3.1 The root idea

The question before us now is what views of well-being states might promote consistently with ensuring that the imperative to respect persons is met. We contend that, given any plausible understanding of agent sovereignty, policies aimed at bettering people's lives must do so according to the beneficiaries’ own standards. They must not impose an external standard of well-being on people. While we will focus on the applications of this point to WBP, we can state it more broadly: insofar as the aim of a policy is to make individuals’ lives better, whether by promoting their well-being, their excellence, beauty, wealth, or whatever, the standards of ‘better’ employed must be those of the individuals themselves. Otherwise it is difficult to see how the persons remain sovereign concerning their personal affairs: someone else is deciding for them, in part, how their lives should go.

The basic idea of pragmatic subjectivism is a simple one: to promote well-being while respecting persons, we must promote well-being as people see it. But this root notion raises some questions, which we consider in this section. The basic idea also needs a blueprint for implementation; in order to put pragmatic subjectivism to work, we’ll need to settle on some interpretation of ‘well-being as people see it’. We argue below that promoting well-being as people see it means paying attention to the values that make up people's conceptions of well-being. In this section we take this for granted, and in the next section (4) we provide an argument for the importance of values. However, it should be noted that the case for pragmatic subjectivism does not depend on the particular interpretation of the view we defend in section 4; one could accept the basic idea behind pragmatic subjectivism without accepting our view about values.

One might think that pragmatic subjectivism amounts to an endorsement of subjectivism about well-being, or what we’ll call substantive subjectivism. In the philosophical literature on well-being, theories tend to be divided into subjective theories and objective theories. According to a common understanding of subjectivism, roughly, what items contribute to well-being depends wholly on the subjective attitudes of the individual; objectivism denies this (Sumner Reference Sumner1996: 38). But person-respecting welfarism entails nothing at all about the nature of well-being; it tells us only what governments may promote in seeking to better people's lives. For all we claim here, the best arguments may well favor a stringently objectivist theory of well-being. That governments may not permissibly impose that view of well-being on their citizens has no bearing at all on its correctness. What's good for you, and what others may do to advance your well-being, are different questions. Antipaternalistic scruples might counsel against imposing our views of well-being on others, but they do not counsel reality not to be objectivist, nor do they advise us to refrain from believing that reality might just be like that. Objectivism about well-being is entirely consistent with any plausible principles of respect for persons. At most, such principles might demand that, in practice, we adopt a stance of respectful humility, taking others’ verdicts about their interests to be worthy of deference—even if we also believe that they might nonetheless turn out to be badly mistaken (for detailed discussion of the ways in which people can be mistaken in matters of personal welfare, see Haybron Reference Haybron2008). So person-respecting welfarism offers no support for subjective theories of well-being.

There are good reasons, as well, not to assume a subjectivist account of welfare in policy. For starters, governments would be wise to steer clear of the long-running debate about the nature of well-being: for thousands of years hedonists, Aristotelians, and many others have failed to generate any sort of consensus about the right view of well-being. While there may be significant agreement about the individual ingredients or causes of well-being, no theory of well-being commands a clear majority among ethical theorists. It would be hubristic and needlessly contentious for policymakers simply to help themselves to a highly tendentious and sharply contested theoretical position in a field where they have no significant competence—at least where there are more modest alternatives, as we will suggest is the case here.

A second concern is that not all people are subjectivists about well-being, and certainly not all agree on any particular variant of it. Governments that assume subjectivism are effectively taking the stance that many of their constituents—Aristotelians and Thomistic Catholics, for instance—are simply wrong about what's good for them, or at the very least, that they are mistaken in their conceptions of well-being. For those drawn to subjectivist accounts on antipaternalistic grounds, this would be a fairly ironic position to take. At any rate, it is not clear that governments should be in the business of endorsing particular conceptions of the human good. Doing so might reasonably be deemed inherently paternalistic, even if it does not strictly infringe individuals’ pursuit of the good as they see it: ‘We think your Aristotelian conception of well-being a groundless superstition, but since we think well-being is actually just a matter of getting whatever you happen to want—however stupid the reasons for it—we will indulge your preferences anyway’. Discussing perfectionist approaches to policy, Nussbaum calls this sort of practice ‘expressive subordination’ and deems it a form of religious establishment (Nussbaum Reference Nussbaum2011). Yet it is a problem not just for perfectionists, but for any policy approach that assumes a substantive account of well-being, including the standard preference-based approach in welfare economics. We suspect many economists would be surprised at the suggestion that their favored approach to policy amounts to a kind of subordination. Of greater concern is the likelihood that such an attitude toward constituents’ values will, in practice, encourage governments to adopt more clearly paternalistic policies that effectively steamroll those values. As a practical matter, policymakers who officially pronounce the personal ideals of many constituents to be mistaken are unlikely to respond to those values in a sensitive and discerning manner.

In short, person-respecting welfarism offers no support for grounding policy in a subjectivist theory of well-being and in fact counsels against it. Person-respecting welfarism instead favors pragmatic subjectivism: governments must take no stand regarding the nature of well-being, deferring entirely to individuals’ own conceptions of well-being in promoting their interests (for related views, see Fleurbaey Reference Fleurbaey2012; Hausman Reference Hausman2010, Reference Hausman2011; Sobel Reference Sobel1998; Wren-Lewis Reference Wren-Lewis2013). This is a kind of subjectivism, but it differs from substantive subjectivism in that it makes no claim about what really is good for people. Pragmatic subjectivism is neutral with respect to theories of well-being—Aristotelian, hedonistic, preference satisfaction, etc.; it thus insulates policy from needing to take a stand on philosophical debates about the character of well-being.

Notice that to say that pragmatic subjectivism is neutral in this way does not mean that all conceptions of the good will in fact be promoted equally. If, for example, Fred's conception of the good involves living in a polity governed by a particular conception of the good, such as a shared religion, he is unlikely to be satisfied by a regime operating on pragmatic subjectivist principles. But policymakers in that regime will at least deem this a cost to his well-being: Fred's well-being is not advanced by their policies, at least in that respect, and he will be regarded as someone who loses out in some way. Such losses are inevitable on any policy approach; the important thing, for our purposes, is that they be recognized as such.

Pragmatic subjectivism differs from the substantive type also in the kinds of factors that shape the theory. Both types of subjectivist might agree that policy should defer to people's values, yet understand this in different ways. For the purpose of an informed-preference theory of well-being, say, a high level of idealization may be warranted, since the goal is to yield the right verdicts about well-being for all cases. Pragmatic subjectivists may reject idealization since their goal is not to give the criteria for well-being but to specify the goals that should drive policy given, inter alia, the demands of respect for persons.

Pragmatic subjectivism resembles liberal neutrality—the idea that the state should be neutral among rival understandings of the good—in acknowledging the importance of people's conceptions of their own good and in enjoining governments not to take sides regarding ideals of the good life. Pragmatic subjectivism may indeed be a variant or corollary of the neutrality doctrine, depending on how the latter is understood (see Dworkin Reference Dworkin and Hampshire1978; Rawls Reference Rawls1993; Sher Reference Sher1997; Klosko and Wall Reference Klosko and Wall2003). However, note that pragmatic subjectivism does not rule out government efforts to promote certain welfare values, even controversial ‘objective’ values such as achievement. If enough of a policymaker's constituents are Aristotelians, then WBP may well include the promotion of distinctively Aristotelian values. Indeed, pragmatic subjectivism is consistent with promoting religious views of well-being. This is pretty distant from orthodox readings of the neutrality doctrine.

Similarly, pragmatic subjectivism does not limit governments to promoting only those goods that are the subject of an overlapping consensus among citizens. If enough citizens take a controversial good to be part of their conception of a good life, the government may have reason to promote this good even if it forms no part of the conception of the good for other citizens. Policies might also accommodate diverse weightings of goods; rural and urban populations, for instance, might tend to differ on the relative importance of goods such as community and achievement, so that different policy regimes make sense for those locales. Moreover, consensus about such matters in any population is unlikely, so that virtually any policy will be suboptimal from the perspective of some citizens.

Further, pragmatic subjectivism focuses narrowly on a particular kind of good and is silent about policy that promotes other kinds of values, unlike some common understandings of liberal neutrality. For instance, policymakers might regard species, ecosystems, or other natural entities as having intrinsic value, and they may protect these for that reason. Such a view of the good might be quite controversial, and hence nonneutral in the present sense, but promoting it is not ruled out by pragmatic subjectivism. Similarly, state efforts to promote the arts, a stock example in the neutrality debates, are compatible with pragmatic subjectivism insofar as the grounds for promoting the arts are not rooted in a conception of well-being. Perhaps certain works of art or certain artistic forms are just intrinsically valuable and ought to be preserved independently of the benefits they confer on people. Again, pragmatic subjectivism has no quarrel with promoting the arts on substantive grounds of value, so long as the value is not that of citizens’ well-being or, more broadly, the goodness of their lives.

The moral ideal that chiefly animates pragmatic subjectivism in some ways resembles the notion of public reason, but it is weaker: it does not entail any general claim that moral or political rules must be justifiable to all citizens. It requires only that, insofar as policies are aimed at improving the lives of citizens, they must be justifiable to the intended beneficiaries, in the sense that they would consent if moderately well-informed and reflective (i.e., roughly, if they were ‘in the room’ with the policymakers and hence fully a party to the deliberations).

Again, most policies will not benefit all citizens—for example, those who don't use or want parks are still taxed to pay for them. Pragmatic subjectivism does not require that those disadvantaged by a policy would consent. But the costs to their well-being, as they see it, must indeed be counted as costs of implementing the policy.

3.2 Is a pragmatic subjectivist approach to well-being policy really well-being policy?

It may seem as though pragmatic subjectivism actually rejects WBP: what it tells governments to promote is not well-being per se, but what people take to be well-being. If the citizenry uniformly embraces a mistaken theory of well-being, then WBP in that polity will not in fact aim at their well-being. In principle, if they are mistaken enough, it could even be disastrous for their welfare.

It would be more accurate to say that, on our view, WBP aims to promote well-being; yet policymakers must defer to individuals’ values to determine the appropriate standard of well-being. Well-being policy as conceived by pragmatic subjectivism is indeed well-being policy even if—like any approach to WBP—it may not always succeed in promoting what is actually good for people.

There is a significant epistemological point to make here as well: governments are not likely to be more reliable than individuals in deciding what values should ultimately govern their lives. We will see below in our discussion of values that the focus on values corrects for many mistakes, such as misplaced priorities; in general, individuals’ personal welfare values probably tend not to be radically mistaken. Notice that all major theories of well-being plausibly tend to agree on what things are most important for well-being, for most people, in practice: relationships, security, achieving major goals, developing and exercising their capacities, pleasure, and so forth (Tiberius Reference Tiberius and Waterman2013b). These are also things that ordinary people tend to value for themselves; this is not surprising, given that theories of well-being are standardly defended through the method of reflective equilibrium, which relies heavily on intuitions about cases (Tiberius Reference Tiberius, Boniwell and David2013a). Even Aristotle, whose eudaimonistic theory is the paradigm of an objective theory, thinks that theorists must pay attention to the endoxa or common beliefs about the good life (Kraut Reference Kraut and Kraut2006). It is unlikely, then, that a group of citizens large enough to attract the interest of government policymakers could be so fundamentally wrong about the components of their own well-being that a policy that genuinely promotes their values would lead to disaster. This being unlikely does not make it impossible, of course, and there may be real examples of this, such as religious groups who refuse life-saving medical treatments. What pragmatic subjectivism implies about such cases is that a policy forcing these groups to accept the treatment is not justified on the grounds that it benefits them—and indeed it should be regarded as detrimental to their interests if it, in fact, conflicts with their conception of well-being. It is a different matter what we should say about the children of people in such religions; moreover, it is possible that other reasons (reasons of justice, for example) should be brought to bear even in adult cases.

Nor is it obvious what would be a more reliable standard for policymakers to apply. Even with expert guidance, the possibility of error about the nature of well-being is substantial. Consider that the standard view of well-being among those experts who have the ear of governments has been a crude form of preference satisfaction theory that virtually—perhaps literally—no philosopher thinks could possibly be correct (Hausman and McPherson Reference Hausman and McPherson2006; Hausman Reference Hausman2011). Hedonism, also rejected by most philosophers, is prominently advocated by other experts (e.g., Layard Reference Layard2005). And on points where experts disagree with ordinary citizens, they are not likely themselves to be in agreement, leaving policymakers with no expert consensus to justify overriding citizens’ values. For these reasons, we think it likely that governments promoting citizens’ values will in fact do better at promoting well-being than governments promoting their own views about what is best for citizens. In short, the epistemology of well-being favors pragmatic subjectivism.

We do not mean to dismiss worries about mistaken conceptions of well-being, however. Given the likelihood that some citizens will have false views, that others may reasonably be less than eager to bear the burdens of promoting those views, that any modern polity is going to include a plurality of conceptions of well-being, and that governments will not be well-equipped to facilitate every detail of their citizens’ welfare, it will be wise for governments to focus WBP, for the most part, on a limited number of important, widely shared values. However obtuse some individuals’ values might prove to be, large swaths of the public are not likely to be indifferent to whether they are healthy or unhealthy, happy or unhappy, and so forth. Policies that promote such homely values will very likely tend to promote well-being, whatever the correct theory. Since most WBP will in fact focus on such uncontroversial values, the worries we’ve been discussing in recent paragraphs are largely theoretical. We will return to this question in section 5.

An interesting question concerns the bearing of our arguments on policies aimed at promoting the welfare of children and adults lacking the capacity for rational self-governance, as well as policies regarding the welfare of nonhuman animals. Insofar as norms of agent sovereignty fail to apply to such individuals, the main argument for pragmatic subjectivism likewise fails to extend to them: treating them with due respect does not require deference to their own views about their interests. In such cases, nothing we have said rules out the use of a substantive conception of well-being in WBP, including perhaps an objective account of welfare. (How policymakers should decide which standard to apply given the lack of expert consensus is another matter we will not address here. We set aside as well questions about the rights of parents and guardians to determine what is best for their dependents.) Since the capacity for self-governance comes in degrees, this suggests that WBP might involve a blend of pragmatic subjectivism and the promotion of substantive welfare values for children and others falling below the threshold for competence. Over the course of childhood, for instance, WBP might gradually shift from applying a firmly objective standard of well-being (supposing that to be the right substantive view of welfare) to an increasingly pragmatic subjectivist approach—that is, becoming more deferential to the child's views about her own good as she matures. Significantly, such a shift would mirror the progression that Richard Kraut attributes to commonsense thinking about ‘happiness’ (in our terms, roughly, well-being): for the newborn we apply an objective standard, becoming gradually more subjectivist as the individual reaches adulthood (Kraut Reference Kraut1979). Perhaps it is not that different standards of well-being apply at different stages of life, but rather that we simply become more deferential in how we judge and promote the interests of others as they mature. The common impulse to defend substantive subjectivism about welfare on antipaternalist grounds might reflect a failure to distinguish adequately between substantive and pragmatic subjectivism.

4.1 From preferences to values

Well-being policy, then, must focus on well-being as people themselves see it. We have so far assumed that the way to understand people's conceptions of well-being is in terms of their standards or values. But this requires some explanation. We have not said why the emphasis should be on values rather than, say, on preferences. We turn now to this question.

Before we can explain the focus on values, we need to understand what values are. Values are relatively robust pro-attitudes, or clusters of pro-attitudes, that individuals take to generate reasons for action (a more detailed version of the account on which we are relying appears in Tiberius Reference Tiberius2000, Reference Tiberius2008. See also Schmuck and Sheldon Reference Schmuck and Sheldon2001; Raibley Reference Raibley2010; Anderson Reference Anderson1995). For example, a mother who values being a parent is relatively robustly disposed to feel proud when she takes her child to lessons he enjoys, ashamed when she forgets to pick him up from school, and so on. Furthermore, she takes her being a parent to justify certain decisions and plans she makes for her life, including decisions that require sacrificing other things she wants, and she takes ‘being a good parent’ to be highly relevant to how well her life is going as well as to her sense of self and self-esteem. Because they play this role in deliberation, planning, and action, values are ‘robust’ in the sense that they are relatively stable and do not evaporate under moderate reflection. A person might like something or judge it to be valuable, but do so only very briefly or in a casual, unreflective manner that would disappear under the merest scrutiny or plays virtually no role in her psychic economy. Such whimsical attitudes do not plausibly reflect what a person genuinely cares about, who she is, where she stands, or what she thinks it is to live well.

Values may be a special subset of preferences if we understand ‘preference’ broadly enough, but there is an important difference between values and ‘mere’ preferences or desires (we use the terms ‘preference’ and ‘desire’ interchangeably in this paper). To value something and not merely prefer it is to see it as generating reasons for you—as tending to justify responding in certain ways to it and limiting how you might reasonably respond to it. This distinguishes values even from intrinsic or ultimate desires, which involve desiring something for its own sake. Perhaps you want to be famous and adored, and your trajectory in life has been such that, although you originally wanted this for the sake of pleasure, you now desire it for its own sake. You might see your desire for fame as childish, shallow, or incompatible with other things you care about more and, therefore, as providing no reason at all to seek it. Rather, you might think that what you have most reason to do is to go to a therapist to help rid you of this annoying ultimate desire.

To distinguish values from mere preferences we might refer to them, rather clumsily, as robust subjective reason-grounding preferences: robust preferences that the agent sees as grounding reasons for her.Footnote ² They are ‘subjectively’ reason-grounding because, at least on some views of reasons, a person's values may not always ground genuine reasons for her. (A sadist might value the suffering of her victims, but some might deny that this gives her any reason at all to seek it.) With this view of values in mind, we can now consider the question: why the focus on values in pragmatic subjectivism?

The reason is that values represent what people see as contributing to a good life for them and what they take to provide practical reasons and standards for evaluating how their own lives are going. Mere desires or preferences, by contrast, may have no intrinsic normative force from the agent's perspective; as that person sees it, they may not be worth fulfilling at all, save to the extent that they relate to his or her values. (Some people may value the satisfaction of their every desire or whim, but not everyone must be like that.) Ordinary consumer preferences may typically be like this: getting the commodities you want may strictly be of no worth to you at all, unless doing so furthers your values. You want a TV, a computer, a car, or whatever, not because you value these things, but because you expect them to promote things you do value—pleasure, accomplishment, etc. Alternatively, you simply have a brute inclination to seek them, say, because you’ve seen the ads—in which case you may see no reason to go for them: to your mind, those desires aren't worth fulfilling at all. And if you genuinely regard those preferences as having no rational force, then governments aiming to promote well-being by your own lights must also do so. To give such preferences weight that the agents themselves firmly reject is to impose an external conception of the good for a person on them, contrary to person-respecting welfarism.

Moreover, there is a close connection between what people value and what they could reasonably be expected to consent to. We should not expect reasonably informed and reflective people to consent to policies that contradict their own values. Knowingly to contravene one's own values would, indeed, seem on the face of it to be irrational. People's values, then, determine the appropriate standard of well-being for WBP.

Note that it is possible to endorse a form of pragmatic subjectivism that does not distinguish between values and mere preferences in this way. We suspect some economists, who may be wary of our focus on values, will be drawn to a pragmatic subjectivism of that sort. For the reasons just given, however, we do not think this a viable or even stable position: the rationale that drives us to be pragmatic instead of substantive subjectivists in the first place seems to require that we distinguish values from other preferences. Without such a distinction, we cannot take seriously people's convictions about what is important in their lives; we cannot take seriously the idea of agent sovereignty.

4.2 Personal welfare values

Having narrowed our concern to values, the question now before us is which values count. A natural thought is to focus on the values implicated in people's conceptions of well-being. Yet, directing policymakers to attend to what people think about ‘well-being’ is problematic. First, people may think nothing about well-being per se. ‘Well-being’ is not a commonly used word, and when it is used, it is rarely employed with much clarity. Therefore, while we think that people do have more or less determinate views about what it is for their lives to go well for them, these views may not be thought of by the people who hold them as conceptions of ‘well-being’. For this reason, policymakers cannot generally rely on people's explicit ideals of well-being as such. Second, insofar as people do have explicit views about well-being, these thoughts might not faithfully or fully represent their evaluative perspectives. How, then, do we distinguish the values that belong to people's conceptions of well-being?

A permissive approach would counsel us not to bother and simply include all values: WBP should promote whatever it is that people value, period (cf. Sobel Reference Sobel1998). In other words, if a policy furthers the achievement of your values, then it succeeds as WBP. However, the permissive approach won't do, partly because it will frequently misrepresent people's views about their own welfare: it is perfectly ordinary for individuals to care about things they see as having little or no positive bearing on their well-being. Artists, social workers, and dissidents, for instance, sometimes choose paths in life that will leave them, in their eyes, worse off than other options before them. Similarly, people care about things that have no relation at all to their own lives, much less to their well-being: the future welfare of a stranger briefly met or the state of the world's ecosystems a thousand years hence. And some individuals may be depressed, detest themselves, or hold religious doctrines on which they actually value their own ill-being. These are standard worries for preference satisfaction theories of well-being (e.g., Parfit Reference Parfit1984; Overvold Reference Overvold1980; Kagan Reference Kagan1992; Darwall Reference Darwall2002; Heathwood Reference Heathwood2011; and Rosati Reference Rosati2009), and whatever one may think of their resolution, the pragmatic subjectivist is compelled to respect these features of commonsense thinking about well-being. WBP needs to focus, not just on citizens’ values, but on what we will call their personal welfare values.

Another reason for WBP to focus specifically on individuals’ personal welfare values and not just their values tout court is that considerations relating to well-being plausibly have normative force that other values lack. It is one thing, for instance, to burden some citizens to help others lead what they see as better lives and quite another to burden them to satisfy others’ completely disinterested preferences for, say, a monument in a far-off city (Scanlon Reference Scanlon1975).

Having established personal welfare values as the proper basis of WBP, the question remains how to distinguish these from other values and from mere preferences. This is a practical question regarding the application of pragmatic subjectivism and is best left for another occasion. In practice, however, the most important values, such as health, enjoyment, and freedom from suffering, will not be hard to classify, and most WBP will focus on such values. We suspect that intuitive judgments of this sort will, by and large, be close enough for government work.

4.3 Does the values approach make a difference in practice?

One might wonder how pragmatic subjectivism differs in practice from a substantive subjectivist view, such as a preference satisfaction theory of well-being. Could not the latter issue the same policy recommendations as the former? In principle it could, but as we’ve already seen, a preference satisfaction theory of well-being, or any other account of well-being, will do so for the wrong reasons: it entails that people with other views of well-being are mistaken, arguably amounting to expressive subordination. Moreover, a substantive subjectivist theory of well-being will needlessly embroil policymakers in contentious philosophical disputes about the nature of human well-being.

But if, as we just argued, pragmatic subjectivism enjoins policymakers to take citizens’ values as the appropriate standard for assessing well-being for policy-making, then any standard form of preference satisfaction account will often fail in practice to get the right results. Suppose, for instance, a government takes up a preference-satisfaction view of well-being, as many policymakers, economists, and other social scientists do (despite, again, the highly controversial nature of this position in ethical theory). This is a paradigm subjectivist view. How does such a regime deal with the very different views of (say) its eudaimonist citizens, who take well-being to consist in a life of achievement and excellence? Quite handily, one might think: their values are simply preferences, and good policy will weight their preferences for goods like achievement accordingly—just like any other preferences. If they care more about achievement than happiness, the preference-satisfaction view can accommodate that.

However, a preference-satisfaction metric of well-being will have to be very different from the formulations standardly used in policy contexts if it is going to respect the structure of people's values. Notice, first, that most people don't just have a strong preference regarding certain values, such as virtue: they regard those values as nonfungible, cherished commitments, not to be traded against mere preferences. An artist might ordinarily be unwilling to compromise her artistic integrity for any amount of money because she regards that integrity as incomparably more important than ordinary commodities. Of course, she might be made to engage in such a compromise in the right circumstances, namely, to protect other cherished values, such as her child's welfare. But such trades do nothing to establish the comparability of such values with ordinary preferences, such as a desire for a flat-screen TV. In short, certain values—call them value commitments—function as constraints on the satisfaction of other preferences and values and cannot simply be treated as strong preferences.

Second, and relatedly, people's value commitments constrain what they are willing to have done for their sake. You may want a convenient new store nearby but think that preference has zero weight in deciding whether to force Granny out of her home to build it. Your convenience preference is not just minor in comparison with her property rights; in your view, it should receive no weight whatsoever in public deliberations about whether to invoke eminent domain against Granny. It is, in that instance, no reason at all even to contemplate forcing someone from her home. In this case, your commitment to property rights silences your convenience preferences. A preference satisfaction view, as standardly formulated, has no way of accommodating this sort of case (see Haybron and Alexandrova Reference Haybron, Alexandrova, Coons and Weber2013).

The focus on values yields a further difference from the traditional emphasis on preferences tout court, one that could prove fairly momentous from a practical standpoint. Namely, far more than their values, the preferences people express in everyday life (particularly their consumer preferences) depend strongly on the menu of options before them as well as on social influences from those around them, and policy cannot help but have pervasive effects on these factors. Employees who don't know anyone who works less than 40 hours per week may not want anything different, simply because it isn't a live issue for them. People who have never known an unstressful pace of life may conclude that ‘stress is just a fact of life’ and form no desire for a more pleasant lifestyle. It is likely that consumer preferences in the United States today, for example, bear the imprint of past policy decisions that in various ways shaped the ‘lifestyle infrastructure’ in which people pursued their lives: a massive increase in productive capacity during World War II that laid the basis for modern consumerism, the financing of a highway system that encouraged suburban sprawl, and agricultural subsidies that encouraged unhealthy diets. It is not implausible that such policies might have influenced Americans’ housing, transportation, dietary, and other preferences, which differ considerably from those of the prewar years.

In these examples, people's preferences change as a result of policy. But their basic values, which most of their ordinary consumer preferences are meant to help fulfill, need not have changed very much, if at all. Furthermore, by altering less fundamental consumer preferences (such as a preference for junk food) policies can make it easier or harder for people to realize their values (such as health). While the American diet has grown less healthy, there is no reason to think that people value health any less. Rising rates of obesity and diabetes may be signs that even as people are better positioned to satisfy their consumer preferences, they are getting less of what they care about, leading shorter, less enjoyable lives. Thus, it may not be that shifting agricultural subsidies and otherwise nudging the food industry toward a healthier and more satisfying model changes people's basic values even as it alters their consumer preferences or tastes. Rather, such a policy change would enable them better to realize their values.

In short, policies based on the sorts of simple preference satisfaction views favored in traditional economic analyses and in many policy circles often fail to treat citizens with respect. By treating all preferences, including value commitments, as commensurable, differing only in strength, and by disregarding the way values bear on agents’ views of many of their preferences, the use of such accounts of well-being to guide policy risks riding roughshod over people's deepest concerns and aspirations.

It might be objected that a revealed preference view—which is one popular form of the preference satisfaction approach—is compatible with any structuring of values, so long as they are consistent and choices are rational and informed: choice behavior will reflect values. However, such an approach has very limited utility for policy unless supplemented with some means of imputing preferences where choice behavior is lacking, as in cost-benefit analysis. (For further discussion and critique of standard economic policy approaches, see Sen [Reference Sen1977]; Alexandrova and Haybron [Reference Alexandrova, Haybron, Hands and Davis2012]; and Haybron and Alexandrova [Reference Haybron, Alexandrova, Coons and Weber2013]. The latter paper argues that such approaches, contrary to common opinion, actually face grave concerns about paternalism.)

At this juncture it will be helpful to emphasize that some influential arguments for preferentialist and other subjectivist accounts of well-being mirror our own arguments for pragmatic subjectivism insofar as they too rely on norms of agent sovereignty. This can lead to some confusion, as our own arguments might easily be taken to commit us to a substantive account of well-being. We see no reason to think that our arguments commit us in this way even if they might be used to support or critique one or another theory of well-being. More important, the arguments advanced here do not commit pragmatic subjectivist policy-making to taking a stand on the right theory of well-being.