Jump to ContentJump to Main Navigation
Why Some Things Should Not Be for SaleThe Moral Limits of Markets$

Debra Satz

Print publication date: 2010

Print ISBN-13: 9780195311594

Published to Oxford Scholarship Online: September 2010

DOI: 10.1093/acprof:oso/9780195311594.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2017. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see http://www.oxfordscholarship.com/page/privacy-policy).date: 17 March 2018

What Do Markets Do?

What Do Markets Do?

(p.15) 1 What Do Markets Do?
Why Some Things Should Not Be for Sale

Debra Satz (Contributor Webpage)

Oxford University Press

Abstract and Keywords

Economists have written surprisingly little about the nature of a “market,” assuming perhaps that it is a simple concept with a clear or obvious referent. There is, for example, no definition of a market in many of the most widely used economic textbooks. Yet, in reality, a market is a complex institution. This chapter explores the virtues of markets in terms of welfare and autonomy and also points to their limits.

Keywords:   welfare, autonomy, efficiency, externality, endogenous effects

Economists have written surprisingly little about the nature of a market, assuming perhaps that it is a simple concept with a clear or obvious referent. There is, for example, no definition of a market in many of the most widely used economic textbooks.1 Yet in reality a market is a complex institution. As we will see in subsequent chapters, my view of markets is that they are even more complex than the basic account I give here suggests.

To begin, markets are institutions in which exchanges take place between parties who voluntarily undertake them.2 Because all human action takes place within limits—I can’t use my arms to fly simply by wishing it so—“voluntary” cannot mean the same thing as “unconstrained.” All human action is constrained, by external and internal factors. There is a rich and subtle philosophical literature on the nature of voluntary actions, attempting to distinguish them from actions that are unjustly constrained.3 For present purposes I will simply assume that in market exchanges both buyer and seller are entitled to the resources with which they transact, have the freedom to accept or refuse an offer of exchange, and can attempt to make another offer or strike a better deal with someone else.4

Additionally a market is not a single exchange between two individuals; indeed an exchange can be noxious without there being a noxious market.5 Markets coordinate behavior through price signals, and to do this there have to be enough exchanges so that people are able to adjust their behavior in response to the actions and anticipated actions of others. If there are only two goods in the world, then you and I might exchange those goods with each other, but unless there is the possibility of coordination on future exchanges we don’t really have a market, at least as I am using the term here.

The New Shorter Oxford English Dictionary defines a market as “a meeting or gathering place of people for the purchase and sale of (p.16) provisions or livestock” and as “the action or business or buying and selling.”6 But markets are not merely meeting places or a series of individual transactions: they are social institutions that must be built up and maintained.7 Initially markets may be thrown up spontaneously, but in the end they are socially sustained; all markets depend for their operation on background property rules and a complex of social, cultural, and legal institutions. For exchanges to constitute the structure of a market many elements have to be in place: property rights need to be defined and protected, rules for making contracts and agreements need to be specified and enforced, information needs to flow smoothly, people need to be induced through internal and external mechanisms to behave in a trustworthy manner, and monopolies need to be curtailed. In all developed market economies governments play a large role in securing these elements.

For this reason it is mistaken to consider state and market to be opposite terms; the state necessarily shapes and supports the process of market transacting. In Lewis Kornhauser and Robert Mnookin’s memorable phrase, all (market) bargaining occurs in the shadow of the law.8 Transacting individuals depend on the state for their basic security when they walk to the corner store to purchase food for their meals; they expect the state to enforce health and safety requirements concerning food production and handling; and they expect the shop owner to be sanctioned if he fails to keep up his end of the transaction. The fact that laws and institutions underwrite market transactions also means that such transactions are, at least in principle, not private capitalist acts between consenting adults, as the libertarian philosopher Robert Nozick famously claimed, but instead a public concern of all citizens whether or not they directly participate in them.

In addition to specific markets, such as markets in land, labor, or luxury goods like a yacht, there is what is sometimes referred to as “the market system” or the market economy. This further abstraction is usually taken to refer to a “society wide coordination of human activities” through mutual transactions.9 Some people also use the term to refer to the integration of markets with “private property in the means of production.”10 But markets can coordinate behavior under very different property rules. I will use the term market in the context of discussing specific types of exchange transactions and market system as the abstraction that is supposed to link the set of all such markets. One important (p.17) argument of this book is that in order to understand and fully appreciate the diverse moral dimensions of markets, we need to focus on the specific nature of particular markets and not on the market system.

Market Virtues

It is difficult to understand how a market system or any particular market works. Like ants in a colony, individuals cooperating in a market “have no dictators, no generals, no evil masterminds. In fact, there are no leaders at all.”11 The participants in a market are not obligated to follow another’s orders with respect to what they buy and sell. Through markets individuals coordinate and mutually adjust their behaviors without relying on a conscious organizer to bring about the coordination. Somehow a market order arises out of millions of independent individual decisions, although such decisions are supported, as I stressed earlier, by an array of government and nongovernment institutions. Nevertheless the fact that coordination occurs largely through individual decisions and not through a central command and control structure explains and supports two particular virtues associated with markets, at least when they are working well: their link to efficiency and their link to liberty. Let us consider each of these virtues in turn.


Market transactions link multiple chains of trades and involve cooperative behaviors spanning the globe. To give an example, workers in India whom I will never meet assembled my cell phone using materials imported from Africa and ordered on the Internet from suppliers, and the phone was transported to me by the employees of a transnational shipping company. Through the use of prices, markets signal what millions of goods are worth to sellers and buyers and intermediaries who will never meet each other. In doing so they function to mete out resources efficiently, indicating to sellers what and how much to produce, to consumers what price to pay, and to investors where to lay down their capital. Because rational individuals will exchange with one another only when they have (p.18) something to gain, markets will (ideally) purge the economy of less desirable goods and move the trading parties to their most preferred positions, given their resources. The continual adjustment of supply and demand, registered in changing prices, allows markets to “clear” what has been produced. When inventory is cleared, there is no excess demand or excess supply: supply equals demand at some price.

A set of remarkable theorems formalizes the link between markets and efficiency. The first is the so-called fundamental theorem of welfare economics, according to which the result of any market equilibrium under perfect competition is Pareto optimal.12 A social state is described as Pareto optimal if and only if no one’s position (measured in terms of their preference satisfaction) can be improved without reducing the position of someone else. The intuitive idea behind the theorem is that people will engage in mutually beneficial exchanges and continue doing so until they cannot improve their positions by exchanging further. When all exchanges cease it is because an optimal allocation has been reached. Once that point is reached any deviation will make at least one person worse off.

A second formal result proves the converse proposition, that every Pareto optimal social state is a perfectly competitive equilibrium for some initial distribution of resources. It is worth keeping in mind that there is typically more than one Pareto optimum for any economy; in addition, given different starting distributions market competition will yield different results. This theorem allows that radical change from the status quo can still be efficient; it suggests that we can always find some initial distribution of resources that, along with the use of a market, will support a given Pareto optimal (efficient) social state.

These two results have intuitive ethical appeal. With respect to the first theorem, it seems obvious that it is better to make people better off and that if one of two prospects is better for someone than the other, and at least as good for everyone else, then it is better.13 Yet although these efficiency results may be powerful in certain respects, they are actually of limited significance from a normative (ethical) point of view. Paretian efficiency does not give us overriding reasons for using markets or overriding reasons against interfering in them. As Amartya Sen notes, “A state can be Pareto optimal with some people in extreme misery and others rolling in luxury, so long as the miserable cannot be made better off without cutting into the luxury of the rich.”14

(p.19) We have good reasons to care about more than Paretian efficiency in our assessment of markets. For example, we have reasons to care that the initial distribution of resources in society is fair. Indeed if you think that individuals are entitled to certain property rights—by considerations of justice—then the fact that a certain social state is efficient relative to a different distribution of property rights has no normative force for you whatsoever. This is why objections to slavery are not undermined if it turns out that a slave system is Pareto efficient (insofar as any change in distributive allocations would make the slave owners worse off).

The second theorem might seem to help here since it allows for the incorporation of the distributive justice objection. If a critic doesn’t like a particular Pareto equilibrium she can always redistribute initial resources the way she wants—abolish slave ownership, for example—and then allow competitive markets to produce another Pareto optimal result. Of course arranging for the redistribution is another matter.

In practice it is very difficult to find policy interventions that do not make at least one person worse off. Consider policies to promote the building of roads, hospitals, bridges, or schools. Somebody almost always prefers that these tasks not be undertaken; for example, a new highway benefits some businesses but hurts others located along the route of the older road. Nonetheless there may be good reasons to build the road. For this reason many economists prefer to think about efficiency in ways that allow the costs to some to be compensated by the extra gains to others. We can define a social state R as a potential Pareto improvement over a social state S if the winners in R could compensate the losers in R and still retain something over and above what they would have had in S. This idea of efficiency is sometimes referred to as Kaldor-Hicks efficiency, and it is effectively a form of cost-benefit analysis. Cost-benefit analysis tells us to adopt the policy (e.g., to build or not build the new road) that has the largest net benefit, other things being equal. However, we should bear in mind that a policy with the greatest net benefit may in reality fail to distribute some of that benefit to the losers, and thus this form of efficiency (unlike Pareto efficiency) can wind up endorsing policies that actually make some people worse off!

Although Kaldor-Hicks efficiency is a more useful concept than Pareto efficiency to use in evaluating economic policies, given that so many exchanges produce both winners and losers both concepts are still normatively narrow ways of assessing economic achievements. Both (p.20) employ criteria that omit consideration of such issues as what is a fair distributive outcome. Indeed the development of these concepts of efficiency was partly motivated by the desire to separate the study of what economists saw as uncontroversial economic improvements from the more controversial questions of ethics and distributive justice.

I believe that such a complete separation is in fact impossible. For example, the acceptance of the Pareto criterion as the measure of economic improvement depends on a key normative assumption: that improvement is to be measured in the space of individual preferences. That is, on this view of efficiency, people are considered better off the more that their own (consistent) preferences are satisfied. Additionally this criterion was formulated to bypass interpersonal comparisons with respect to different individuals’ preference satisfaction since such comparisons are considered meaningless because there are “no means whereby such comparisons can be accomplished.”15

But surely not all preferences are equally worthy of satisfaction. First, some preferences are really urgent needs, whereas others are altogether frivolous. It is surely more important to satisfy the needs of those in extreme misery in Sen’s example than to add more to the coffers of those already rolling in luxury. The fact that income transfers to the poor would make the wealthy worse off does not settle the case against such transfers. Second, some preferences, such as the preference for hurting others, would be accorded no weight at all from a moral point of view. Is it really an improvement if, all things being equal, the slaveholder’s preference for more slaves is satisfied or the sadist’s preference for inflicting pain?

For these reasons most political and moral philosophers (indeed most people) use criteria for assessing social policies that go beyond Paretian and even Kaldor-Hicks efficiency. They appeal to fairness as well as to conceptions of human well-being that allow us to compare the benefits and costs of different policies to different individuals. In comparing people’s well-being we might be led to decrease the preference satisfaction of the millionaire to satisfy the urgent needs of the desperately poor. Indeed we might be led to reject preference satisfaction as the right metric for making and assessing interpersonal comparisons and for evaluating economic states of affairs. (Later in this book I discuss in more detail the limitations of focusing on preference satisfaction as a standard for assessing markets).

Nevertheless the efficiency theorems do give us some insight into the individualistic basis for the mutually advantageous nature of trade. (p.21) Individual decisions function, in the context of markets and prices, as signals for coordinating action to satisfy maximally agents’ wants under given sets of constraints. In a market’s best-case scenario, where information flows, there are no third-party effects of exchanges, no monopoly power, and the parties are completely trustworthy, the network of individual trade serves to generate improvements in getting people what they want. It thus produces efficiency relative to those wants; it limits waste and uses human and nonhuman resources efficiently. However, in real-world scenarios we cannot automatically conclude that the market is more efficient than alternatives. In almost all actual market contexts there are problems with information and enforcement that mean that intervention can improve on efficiency, a point to which I will shortly return.


From a normative point of view, one of the key attractions of markets is their relationship to individual choice and decision. Markets:

  • Present agents with the opportunity to choose between a set of alternatives (partly by providing individuals with incentives to create the material wealth which is a precondition of having an extensive array of choices)

  • Provide incentives for agents to anticipate the results of their choices and thus foster a kind of instrumental (means-ends) rationality

  • Decentralize decision making, giving an agent alone the power to buy and sell things without requiring him or her to ask any one else’s permission or take anyone else’s values into account

  • Place limits on the viability of coercive social relationships by providing (at least formally) avenues for exit

  • Decentralize information, thereby making abuses of power by authorities less likely

  • Allow people to experiment, to try new commodities, to develop new tastes, to opt out of traditional ways of life

  • Contribute to the undermining of racial, ethnic, and religious discrimination by appealing to the reciprocal self-interest of individuals in exchanging goods with one another and by fostering anonymous exchange

(p.22) Liberal theories that assign substantial weight to individual freedom thus tend to allot a central role for market allocation, pointing to the market realm as a place where the capacities for individual choice, indeed where the liberal individual herself is developed. Markets call up our powers as individual decision makers who can veto as well as sign on to exchanges, and they give scope for the exercise of these powers. In this sense markets can be instruments for promoting freedom: they develop our capacities to choose. Additionally markets can be components of freedom. As Amartya Sen has noted, the freedom to engage in transactions with others, to decide on where to work, what to produce, and what to consume, are important parts of a person’s overall freedom.16 Choosing often has an intrinsic value; many of our actions have a special meaning for us precisely because we chose them. Think about buying a birthday gift for a devoted friend. Even if I could hire someone to make the choice and purchase for me, I may want to do it myself as a way of expressing and communicating my own feelings. Even if a well-designed computer program allotted people into careers that matched their talents, this would be quite different from allowing people to choose (perhaps with less happy outcomes) their own occupations. Many of us want our own values and judgments to be reflected in what work we do, what we consume, and which of (what Max Weber termed) the warring gods we serve in how we live.

Many political and social theorists have valued markets precisely because they believed that markets assist in the development and exercise of our capacities as individual decision makers. For even if, as Locke and Rousseau thought, we are born to a state of freedom, it is widely recognized that to develop and realize various freedoms requires education, planning, practice, and cooperation with others. The development of the free individual is in fact a tremendous social achievement. Markets have had an important role to play in facilitating freedom’s achievement by stimulating the capacities we need to choose and providing these capacities with a wide arena for their employment.

Reliance on markets for the distribution of goods and services can also be an important way of respecting individual and divergent values. Two people do not have to agree on the importance of a good, or its place in a worthwhile life, in order to exchange that good on a market. Think of the buyer and the seller of a religious text such as the Bible. Buyer and seller may disagree radically as to the Bible’s importance as well as about the (p.23) appropriate attitude a person should take to the Bible, but they can still agree on its price. In a market system there is no preordained pattern of value to which individuals must conform; markets allow people to make their own judgments about what they want to buy or sell, how hard they want to work, how much they want to save, what they value and how they value it, and what they wish to consume. Indeed the market system institutionalizes the idea that, potentially, anything might be traded for anything and anyone might enter into the great trading game.

In a justly famous passage in The Communist Manifesto Karl Marx celebrated this cosmopolitan and liberating character of a market system:

All fixed, fast frozen relations, with their train of ancient and venerable prejudices and opinions are swept away, all new-formed ones become antiquated before they can ossify. All that is solid melts into air, all that is holy is profaned, and man is at last compelled to face with sober senses, his real conditions of life and his relations with his kind…. In place of the old wants, satisfied by the production of the country, we find new wants, requiring for their satisfaction the old products of distant lands and climes. In place of the old and national seclusion and self-sufficiency, we have intercourse in every direction, universal interdependence of nations.17

True, Marx was ambivalent about the liberating effects of the market system—he thought that under capitalism too many of those who work were under the subjection of their employers and limited by their own poverty—but as this passage makes clear, he also saw the potential for markets to link people together in a fundamentally new way, in opposition to the “venerable prejudices” that had previously bound people in traditional “fixed and frozen” roles. The idea that markets place people in new social relationships with one another—relationships that are horizontal, egalitarian, and anonymous—is a theme sounded by the market’s earliest defenders as well as by its detractors.

Sometimes it is thought that the type of freedom that markets support is essentially negative freedom, freedom from interference by others. In the marketplace the consumer is held to be her own “sovereign,” not subject to anyone else’s authority. (As I noted, this is literally false: markets always depend on property rules, enforced through public coercion, that interfere with some individual liberties. If you own the car, then I am not simply free to use it. Ownership of real estate and land, likewise, puts (p.24) enormous restraint on people’s freedom of movement. But markets also can support a more positive kind of freedom, the freedom to be in control of one’s own life, by reducing servile dependency and undermining hierarchical social relationships. Adam Smith singled out this feature of markets as their “most important” effect, along with “good government”:

Commerce and manufactures gradually introduced order and good government, and with them, the liberty and security of individuals, among the inhabitants of the country, who had before lived almost in a continual state of war with their neighbours and of servile dependency upon their superiors. This, though it has been the least observed, is by far the most important of all their effects.18

Under feudalism wealthy landlords employed hundreds of retainers, servants, and peasant farmers, all of whom depended on them for both their subsistence and protection.19 By contrast, Smith points out, commerce and manufacturing liberate individuals from such degrading servility because in a well-functioning labor market, no one is dependent on any one particular master. Any worker can, at least theoretically, move to another employer in the event of humiliating or arbitrary treatment.20 And in a competitive market no single person has the power to set prices: prices depend on the choices of all.

Of course, it is important not to overstate this contrast between market freedom and feudal dependence; many laborers did and still do have to obey an arbitrary master on the factory floor. Bosses wield power over their employees that these same employees do not wield over employers.21 But two features of competitive labor markets work to temper the degree of humiliating servitude that workers face on the job.

The first mitigating feature is that market relationships are impersonal relationships based on mutual self-interest. As Smith reminds us, “It is not from the benevolence of the butcher, the brewer or the baker that we get our dinner, but from their regard to their own interest.”22 The motivation of self-interest that fuels a market differs from those motivations involved in the exercise of an arrogant and personal power. In Albert Hirschman’s words, in a market society “passion” tends to be tamed by “interest.23 Under the pressure of competition the motivation for abusing and lording it over inferiors and the temptation of unleashing volatile emotions such as vengeance, honor, and envy have to be disciplined by the need for productive efficiency.24 Moreover markets (p.25) link anonymous strangers, people who have no personal relationships with one another, and therefore no personal axe to grind.

The second mitigating feature of competitive labor markets is that they allow, to varying degrees, for the possibility of exit. The need actually to enlist loyalty, commitment, and accountability on the part of their employees gives employers a reason to mitigate the power that they might otherwise wield. Exit is a powerful influence on the shape of human relationships and interactions. In many circumstances the mere threat by a person to exit a relationship may lead others to consider her interests more carefully and to treat her better.

Employees also exit from their employers when they leave their job, in contrast with feudalism; they go home to a realm in which their employer is not assumed to have authority. Feudalism gave the owners of land (the lords) the rights they needed to exercise direct control over the people who lived on their land (the peasants), including the right to punish them and to give them orders to go to war with neighboring landowners. Although there have been “company towns,” capitalist managers do not ordinarily give orders to their employees outside of their working hours and have little direct control over non-work-related aspects of an employee’s life. In developed capitalist economies residence is largely separated from work, although as we shall see later in this book, in some parts of the developing world such exit can be foreclosed by the shape of the labor market itself.25

Of course, much of the curtailment of employers’ arbitrary and abusive power was achieved not only through the employers’ own prudent decisions about the requirements of maximizing productivity, but also, and perhaps especially, through the advent of labor unions. A critical function of unions on the factory floor has been to protect the freedom and equality of workers by providing a counterweight to employer power.

And even though markets can be seen to promote independence and individual freedom we should not lose sight of the fact that they can also coexist with political regimes that deny or curtail basic political freedoms. Finally, those who fare very badly in the market system—who hold down personally unrewarding jobs for little pay, have no viable alternatives with which to support themselves, lack information, and so on—might reasonably claim that they have only a minimal and degenerate form of freedom.

Nor are markets the only route to personal freedom and independence. A person can experience important freedoms within a nonmarket context, (p.26) such as when she participates in a collective political endeavor or shares in a project with her friends and family. Many of our important collective and individual freedoms do not rely directly or even indirectly on a market. Indeed some of these freedoms, such as the freedom to participate in a tight-knit homogeneous community or to be able to escape competitive interactions with others, may be effectively undermined by the existence of a market.26 Nor is there any guarantee that all of the freedoms that markets enable will be meaningful freedoms; freedom from servitude and abuse is crucially important, but having the opportunity to choose between dozens of toothpaste brands does not significantly advance a person’s freedom.

The Background Conditions for the Market’s Link to Efficiency and Freedom

Markets do not automatically or spontaneously realize the virtues of efficiency or freedom. For markets to promote these values, there has to be a suitable platform in place. Theorists from Adam Smith to David Hume have recognized that economic activity presupposes property, rules of exchange, and contract and enforcement. Moreover different platforms will have dramatically different effects on the compatibility between markets and the values of freedom and efficiency. In other words, a positive relationship between particular markets and the values of freedom and efficiency is contingent: it depends, at least in large part, on the platform on which markets are erected. I describe in generic terms the most important elements of this platform below.27

Property Rights

Markets work efficiently only where there are established and protected property rights. This requires the existence of legal and regulatory frameworks to ensure that contracts are enforced and the given property rights are respected. But functioning markets require that the state do more than simply intervene to prevent theft and fraud. There also need to be mechanisms for resolving commercial disputes; there has to be a sound banking system that provides businesses with access to credit; (p.27) and there needs to be a system of taxation to pursue necessary collective goals such as education, building and maintaining infrastructure, and the administration of justice.

Property rights are also relevant for the real freedoms a person can achieve. For example, a market in which some people can be owned limits the freedom of those who may become the property of others. A market that leaves people with few social entitlements may undermine the ability of the poor to achieve important substantive freedoms. Even if, to paraphrase Anatole France, in a market system the poor and the rich are equally free to dine at the most expensive restaurant in New York City, this freedom is not worth very much to the poor. Before a person can be said to have the effective opportunity—the real freedom—to be and do many things, she must have access to a number of goods that markets may or may not provide. A person may be unable to participate in collective decision making, achieve a kind of personal independence, or even function as a market agent if she is hungry or illiterate or cannot escape premature morbidity.

More generally all property rights enable certain freedoms and place limits on other freedoms. Some private property rights endow individual owners with exclusive authority over their property and thus simultaneously exclude all others.28 In addition all property rights are the products of laws and conventions that back them up and enforce them.29 My ownership of a good means little if I am powerless to prevent others from seizing it. An important implication of this observation is that the free market is necessarily based on the coercive power of property rules, government regulations, and social conventions. True laissez-faire is not even logically possible.

Free Information

Unless buyers know what commodities are selling for, they may overpay for them. If a seller controls price and product information, then buyers can be misled into buying a shabby product, and there is no incentive for the seller to cut his prices. In the presence of reliable information, an exchange that looked to be in the buyer’s benefit might turn out to have been a mistake. Efficiency requires that decisions be made with adequate information on benefits and costs.

(p.28) Information does not always flow freely in a market. It is costly to come by, it takes time and effort to learn what goods are available and what their prices are, and it is even more costly to determine their quality. And both buyers and sellers have an incentive for holding on to information to increase their own market power. To ensure the flow of information requires many institutions, conventions, regulations, and norms. Services such as the Yellow Pages, Google, and company trademarks lower the costs of finding information for consumers. Wholesalers and trading companies lower costs of information to businesses. Government regulations attempt to assure quality control and accurate information about products.30 Nevertheless in some exchanges large asymmetries of information are likely to remain between buyers and sellers; examples include health care markets and the market for used cars.31 In these cases it is hard for buyers to ascertain the quality of the goods that are for sale. As we shall see, child labor markets are also often closely tied to poor information.

Even when information on products is available people are notoriously bad at processing it; they regularly distort the probabilities of risks associated with different products, and they are easily overloaded by too much information. For example, even reasonably informed people may choose to downplay the risk of cancer due to smoking because, although knowing the statistics, they do not see cancer as something that can happen to them. Recent work on biases in decision making has demonstrated that people routinely overestimate the importance of nominal losses, overestimate their probability of success, and respond to “framing effects” in the ways that decisions are posed.32 In such cases biases mean that it may be possible to improve on market outcomes through some kind of intervention (educational campaigns, changing default starting points, marketing). There is no invisible market hand automatically producing efficient outcomes; as Joseph Stiglitz has remarked, Adam Smith’s “invisible hand” is invisible because it is not there.


Markets function well only when the participants are trustworthy. Because in many transactions there is a time lag between purchase and sale, buyer and seller depend on each other to honor their agreements. Because obtaining information and monitoring are costly, markets are (p.29) more efficient when the parties do not aim to deceive one another. This means that although it is often said that markets are fueled by a maximizing self-interest, they must also be underwritten by social sentiments and norms. Homo economicus may be out only for himself, but he must not generally steal, lie, cheat, or murder in order to maximize his gains if markets are to work.33 Theft involves an exchange of goods, but clearly it is not a market exchange.

Interestingly markets have different and opposing effects on the possibilities for trust and trustworthiness in a society. On the one hand, to the extent that a trustworthy reputation is important to market success, markets encourage intelligent pursuit of interest over reckless passion.34 When one party behaves in an untrustworthy manner, other parties may refuse to trade with him in the future. Knowing this, it is not in his self-interest to default on his contractual agreements. In this way self-interest can serve as a basis for mutually beneficial behavior. On the other hand, however, the possibilities for trust depend on several factors that are themselves affected by markets. People seem to be more likely to trust those with whom they repeatedly interact, with whom they share beliefs and values, and with whom they are able to engage in direct communication. Markets negatively affect all of these factors by increasing the number and heterogeneity of trading partners.35 The anonymous nature of market exchanges tends to favor short-lived exchanges and a pairing of individuals that is more random than in a small community of friends. As the number and heterogeneity of trading partners increase, the monitoring and enforcement costs also increase, and self-interest becomes a less reliable basis for producing socially good results. Although markets enable participants to economize on virtue, those exchanging cannot economize too much.


An efficient market needs to keep the tendency to monopoly in check; in particular, competition is necessary for the two theorems of welfare economics to hold. In perfect competition no one has any power over anyone else, all are assumed to act independently of one another, and no one can determine prices. Competition thus disciplines companies; they must produce high enough quality products at low enough prices (p.30) to stay ahead of their competitors. Monopolists face no such incentives; they can persist in offering shoddy products at inflated prices, and they can impose arbitrary prices because there are no alternatives. To forestall the formation of monopolies societies must sometimes rely on antitrust legislation, laws against price fixing, and regulations regarding mergers and takeovers.

Even with such measures many markets are not perfectly competitive. Economies of scale convey advantages in production that lead large producers to corner the market. Some industries are natural monopolies in which it makes little sense to have multiple suppliers. For example, there would be greater social costs to have two water systems running in parallel than having just one, given the costs of digging pathways that go to the same place.

Important freedoms are also undermined by the absence of alternatives. Under monopoly buyers cannot get what they want from many sellers, and in cases of needed goods they are completely dependent on one supplier. Consider the power of the person who owns all the water in a desert. Monopoly is a particular form of compulsion, re-creating a feudal relationship of dependency right in the heart of a liberal market society.

In sum, well-functioning markets require supports. Such supports are not all or nothing, but plainly admit of degrees. The great majority of actual markets lie somewhere in between the textbook extremes of perfect competition and pure monopoly. Individuals come to the market with very differing assets and differing knowledge about alternatives, which can make some parties far more dependent on the transaction than others.

State regulations, redistribution, and widespread acceptance and use of norms such as sympathy and honesty can bring markets closer to their ideal conditions. For example, the state can enforce property laws, curtail monopolies, regulate communications systems, and underwrite compulsory education. Yet even if these four props are in place—even if there are established property rights, free information, trust, and competition—markets can still fail to be efficient or realize liberal freedoms. And even if they do support efficiency and liberal freedoms we may still find markets in some goods unsettling. I will postpone detailed discussion of how markets can fail to link to freedom until a later chapter, when I discuss how labor markets can be compatible with extreme (p.31) servitude and dependency.36 And I will postpone discussion of how even efficient and freedom-enhancing markets can nevertheless be problematic until chapter 4, when I discuss markets in specific goods like safety and education. I conclude this chapter by focusing on the main contemporary economic concern with markets: their efficiency. Why does the link between markets and efficiency sometimes fail, even when good supports for the market are in place?

Market Failure

It is well recognized in economics that market transactions can sometimes impose costs on uninvolved third parties. These costs are usually referred to as “externalities,” and they form the core of the economist’s theory of market failure. As an example, consider that the effects of pollution cannot be restricted only to the parties whose exchanges produce it. Many of the world’s greatest environmental problems today are due to the external unpriced effects of increasing industrial production and fuel consumption. Likewise the sales of international weapons can spill over to have effects on people who are far removed from the parties to the transaction. Other bases of market failure include non-zero transaction costs and technologies that give rise to economies of scale, making only monopolistic or oligopolistic firms viable, as well as the existence of natural monopolies.

When markets fail because of externalities it is because there are some costs that have been introduced that individuals acting in the market have not accounted for. Some of these costs may actually be beneficial—public goods and not public bads—but the ones that concern us are usually not. The production of public bads as a byproduct of market exchanges forms the basis for the economic case for their regulation.

At one time economists proceeded as if externalities were unusual, and the rule was that most transactions had little effect on the individuals who were not direct parties to the exchange.37 But a little reflection will show that this assumption is mistaken. Almost any exchange in a dense, interdependent, and complex society is likely to impose a cost on third parties. Building high-rise apartment towers block the sunlight for neighboring houses. Cars bring congestion. Cigarette smoke circulates. In fact whenever I have preferences over your actions or their effects we (p.32) also have an externality. If I disapprove of a particular religious text because I despise that religion, then your buying or selling this text generates an externality for me, a negative cost that I must now absorb.38

In practice economists tend to be quite opportunistic as to where and when they invoke the concept of externality.39 Indeed they usually appeal to externalities as a basis of regulation in ways that track the traditional “harm principle” of liberal theory, according to which the bare fact that I do not like a certain outcome does not constitute harm, that is, a genuine cost to me that calls for redress.40 But nothing in economic analysis generates or supports this particular interpretation of costs or harm; the economic argument for identifying inefficiencies in the case of only certain externalities—pollution but not intolerance of religious diversity—feeds off moral theory done elsewhere.41 That’s not necessarily a problem, as long as we attend to the moral theory and make it explicit in our understanding of inefficiency.

Markets can also fail to provide needed public goods, where these are understood to include goods (such as national defense) that provide positive externalities, are nonexcludable, and are costly to produce. In such cases, although it is to everyone’s benefit that the good be provided, it is in no one’s individual benefit to provide it. If national defense is provided it will benefit all those who live in a country, even those who do not pay their share of the costs of maintaining it. Many goods are purely or partially public in nature. (And sometimes we face decisions about whether to consider a good a public or a private good. Although education is often treated as a public good, it could be treated as a private good.) Of course even if markets generate inefficiencies due to externalities, the alternatives might be worse. Perhaps some market inefficiency is preferable to a lot of government regulation, with its slow, clumsy, and lumbering bureaucracy. That is why market failure generates only a prima facie case for intervention, not an-all-things considered case.

The logic of the economic approach to markets leads us to view market failure as an indicator not that the market’s system of allocation is defective, but as a sign that the market system is not complete.42 If the scope of the market could be enlarged to include the external third party effects—if sunlight, congestion, pollution, secondhand cigarette smoke, and religious distaste could be priced and sold—then the externalities could be reabsorbed. A complete market, universal in scope and across all future temporal states of the world, promises in theory to eliminate (p.33) all externalities. Indeed much economic reasoning is at least theoretically imperialistic about the range of the market. In the standard Arrow-Debreu general equilibrium models, for example, there is assumed to be a market for every conceivable good, present and future, and every conceivable circumstance.43

Economists’ response to the inefficiencies of actual markets suggests that they have some independent normative commitments and beliefs—a belief, for example, that the market’s inefficiency costs will turn out to be less burdensome than the intrusions of state regulation, and the assumption that third-party cost is defined by only certain kinds of losses. It is open to any of us to endorse a different and more complex view of the concept of market failure.

Looking Ahead

To this point I have stressed the idea of markets as economic and social mechanisms for setting prices, coordinating behavior, and promoting individual choices. As we have seen, contemporary economics offers some powerful arguments in favor of the market mechanism. Markets are often (but not always) better in a technical sense than alternatives, superior as an outcome (in terms of individual preferences) for everyone involved. Markets help develop and give range to individual choice and decision. This chapter explains and defends (in part) these arguments. But it also cautions us to not treat these arguments as a priori. Markets are not necessarily better at promoting these values than alternatives, including, in many instances, in-kind redistribution by the state. To evaluate markets and their alternatives we need to examine messy empirical cases.

The economic arguments in favor of markets proceed without attaching any independent moral value to the commodities being produced and exchanged. It doesn’t matter whether the goods on the market are bibles, guns, butter, human organs, “blood diamonds” that fuel bloody civil wars, or sex. Nor is the quality of the goods relevant. It all looks the same in the economist’s equations. As Lionel Robbins explained in 1932, economics deals with the ubiquitous elements of scarcity, means, and ends, and the means and ends can be filled in with any content whatsoever.44 All markets are explained in the same terms.

(p.34) Moreover market failure is understood in the same terms in all of these different cases. Rather than address questions of ethics, most economists purport to employ a division of labor whereby they explain only the economic consequences of the use of particular markets for efficiency while others worry about ethics. But, as I have argued, such a division of labor is impossible: what counts as an inefficiency or an economic improvement involves prior ethical judgments. For if the only resource we have for thinking about efficiency is subjective preference, then we will have to count dissatisfactions based on envy at another’s success as economic costs. But this seems ludicrous. It follows that any plausible measure of the costs of various activities presupposes a substantive conception of what is important to human welfare, of which subjectively felt harms count as costs. Efficiency turns out to have a moral dimension after all.

In this book I will argue that neither standard efficiency analysis nor the generic concept of market failure can tell us when we should use markets to allocate particular goods and when other mechanisms are more appropriate. Let me anticipate my discussion in the coming chapters with a few simple examples.

Consider the vote. As James Tobin notes, “Any good second year graduate student could write a short examination paper proving that voluntary transactions in votes would increase the welfare of the sellers as well as the buyers.”45 But no one seriously proposes that we distribute a society’s votes through a market; the legitimacy of the political process rests on the prohibition of such transactions.

Consider the labor market. Should employers be allowed to demand sexual favors in compensation for a higher wage?46 Should individuals be allowed to sign slavery contracts with one another? Both quid pro quo sexual favors and slave contracts are widely held to be reprehensible. The interesting question is why this is so and whether efficiency or the standard analysis of market failure is in any way at issue.

Military service is often viewed as a civic duty and something to be praised when undertaken. At the same time, the hiring of mercenaries is widely condemned. Why do people condemn an act when done for pay that they would praise if done for duty?47

A central thesis of this book is that we must expand our evaluation of markets, along with the concept of market failure, to include the effects of such markets on the structure of our relationships with one another, (p.35) on our democracy, and on human motivation. Even if markets in sexual favors or votes or mercenaries turned out to be efficient, and even if they arose from voluntary agreements, such markets might still be objectionable—would be objectionable, I shall argue—insofar as they arise from weak agency, exploit the underlying vulnerabilities of the most vulnerable, or have extremely harmful consequences for individuals or their societies.

In the next two chapters I explore alternative frameworks for thinking about markets. In chapter 2 I present the neglected and rich approach of the classical political economists. Whereas contemporary economics has tended to think of markets in very abstract terms, the classical economists saw markets as heterogeneous, and they sharply distinguished between markets in land, labor, and capital. Their assessment of different markets explicitly called attention to the structure of power and to the effects of markets on human motivation, human capacities, and social relationships. This tradition has been neglected in economics, and I argue that we have much to learn from it. Chapter 3 examines some contemporary egalitarian frameworks for considering the role of the market and its moral limits, including those of Ronald Dworkin and Michael Walzer. In chapter 4 I present and defend my own view of these limits. (p.36)


(1) . There is, for example, no definition of a market in Hal Varian’s widely used Intermediate Microeconomics. It is striking how little discussion there is about the features of a market in economics, despite the fact that this is the central institution that economists study.

(2) . Friedman, Capitalism and Freedom, 13–15.

(3) . See Nozick, Anarchy; Wertheimer, Coercion.

(4) . I consider the question of voluntariness in more detail in chapter 8, in my discussion of bonded labor.

(5) . See Kanbur, “Obnoxious Markets,” 42.

(6) . New Shorter Oxford English Dictionary, 1699.

(7) . Might markets have a natural basis? Adam Smith thought so. So did Hayek. But both thinkers were also well aware of the market’s evolved dependence on rules of property, social conventions, and state enforcement.

(8) . Mnookin and Kornhauser, “Bargaining.”

(9) . See Lindblom, The Market System, 4.

(10) . A. Buchanan, Ethics, 2.

(11) . Gordon, Ants at Work, x.

(12) . This concept is named after Vilfredo Pareto, an Italian economist who used the concept in his studies of economic efficiency and income distribution. For discussion, see Amartya Sen, On Ethics and Economics (Oxford: Basil Blackwell, 1987), 30.

(13) . “Obvious” does not mean true. For criticisms of the Pareto principle, see G. A. Cohen, “The Pareto Argument”; Philips, Which Equalities Matter? See also G. A. Cohen, Rescuing Justice and Equality.

(14) . Sen, On Ethics and Economics, 32.

(15) . Lionel Robbins, quoted in Sen, On Economic Inequality, 81.

(16) . Sen, Development as Freedom, 25–30.

(17) . Marx, “The Communist Manifesto,” 476.

(18) . Smith, An Inquiry, vol. 1, book III, chap. 4, paragraph 4, p. 412, my emphasis.

(19) . This dependence was secured by both the lord’s possession of force as well as an ideology that made hierarchical relationships seem natural.

(20) . See the discussion of Smith’s view of markets in chapters 2.

(21) . See Putterman, “On Some Recent Explanations.”

(22) . Smith, Wealth of Nations, 27.

(23) . For discussion, see Hirschman, Exit.

(24) . But see Marglin, “What Do Bosses Do?,” for arguments about the persistence of hierarchy in the capitalist firm.

(25) . I discuss bonded labor arrangements in chapter 9.

(26) . See Marglin, The Dismal Science.

(27) . See McMillan, Reinventing the Bazaar. I am indebted to McMillan’s excellent discussion of the background institutions and norms presupposed by the market system.

(28) . Some property rights are distributed among different individuals or entities.

(29) . See also the discussion of the conventional nature of ownership in Murphy and Nagel, The Myth of Ownership.

(30) . See McMillan, Reinventing the Bazaar, chap. 4, p. 45.

(31) . Akerlof, “The Market for ‘Lemons’”; Arrow, “Uncertainty.”

(32) . See Kahneman and Tversky, “Prospect Theory.”

(33) . See Bowles, “Mandeville’s Mistake.”

(34) . Hirschman, The Passions.

(35) . See Taylor, Anarchy and Cooperation; Ostrom, Governing the Commons.

(36) . As I mentioned earlier, markets have coexisted with political regimes that have denied basic liberal freedoms. Pinochet’s Chile and Nazi Germany are prominent examples.

(37) . Recent work recognizes that market failure may be more widespread than previously supposed, due to transaction costs, information failures, and strategic actions by the parties.

(38) . Economists also consider the effects of a market transaction on the relative prices of goods for others, effects referred to as “pecuniary externalities.” Such effects are also omnipresent, and I discuss their significance in later chapters.

(39) . Herzog, “Externalities.”

(40) . The canonical statement of the harm principle is in John Stuart Mill’s On Liberty (Harmondsworth, England: Penguin, 1984).

(41) . One might try distinguishing between a person’s moral judgment and his happiness or utility, but such a distinction seems at best unmotivated from within the framework of economics, which tends to view moral judgments in terms of preference functions.

(42) . Elizabeth Anderson notes of the theory of market failure that it is “a theory not of what is wrong with markets, but of what goes wrong when markets are not available” (Value, 192).

(43) . The “complete markets” model assumes away the problems of contract enforcement. This assumption is actually hugely misleading, because once we assume that there is trade occurring extended over time, then it may not be in the economic interest of agents to fulfill their contract. This means that we have to rely on norms or other mechanisms to secure compliance: the other mechanisms may interfere with efficiency.

(44) . Robbins, An Essay.

(45) . Tobin, “On Limiting,” 269. Of course, it is possible to argue that vote selling creates externalities, but I won’t pursue that argument here because I find it unpersuasive. But see http://gregmankiw.blogspot.com/2007/11/on-selling-votes.html.

(46) . Basu, “Economics.”

(47) . This condemnation has found its way into international law. Article 47 of the Protocol Additional to the Geneva Conventions of 12 August 1949 and Relating to the Victims of International Armed Conflicts states, “A mercenary shall not have the right to be a combatant or a prisoner of war.”