Why Worry about Incentives?
Why Worry about Incentives?
Abstract and Keywords
This chapter considers the ethics of incentives. In the modern world, incentives are becoming the tool we reach for when we wish to bring about change. In government, in education, in health care, in private life, and between and within institutions of all sorts, incentives are offered to steer people's choices in certain directions and to bring about desired policy outcomes. So what? you might well ask. Where is the ethical issue here? From a certain point of view, there is none. Incentives could be viewed as a form of trade. A person is offered something of value to him or her in exchange for doing something valued by the person making the offer. However, not all incentives and disincentives are alike. For instance, we recognize bribery and blackmail as wrong even though both can be described in neutral terms as situations in which a simple trade takes place. The discussion then turns to incentives as a form of power. An overview of the subsequent chapters is also presented.
EXPRESS TRAFFIC LANES are set aside during rush hour for cars with more than two passengers. A will stipulates that a daughter will inherit only if she agrees to be a stay-at-home mom. West Virginia pays married couples on welfare an extra $100 per month, funded by a federal program to promote marriage. The government authorizes tax deductions for charitable contributions. Companies pay schools to install soda machines or televisions in their lunchrooms. Schools pay students when they get good grades. A prominent economist suggests that the government tax calories in order to reduce obesity. Legislators in South Carolina discuss a proposal to reduce prison sentences for inmates who donate organs. A soup kitchen feeds the homeless only if they attend a church service first. Cities across America offer large tax breaks to entice businesses to relocate. A donor funds college courses on the condition that Ayn Rand’s Atlas Shrugged is on the reading list. A state legislator suggests paying poor women $1,000 to have their tubes tied while others debate making welfare conditional on the use of the Norplant contraceptive (p.2) device. All of these are real examples, and the list could be multiplied endlessly.
Increasingly in the modern world, incentives are becoming the tool we reach for when we wish to bring about change. In government, in education, in health care, in private life, and between and within institutions of all sorts, incentives are offered to steer people’s choices in certain directions and to bring about desired policy outcomes. So what? you might well ask. Where is the ethical issue here?
From a certain point of view, there is none. Incentives could be viewed as a form of trade. A person is offered something of value to him or her in exchange for doing something valued by the person making the offer. If the offer is accepted, both parties are better off according to their own lights. If that were not the case, and the benefit being offered were not sufficient, the offer would be rejected. This looks like a trade, and a trade is inherently ethical. It is a voluntary transaction that will occur only if both of the parties involved believe that they benefit from it. Thus, trading is free and rational and, for that reason, it can be considered an ethical relation between persons.
Nonetheless, all incentives and disincentives are not alike. We do recognize bribery and blackmail as wrong even though both can be described in neutral terms as situations in which a simple trade takes place: how much is it worth to a customs official to let his duty slide and ignore a smuggling operation? How much is it worth to one person to know that another will not reveal his criminal past?
But are these cases really the same as our trading your two apples for my three oranges? How can we justify distinguishing between legitimate incentives and disincentives on the one hand, and bribery and blackmail on the other? Viewing incentives as simple trades will not get us very far in answering that question.
Moreover, the question is broader than that: there are incentives and disincentives that we might judge illegitimate that nonetheless cannot be classified as bribery or blackmail. The use (p.3) of incentives in public policy often leaves people with vaguely defined ethical qualms. I expect that some of the examples in the opening paragraph elicited some discomfort in you. What do those “gut reactions” tell us? Should some incentives elicit ethical concerns? How do we make sound ethical judgments in the gray areas?
We often meet these issues in everyday life. Some cases are clear. Most people do not object to rush hour express traffic lanes for multi-passenger cars, for example; most people do condemn actions like bribing a judge. But many cases are not so easily agreed upon. In North Carolina, at one time, a licensed driver of high school age could lose that license temporarily if he were failing a course. On hearing of this regulation, my young daughter said, “That’s a good idea.” My teenage son said, “That’s blackmail!”
Examples in the realm of politics are equally controversial. Environmental policies allow companies to buy and sell pollution credits—but does treating pollution as a commodity distort the moral claim that supports its regulation in the first place? The federal government routinely shapes state policies through the use of federal grants in areas it certainly could not constitutionally regulate by federal law—but is this an illegitimate encroachment of power or not? State and local governments offer benefits to businesses to relocate in their area—is this a use of public resources for the public good or an unfair advantage for new businesses? None of these are examples of bribery or blackmail, but all involve the use of incentives in ways that some people find unprincipled and others find perfectly justifiable.
What is the ground of the moral sensibility that so often finds the use of incentives offensive? Some people object, for example, to offering incentives to encourage participation in medical research. In their view, participants ought to be willing volunteers committed to furthering the research enterprise. Otherwise, they are being objectified, used like lab rats for other people’s purposes.1 On the other end of the spectrum of moral sensibilities (p.4) are those who don’t even condemn blackmail. A blackmailer who asks for something in exchange for refraining from revealing an extramarital affair is only threatening to tell the truth, after all. What is wrong with threatening to do something that is perfectly acceptable to do?
The question of the ethics of incentives goes to the heart of a longstanding confrontation between two sorts of moral attitudes. The first might be called the “moralistic attitude,” according to which the quality of character of the members of society ought to be a central public concern. Since societies can only function at their best if their members, especially their leaders, are capable of virtues like self-restraint, personal sacrifice, and public responsibility, matters of motivation and character formation are critical for politics. The contrasting view I will call the “economic attitude” or the “Mandevillian attitude.” The latter refers to Bernard Mandeville, a Dutch author who famously argued in The Fable of the Bees (1714) that private vices often yield public benefits. In this view, our proper concern should be the aggregate outcomes of individual choices and not their motivation or moral quality. The skillful politician is the one who so manages society that even the self-indulgence and vanity of its members produce public goods. The “Mandevillians” scorn the “moralists” as soft-headed and irrational, willing to sacrifice all sorts of beneficial developments on the altar of an illusory project of moral perfection. The “moralists,” in turn, condemn the “Mandevillians” as reductionist cynics who destroy, by denying, the higher human possibilities.
It is an argument that goes back a long way and still takes many forms. One can find it today whenever the ethics of incentives arises as an issue. For example, in the debate over whether to offer payment as an incentive for people to give blood, some worry that altruistic motives will disappear once payment becomes accepted practice (which will lead to blood shortages as well). Others question whether a system that relies on altruism can efficiently ensure a sufficient supply of blood.2
(p.5) Controversies like these have been around for a very long time, and there is every reason to believe that they will continue in one form or another. There are two recent versions of “Mandevillian” thinking worthy of note. For the last ten years or so, “conditional cash transfer programs” have been popular in Latin America and the Caribbean and have recently been tried in the United States as well. These are programs where poor mothers are given cash payments on the condition that they get their children vaccinated, or send them to school regularly, or some similar requirement.* The term is new, but the general idea is not. Similarly, there has been much discussion recently of “libertarian paternalism.” This approach seeks to change people’s behavior by structuring choice situations in certain ways. For example, one can ensure that people will save more for retirement if they are automatically enrolled in a 401k plan and have to make the effort to “opt out” than if they have to make the effort to “opt in.”3 These approaches seek to increase responsible behavior without dealing directly with responsibility as an aspect of character. This “Mandevillian attitude” obviously favors the use of incentives of all kinds, while the “moralists” condemn them. But neither position gives grounds for making ethical distinctions among incentives themselves.
I hope to do exactly that by adopting an alternative approach to the question of the ethics of incentives—by looking at incentives as a form of power. The use of incentives is one possible answer to the following question: How can one person get another person to do what he wants him to do? When considering forms of power, the classic alternatives are force and persuasion: people can make you do what they want you to do, or they can convince you to want to do what they want you to do. But bargaining—including incentives—is a third form of power. People (p.6) can give you something that you want in exchange for your compliance with what they want. Suppose I want you to do X, and you are reluctant to do it. If I cannot persuade you to do it and I do not have the capability to coerce you to do it, I may still be able to induce you to do X by offering you an incentive. Coercion, persuasion, and bargaining are different forms of power. Each is sometimes legitimate and sometimes not. Examining the standards for the legitimacy of all kinds of power will help clarify the criteria for distinguishing legitimate from illegitimate uses of incentives.
Considering incentives as an exercise of power raises ethical issues that are not brought to light by the typical approach. When incentives are considered exclusively as a type of trade, the crucial ethical question is, “Is this transaction voluntary?”4 The approach suggested here goes further. It raises many additional ethical questions in considering the use, and abuse, of incentives. It explains why some incentives are generally recognized as problematic despite their formal similarity to other kinds of trades. It takes seriously the ethical impulse behind the discomfort that many people experience in reaction to incentive programs, such as some of those presented at the beginning of this chapter. And, most important, it allows us to establish standards for making crucial ethical distinctions. Different kinds of incentives are not alike in the ethical considerations they raise; if we consider incentives exclusively within an economic framework—as simple trades—these important differences are obscured.
Thinking about incentives as a form of power, along with coercion and persuasion, also brings to light important concerns about democratic politics. We are accustomed to thinking about incentives as an alternative to coercion: economic sanctions, rather than military attack, for example, or pollution credit markets, rather than regulation (sometimes called “command and control”). And incentives certainly seem to have the moral high ground over coercion as an alternative. But coercion is not the only alternative.
(p.7) Incentives might just as readily be considered in contrast to persuasion. Persuasion, after all, is also a means of exerting power.5 Incentives attempt to circumvent the need for persuasion by giving people extrinsic reasons to make the choices that the person or institution offering the incentive wishes them to make. When incentives are employed, there is no need to convince people that collective goals are good or to motivate them to pursue those goals by appeals to rational argument, personal conviction, or intrinsic motivations. Experts and powerful elites can thus direct institutions and shape people’s choices without the sort of public discussion and consent that ideally characterize democratic processes of decision-making. To take an example close to home: at many colleges and universities, collective bodies of faculty members have a primary role in designing the curriculum. At the same time, individual faculty members often receive incentives from private donors to develop specific kinds of courses. At what point does this practice of private incentives preempt or undercut collective faculty deliberation over educational goals and practices? At what point does the faculty as a collective body lose control over the curriculum? In this case, incentives seem problematic indeed. When the alternative to an incentive is persuasion rather than coercion, the ethical superiority of the use of incentives is not obvious at all.
Yet there is always pressure toward the use of incentives in politics and government. I have drawn examples from all arenas of social interaction, private and public. But the approach taken here is particularly important for politics. Politics—especially in democracies—is at least as much a matter of nonco-ercive forms of power as it is a matter of coercion. Politicians cannot govern without popular support or the cooperation of coalition partners and allies of all sorts—and that cooperation cannot be compelled. Government must operate with carrots as well as with sticks. The only options besides coercion are bargaining and persuasion, and persuasion is often limited in its effectiveness. This is why some form of bargaining often will be the only effective method available—usually incentives. It is (p.8) particularly important, then, to understand and articulate the complex ethical issues involved in their use as a tool of government: to recognize that incentives are not necessarily preferable to all forms of coercion; that incentives sometimes substitute for persuasive processes, which is a real cost in a democracy; and that the fact that incentives are voluntary transactions does not settle the ethical questions raised by their use.
It is impossible to address ethical questions without first making them visible. The danger is that once incentives are introduced in certain areas and people become habituated to their use, the important questions simply no longer arise. Consider plea bargaining. There was a time when the courts condemned plea bargaining. Today, nearly 95 percent of felony convictions involve a guilty plea. Plea bargaining can appear to be acceptable because it is a voluntary agreement where both parties seem to be satisfied with the terms of the agreement. But if the criminal justice system is meant to mete out punishment that is deserved, plea bargaining ensures that that goal will never be met. Either innocent people plead guilty falsely or guilty parties receive punishment that is less than they purportedly deserve. The important point here is that today, these are not major issues for general public discussion. Over time, we have moved from consensus around the idea that plea bargaining is illegitimate to tacit acceptance of the practice.6 I hope that this book will make its readers worry about things they did not worry about before.
Once worried, we need to reach some kind of judgment. This book assesses incentives, along with the various forms of coercion and persuasion, in order to articulate standards for making those judgments. This is the task of chapter 4, and that is where the theoretical heart of the argument can be found. Before embarking on that task, I present, in chapter 2, an historical account of the use of the term “incentives” and of the introduction of incentives in scientific management and behavioral psychology. This history, surprising in many respects, lends considerable support to my approach. “Incentives” came into the language in (p.9) the early part of the twentieth century in America. During this period, the language of social control and of social engineering was quite prevalent, and incentives were understood to be one tool in the social engineers’ toolbox—an instrument of power. Not coincidentally, incentives were also extremely controversial at this time and were criticized from several quarters as dehumanizing, manipulative, heartless, and exploitative. When incentives are viewed as instruments of power, the controversial ethical aspects of their use come readily to the fore.
This history is followed, in chapter 3, by a discussion of the meaning of “incentives.” The term has become so ubiquitous that it has almost lost all boundaries and definition. I try here to distinguish incentives both from other forms of motivation and from other forms of trade or exchange, reward or compensation. The discussion allows me to isolate a few core characteristics of the kind of incentives that need to concern us.
Having laid this groundwork, in chapter 4 I suggest three basic standards for distinguishing ethical from unethical uses of incentives. These are legitimacy of purpose, voluntariness, and effect on the character of the parties involved. These standards, in turn, rest on the notion that ethical uses of any kind of power must treat human beings as free and rational agents.
Chapter 5 takes up the problem of making practical judgments: how to apply abstract standards in practice. In addition to evaluating any incentive against the three basic standards, we need to look at its context and alternatives. We will find that there is no “rule of thumb” that can be applied without consideration of the circumstances. One cannot know whether or not it is good to put a cast on someone’s leg without knowing first whether or not the leg is broken. Examples from medicine, business, education, government, and so forth show what sorts of questions need to be asked in particular cases. These include the following: Which of the standards is most important in this case? Does the incentive work better than the alternatives? Is it fair? Does the incentive mask accountability? Is this a case of undue influence?
(p.10) Taking all of these questions into consideration, in chapter 6 I explore in more detail four very different domains where incentives have been controversial: plea bargaining; recruiting medical research subjects; the loan policies of the International Monetary Fund (IMF); and motivating children to learn. Each of these cases illuminates a web of concerns surrounding the ethics of incentives and illustrates how legitimate and illegitimate incentives can be differentiated.
The discussion of plea bargaining explores the contrast between viewing plea bargaining as a contract between two individuals and viewing it as an exercise of state power. The latter perspective focuses attention on whether this practice serves the proper purposes of the criminal justice system: establishing guilt or innocence and appropriately punishing the guilty. I argue that when the state offers leniency in exchange for a guilty plea, it subverts these purposes and acts illegitimately.
The next case is the use of incentives to recruit subjects for medical research. The ethics of this practice has been heavily debated, with critics primarily concerned that large incentives offered to vulnerable populations are coercive, violating the requirement of voluntary consent. I argue that the exclusive focus on voluntariness has limited the discussion. Even though, in my view, these incentives are not coercive, there are other ethical issues that come into play. I conclude that, with some important exceptions, the use of incentives to recruit research subjects is, perhaps surprisingly, benign.
IMF loan conditionality is another area where voluntariness is only one among a number of complex ethical issues. Everyone involved in this debate seems to agree that in order to avoid undue influence, IMF loan conditions must be limited to requirements directly related to the purposes of the fund. But this turns out to be a very difficult line to draw. How far can the IMF go in stipulating national policies and institutional structures—even if these might ensure that the borrowing country will have a strong enough economy to repay its loans and avoid future debt? How effective are IMF loan conditions in practice? (p.11) Is it fair for the poor to bear the burden of restoring economic health? And what responsibility does the IMF have when its programs fail?
Finally, I consider the use of incentives to motivate children to learn, particularly recent “pay for grades” programs. Here, we can see that incentives have inherent limitations. In educational settings, they can work for some purposes (e.g., performing routine tasks) but not for others (e.g., problem solving). And their effects tend to be short-lived. Moreover, where children are self-motivated, incentives predictably “backfire,” turning play into work and decreasing a child’s interest. In evaluating the use of incentives with children, we need to attend to considerations of character as well. How do different types of motivators affect not only learning but also, for example, self-motivation and the sense of mastery? The discussion of this case allows me to distinguish when incentives in education are likely to be useful and when they are likely to be counterproductive.
Examination of these various cases reveals the multiplicity of questions that must be asked before coming to a judgment. And so chapter 7 returns to the contrast between treating incentives simply as a form of trade and treating incentives as a form of power. The problem with the “trade” approach is that it tends to focus on voluntariness as the only important ethical issue to consider. To increase a person’s choices always seems to be a good thing, but it does not settle every ethical question. Here I try to show that there are some offers that should not be made, even if a person is perfectly free to reject them.
To take an extreme example, in William Styron’s novel Sophie’s Choice, a Nazi concentration-camp officer offers a woman the opportunity to save the life of one of her two children. If she refuses to choose one, both will die. But the same act of choosing life for one child is also the choice of death for the other. Had the officer actually wished to save the life of a child, he could have made the choice between the two children himself. His decision would have been authoritative, but it would have been far preferable to giving the choice to the mother. In this (p.12) case, what seems to be a gift of choice is a perfect expression of cruelty.
In less extreme cases, offers can be paternalistic, manipulative, seductive, exploitative, or irresponsible in a wide variety of ways. Hence, I try to take the discussion in this chapter “beyond voluntariness” to include additional ethical standards.
I also consider the contrast between two ways of thinking about what voluntariness itself is: having a choice or acting autonomously. Incentives always present people with choices, but they can be an affront to their autonomy at the same time. This observation explains why incentives predictably backfire in certain situations.
Finally, in chapter 8, I raise the broader question of the relation between incentives and democratic politics. This question arises first in the historical investigation of chapter 2. During the Progressive era, incentives were considered a tool of social engineering, and social engineering was hotly contested. Is social engineering democratic or anti-democratic? The use of incentives as a tool of government policy appears to increase our choices and protect a space of freedom. We can always refuse the offer. This seems to be better than government regulations that fore-close options and establish penalties for transgressions. But, to the extent that incentives are one of the ways in which experts seek to manipulate behavior and to the extent that incentive systems substitute for persuasion and foreclose deliberation and debate, a democratic people ought to be deeply suspicious of them. At the very least, the question of the ethics of incentives leads directly to the question of the role of experts in a democracy and finally, further still, to the question of what kind of citizens we aspire to be.
My first aim in this book is to make visible the problematic ethical issues involved in the use of incentives. It is impossible to do that without uncovering a host of other issues as well. My second aim is to find ways to distinguish legitimate from illegitimate incentives. That also proves to be a complicated matter. We (p.13) encounter incentives frequently in daily life and are called upon to make judgments about them. We often make those judgments with relative ease. But if you scratch the surface of the matter, you will soon find yourself thinking about psychology and ethics, democracy and expertise, power and freedom.
(1.) Hans Jonas, “Philosophical Reflections on Experimenting with Human Subjects,” Daedalus 98, no. 2 (1969): 219–47.
(2.) Kenneth Arrow, “Gifts and Exchanges,” Philosophy and Public Affairs 1, no. 4 (1972): 343–62; Peter Singer, “Altruism and Commerce: A Defense of Titmuss against Arrow,” Philosophy and Public Affairs 2, no. 3 (1973): 312–20; Richard Titmuss, The Gift Relationship (New York: The New Press, 1997).
(*) Can you imagine a similar program offering tax breaks to middle-class mothers who keep their teenage children drug-free?
(3.) Richard H. Thaler and Cass R. Sunstein, Nudge: Improving Decisions about Health, Wealth, and Happiness (New Haven: Yale University Press, 2008).
(4.) Additional ethical issues may arise over “externalities,” i.e., the impact of the transaction on third parties. See, for example, G. A. Cohen, “Robert Nozick and Wilt Chamberlain: How Patterns Preserve Liberty,” Erkenntnis 11 (1977): 5–23; Michael J. Trebilcock, The Limits of Freedom of Contract (Cambridge, MA: Harvard University Press, 1993), chap. 3.
(5.) Some identify persuasion entirely with rational conviction, which does not compromise free choice, and hence do not view persuasion as a form of power. See, for example, Jürgen Habermas, Moral Consciousness and Communicative Action, trans. Christian Lenhardt and Shierry Weber (Cambridge, MA: MIT Press, 1990), 88–99. But persuasion often involves irrational mental processes as well. See Robert B. Cialdini, Influence: Science and Practice, 5th ed. (Boston: Pearson, 2009). Moreover, in my view, a person who can influence others’ (p.142) choices through intelligence and eloquence is more powerful than one who cannot. On this subject, see also William E. Connolly, The Terms of Political Discourse, 2nd ed. (Princeton: Princeton University Press, 1983), chap. 3. The various forms of power are discussed in chapter 4.