the groups that socialize individuals, we must not forget, are made up of individuals. Also, such groups had to emerge as a result of the interactions of individuals: they could not exist before the individuals who, purposely or inadvertently, formed them. So we can use the broader, integrated approach on display here to study groups and the influence that they have, through socialization, on individuals. The ultimate goal is a theory of general social equilibrium that simultaneously considers the individuals and groups in society and their interactions. Some rational‐choice sociologists, such as James Coleman, have already explicitly endeavored to do this and have made significant progress.20 If the chapters by Montgomery and Hardin and the (p.30) related work discussed above are right, a beginning has already been made to try to explain the origins and characteristics of some kinds of social groups.
7 The State Versus the Market
The remaining two chapters, Robert Cooter's and Pranab Bardhan's, are also concerned with social norms and social groups. But they consider how these norms and groups should affect the respective roles of the private and the public sectors. In different ways, these two chapters bring us back to the long‐standing debate between the right and the left about the proper roles of governments and markets. This is not just the question of how large or small the government should be, but also whether government should be viewed in substantial part as a body that strengthens and enforces norms that evolved first in the private sector—or as the only instrument that can overcome some entrenched impediments to progress in fragmented societies with dysfunctional institutions.
Cooter shows how private‐market interaction can sometimes give rise to certain social norms that are better than those that would arise through an initiative from the central government, so that government at its best sometimes codifies, strengthens, and enforces a social norm emerging from a private sector which shares in the task of enforcing this norm. Bardhan points to impediments to economic development and social dysfunctions that market mechanisms cannot cure because the needed large‐number collective action cannot emerge spontaneously. The difference in emphasis in each chapter may be due partly to differences in the societies and problems analyzed.
7.1 The Spontaneous Emergence of Socially Useful Norms
Cooter is inspired by successful advances in commercial law in relatively competitive and thriving economies. He starts in Chapter 9 with Judge Mansfield's modernization of English common law in the eighteenth century (that is, in the country and the century in which the Industrial Revolution began) and with Professor Llewellyn's effort to identify the best commercial practices in twentieth‐century United States and to write them into the commercial code. Cooter finds that the state built on pre‐existing social norms, so that, as in his title, there is “law from order.” He argues that “in an environment of open competition, business practices tend to evolve rapidly towards (p.31) efficiency. Without open competition, however, harmful business norms can create monopoly power or distort consumer information, and incomplete markets can impose external costs.”
How can a desirable social norm emerge in the private sector? Though Cooter does not emphasize this, such a social norm is a public good. If the numbers involved are large, we should expect that the difficulties of collective action could not be overcome, so the social norm would not evolve under laissez‐faire. Cooter considers cases where principals hire agents, but where in any period an agent can appropriate the principal's investment for himself and thereby obtain much more in that period than a cooperative agent would have earned. In future periods a principal would retain only agents who did not do this, so the offending agent has to seek other principals who might hire him and thus does not earn income in as many periods as a cooperative agent. Drawing on evolutionary game theory, Cooter makes it clear that, if there is an equilibrium in the society with both types of agent, the appropriators and cooperators must in this equilibrium earn the same average rate of return.
Since principals would never knowingly hire an agent that intended to appropriate from them, it pays all agents to claim emphatically that they are cooperators, even if they plan to appropriate. This uniform signaling means that the self‐interest of the agents assures that the norm that agents should cooperate gets a lot of good free publicity!
Cooter also points out that many situations do not lead to uniform signaling. Hard bargainers, for example, may gain from a reputation that they won't give in, so even though society would presumably have lower transaction costs with a social norm that everyone should be a soft bargainer, there is no tendency for such a social norm to emerge.
Where there is uniform signaling, it not only affects speech and beliefs, but also leads, Cooter hypothesizes (through psychological processes that he leaves to the psychologists to analyze), to an internalization of some degree of willingness to act on and help to enforce the norm. Cooter is aware that, when an individual punishes someone who violates a norm, the benefits accrue mainly to others; but he assumes that, when a group has sufficient “coherence,” individuals are sometimes motivated to enforce social norms. He assumes that, in groups with sufficient coherence, externalities of many kinds will at times be internalized, at least if the externalities are “symmetric” ones that leave most individuals as victims as well as perpetrators of the diseconomy.
These conclusions, and the principal–agent and commerical law contexts that Cooter emphasizes, leave the reader with the impression that he is thinking mostly of groups that are not very large. Though he does not go into the numbers issue explicitly, it is clear that if the numbers involved are sufficiently small his conclusions are beyond challenge. But if the numbers involved are (p.32) sufficiently large, Cooter has no alternative but to rely on the psychologists to fill in the gap in rational behavior.
Here the contrast between Bardhan and Cooter is especially instructive. Bardhan explicitly emphasizes not only large numbers but also social fragmentation, and concludes that spontaneous market behavior does not prevent very bad outcomes. Cooter, by contrast, emphasizes contexts where numbers are probably small and finds that in the right circumstances, with uniform signaling and social coherence, useful social norms may emerge.
Still, even when conditions are favorable to the development of useful social norms and efficient business practices, Cooter finds that law and government should often be brought in. When a practice that runs against a social norm is outlawed, the government not only plays an important role in enforcement, but also strengthens private enforcement: individuals are more aggressive in criticizing and punishing violators of social norms when these violations are illegal. Cooter argues that the law works best when, as with the common law in eighteenth‐century Britain or the Uniform Commercial Code in the twentieth‐century USA, it reinforces and even follows social norms. By contrast, when the laws and social norms go in opposite directions, the results will probably be bad. Cooter's most basic point—that the law works best when it is consistent with the outcomes in competitive markets and the social norms of a coherent society—is surely valid.
7.2 Tenacious Institutional Impediments to Development
The last chapter in this collection, Pranab Bardhan's “The Nature of Institutional Impediments to Economic Development,” like all of the others herein, goes beyond the classical analysis of preferences, resources, technology, and markets; it considers governance, institutions, and (especially) the difficulties of collective action as well, and their importance in determining whether there is stagnation or progress.
Bardhan not only illustrates the broader economics, but also takes us back to the substantive issues with which this introductory chapter began. There we challenged a tacit and almost unconscious assumption that characterized much economic (and lay) thinking, at least until fairly recently. This second‐nature assumption was that, unless government intervention prevents it, the markets required for an efficient market economy will arise spontaneously and automatically: that markets are not an artifact of government. We have attempted to show that this assumption, though correct for self‐enforcing transactions, was not true in general because many crucial transactions, such as most of those in the capital market, require third‐party enforcement at least as a last resort.
(p.33) Bardhan also emphasizes that there is no necessity for spontaneous or automatic processes in order to obtain socially efficient or desirable outcomes. Some economists have supposed that large changes in the proportions of the production factors make a relatively scarce factor more valuable, and this in turn motivates the development of better property rights for the more valuable factor. Bardhan points out that this need not necessarily happen. The difficulties of collective action could prevent the realization of the gains that better property rights or other institutional improvements would have brought about. The rationality of individuals need not, because of the infeasibility of collective action or other difficulties, bring about the elimination of even dramatically dysfunctional institutions, much less guarantee social efficiency. Bardhan's starting point is not English commercial law in the century of the Industrial Revolution or the propitious development of the Uniform Commercial Code in the United States: it is the poverty‐stricken millions and social fragmentation in India.
In keeping with his argument that some serious impediments to development do not disappear spontaneously, Bardhan finds simplistic conceptions of the state—such as the theory of the state as simply a predator on the rest of society, or alternatively as just an instrument of domination for one social class over another—to be insufficient. He prefers a more nuanced view that recognizes not only predation and the exercise of dominance, but also (as at times in East Asia) relatively constructive action which promotes prosperity. In keeping with other analyses in this volume, he points out that such constructive action could arise from an encompassing interest.
Bardhan also finds that even the familiar tradeoff between efficiency and equity may in some cases be too simple. The mechanisms for third‐party enforcement of contracts are often poor, so that those who would make productive use of working capital cannot borrow it unless they can provide lenders with very good collateral, such as equity in land. In such circumstances, a less unequal ownership of land could give more cultivators the capacity to borrow working capital, and thus perhaps increase both equity and efficiency.
Bardhan points out how the obstacles to collective action can not only keep spontaneous or market forces from automatically eliminating dysfunctional institutions, but also keep societies from obtaining adequate supplies of public goods. Social fragmentation can increase the difficulties of collective action, and that can mean that (notwithstanding a pre‐existing government) the political and governmental processes do not work in ways that assure an adequate supply of public goods. Bardhan points out that in India there is considerable social fragmentation (and, of course, very large numbers), and that this sometimes keeps even local communities and governments from making adequate provision for such basic needs as elementary education. It is difficult to see how (p.34) anyone could explain the continuing high rates of illiteracy in parts of India without taking account of Bardhan's argument.
8 Theory, Ideology, and Cumulative Research
The issue of what should be the proper roles of the government and the market brings up the insistent question of how the suburbanization of economics relates to the ideological struggles in society. The broader and the more general the ideas, the more often people ask how they relate to the ideologies of the time. So, many people ask: is the broadening of economics an outgrowth of an ideological agenda? Or is it a triumph for one side or the other in the ideological struggle? The opposite ideological implications of the final two chapters hint that the answer should be no. Yet some observers think that at least one of the answers should be yes. Since these questions keep coming up, we must deal with them.
If what we have said so far is correct, the not‐so‐dismal science cannot always favor one ideology or another. As we saw at the beginning, one source of the broader approach is the awareness that some goods cannot be provided, at least not to sufficiently large groups, by voluntary or market mechanisms. The need to accommodate public goods (or, more generally, to analyze market failure) is one of the two main sources of the broadening of economics. Therefore, at least in comparison with narrower or city‐center economics, the not‐so‐dismal science is not single‐mindedly focused only on markets. It begins with the difficulties of collective action and the need for governmental or other collective mechanisms.
The second main source of the broader economics is the recognition that, to understand the growth and distribution of income, the economist must also consider the other side of the coin of market failure: it is usually remedied only by the use of coercive power, and wherever there is power, there is the power to take, and often lots of taking. The proceeds of this taking usually go to those with the power to take, rather than to the relatively powerless poor. Obviously, behavior in markets, which is inherently voluntary, cannot lead to such bad outcomes as can occur from the power to take. In a particular transaction, an individual may be the victim of fraud, and thus be made worse off by the transaction. But rational individuals will never voluntarily participate in repeated fraudulent transactions—you can't sell anyone the Brooklyn Bridge twice. Any voluntary market interactions that continue to occur must make people better off.
(p.35) Not so with the power to take: there is essentially no limit to the damage that this can do. Thus, there is government failure as well market failure, and the consequences of government failure are often incalculably more harmful than anything that can occur from voluntary interaction in a market, however imperfect the market.21 However, since we cannot get along without governments with the power to coerce, the fact that governments have incomparably greater capacity to do harm than markets is not the whole story. As Oliver Williamson's chapter pointed out, we should not condemn a practice or institution as irrational unless we can find a remedy.
Accordingly, when we take the two sources of the suburbanization of economics together, we see that the logic that led to the broader economics is inherently two‐sided and eminently suitable for analyzing both the contribution of markets and market failure: it emphasizes both the need for government and the prevalence of government failure. If devotees of any ideology or political tendency should find the results of good integrated social science work in conflict with their preconceptions, they should ask whether these preconceptions are valid and well balanced.
In any case, the question of whether the suburbanization of economics will favor this or that ideology is the wrong question. What we really need to know about any research results is how much truth, if any, they contain and whether they inseminate new and fruitful inquiries. The professor of theology at a seminary might have the responsibility of coming up with arguments that help the village priest out‐argue the village atheist. But economists and other social scientists (unless they have sold themselves to some interest group or party) have no corresponding responsibility to any of the contending ideologies. Our task is rather to improve the understanding of the realities that individuals and societies confront and thereby to help them find means that are better adapted to achieving the ends that people seek. Many individual researchers do, of course, start with strong preconceptions growing out of one ideology or another, but if they find new and fruitful truths, these truths are not either poisoned or blessed by the creed that inspired them. They are, on the contrary, valuable to us all: they reduce the extent to which illusions govern choices.
8.1 Cumulative Knowledge
Some say that the truth of today is the error of tomorrow, and this is sometimes the case. But, as others before us have pointed out, in science and scholarship the truth of today is much more often the special case of the truth of tomorrow. (p.36) Science and scholarship, whenever they are making much progress, are inherently cumulative: later work generalizes, extends, amends, and improves what was done earlier. There are some lines of literature in the social sciences in which successive writings begin by demolishing or belittling what has gone before, but there is usually not much of value in them. By contrast, the suburbanization of economics and the emergence of an integrated social science are examples of decidedly cumulative science and scholarship: the truths of yesterday are indeed special cases of the truths of today.
The suburbanization of economics does not discredit or contradict anything in the city center. On the contrary, it suggests that the two‐and‐a‐quarter centuries of cumulative work in economics, some of it by men of unquestioned genius, has an even greater value and wider potential than has usually been supposed. The modern economist can say, as Paul of Tarsus did (Acts, 21), that he comes from “no mean city.”
(1) We are deeply grateful to Kimberly Brickell, Brian Steinhardt, and especially Maria Coppola for help in organizing this volume and the conferences out of which they grew.
(2) Though Thomas Hobbes pointed out in Leviathan in 1651 that, in the state of nature, “he that performs first has no assurance that the other will perform after,” few writers notice the dependence of many markets on third‐party enforcement of contracts or distinguish these markets from those with self‐enforcing transactions. The economic historian Douglass North, by contrast, has explicitly distinguished self‐enforcing transactions from those that require third‐party enforcement; see his Institutions, Institutional Change, and Economic Performance (Cambridge University Press, 1990). There is a full analysis of this distinction and an array of econometric tests showing that it is extraordinarily important for the structure of economies, investment, and economic growth in Christopher Clague, Philip Keefer, Stephen Knack, and Mancur Olson, “Contract‐Intensive Money: Contract Enforcement, Property Rights, and Economic Performance,” IRIS Working Paper no. 151, 1995.
(3) The justifications for this assertion and for many of those in prior paragraphs are given in Mancur Olson, The Logic of Collective Action (Cambridge, Mass.: Harvard University Press, 1965), The Rise and Decline of Nations (New Haven: Yale University Press, 1982), and “The Varieties of Eurosclerosis”, in Economic Growth in Europe since 1945, ed. Nicholas Crafts and Gianni Toniolo (Cambridge: Cambridge University Press, 1966), pp. 73–94.
(4) Works of Thomas Carlyle, (New York: Charles Scribner & Sons, 1904), vol. 29, pp. 348–83. Our account is inspired by and draws heavily upon Joseph Persky's “A Dismal Romantic,” Journal of Economic Perspectives, 4 (Fall 1990): 165–72.
(5) To set out the complete logic behind this sentence and the many major qualifications it requires would take us into some complex, lengthy, and fascinating issues that do not have that much to do with our criticism of the no‐free‐lunch theory. In part, the sentence is supported by one of the two basic theorems of welfare economics, which demonstrates that, if a competitive equilibrium exists, it is Pareto‐efficient. In part, the assertion is misleading because it assumes away not only the public goods, externalities, and missing markets that we have already discussed, but also some problems that arise from economies of scale, asymmetric information, and the theory of the second best.
(6) Joel Mokyr, The Lever of Riches (New York: Oxford University Press, 1990).
(7) Oliver Williamson, The Economic Institutions of Capitalism (New York: Free Press, 1985).
(8) Adolph Berle and G. C. Means, The Modern Corporation and Private Property (New York: Macmillan, 1932).
(9) Williamson's argument brings to mind the seminal and now‐classic work on constitutions by James Buchanan and Gordon Tullock, The Calculus of Consent (Ann Arbor: University of Michigan Press, 1962). This book argued that restrictive constitutions that permitted action only when there was much‐more‐than‐majority support would largely prevent such redistributions. Though Williamson does not discuss constitutions, his conclusion that sometimes there is no remedy may suggest that there are limits to what we can expect from constitutional reform. The same opportunism and other difficulties that make many other ostensibly attractive deals or contracts unworkable might also bedevil constitutional construction and interpretation.
(10) Mancur Olson, “Autocracy, Democracy, and Prosperity,” in Richard Zeckhauser, ed., Strategy and Choice (Cambridge, Mass.: MIT Press, 1991), pp. 131–57, which develops the argument about encompassing and narrow interests in Olson's previously cited Rise and Decline of Nations. There are very important early insights in this line of thinking in various unpublished drafts by Martin McGuire in the early and mid‐1990s.
(11) Martin McGuire and Mancur Olson, “Economics of Autocracy and Majority Rule,” Journal of Economic Literature, 34 (March 1996): 72–96.
(12) This paradoxical possibility that self‐interest could make a sufficiently encompassing (“super‐encompassing”) ruling interest avoid any redistribution to itself is demonstrated in the McGuire–Olson article cited above, but was not understood when the paper in this book was written and is not mentioned in it.
(13) Narrow special‐interest groups have considerable influence in most autocratic societies. Guilds, for example, were important even under the absolutist monarchies in early modern Europe, and the Soviet‐type societies in their later years were dense with insider lobbies. Still, dictatorships that are both new and strong may sometimes emerge, and they may repress special interests. This happened for a time in South Korea, Taiwan, and Chile.
(14) J. Bradford De Long and Andrei Shleifer, “Princes and Merchants,” Journal of Law and Economics, 36(2), (October 1993): 671–702.
(15) Mancur Olson, How Bright are the Northern Lights? Some Questions about Sweden (Lund: Institute of Economic Research, Lund University Press, 1990), and the Rise and Decline of Nations.
(16) Mancur Olson, “A Theory of the Incentives Facing Political Organizations: Neo‐Corporatism and the Hegemonic State,” International Political Science Review, 7(2) (April 1986): 165–89; “The Devolution of the Nordic and Teutonic Economies,” AEA Papers and Proceedings, 85(2) (May 1995): 22–7, and in the already cited “Varieties of Eurosclerosis.”
(17) Gudmund Hernes, “The Dilemmas of Social Democracies: the Case of Norway and Sweden,” Acta Sociologica, 34(4) (December 1991): 239–60.
(18) Russell Hardin, One for All (Princeton: Princeton University Press, 1995) and Russell Hardin, Collective Action (Baltimore: The Johns Hopkins University Press, 1982).
(19) Olson, Rise and Decline, ch. 6.
(20) James Samuel Coleman, Foundations of Social Theory (Cambridge, Mass.: Harvard University Press, 1990).
(21) The complete absence of government leads not to voluntary market interaction, but to the private taking of Hobbesian anarchy.