It is not fair to ask of others what you are not willing to do yourself.(Eleanor Roosevelt). Park. As a runner, Miguel benefits from the expansive neighborhood watch that keeps his local park safe for nighttime jogs. Yet, his neighbors complain, he never takes a turn on patrol. Something seems troubling about Miguel's behavior. Yet it is hard to say what he is doing wrong. He did not ask anyone to act for his benefit. He never agreed to any park-walking plan. Given the number of people involved in the watch (imagine it is a town with thousands of volunteers at the ready), he is not meaningfully adding to anyone's burden. when a number of persons engage in a mutually advantageous cooperative venture according to rules, and thus restrict their liberty in ways necessary to yield advantages for all, those who have submitted to these restrictions have a right to a similar acquiescence on the part of those who have benefited from their submission.1 Put more precisely: Much about this claim is controversial: does the mere receipt of benefits really trigger obligations? Does this hold if beneficiaries do not acquiesce? Must the relevant benefits track objective or subjective good? But my interest lies in a further puzzle, one that arises once we accept that something in this vein explains the wrongness of Miguel's actions. Snow. Neda hates shoveling snow. One morning after a big storm she wakes up to learn that enterprising elves have cleaned her sidewalk and driveway for her. She is delighted. She is less delighted when they drop off their (reasonably priced) bill.3 Many scholars treat Park and Snow as different in kind. While the former is said to involve wrongful free-riding, the latter is viewed as a predatory demand. Consequently, while Miguel is said to have a duty to join in, Neda is thought to do nothing wrong if she refuses to pay the bill. The problem is that the two scenarios look remarkably similar. In both situations, a group of people provide an unrequested benefit and take their doing so to trigger an obligation for beneficiaries to repay in a manner specified by their benefactors. My goal in this essay is to make sense of these competing intuitions by developing an account of what differentiates predatory demands from practices that properly generate duties of fair play. In fact, I will argue, cases like Snow and Park are even more similar than theories of fair play have acknowledged. Nonetheless, we can distinguish the two by properly situating fair play in the broader moral landscape. Doing so better grounds the duty and, more precisely, illuminates its scope—but it requires profoundly reimagining what fair play asks of us in a way that calls into question long-standing assumptions about civic ethics. My argument proceeds as follows. In Section I, I detail an account of the moral motivation that underlies the duty of fair play. Building on recent work by Garrett Cullity and others, I argue that such obligations arise from a concern for fairness best understood as a demand for appropriate impartiality. Those who free-ride make unjustified exceptions by granting themselves a privilege that they would deny to others. In Section II, I raise a challenge to recent attempts by Isabella Trifan to flesh out the relevant notion of impartiality. We can, she suggests, distinguish cases of free-riding and predatory demands by looking to participants' attitudes. On her account, people are similarly situated such that non-contributions constitute violations of impartiality so long as they share a preference for receiving the same good without contributing to its production. But this approach, I show, fails to capture widespread and appealing beliefs about fairness. In Section III, I propose an alternative. Fairness concerns arise whenever people are willing to accept others working to their good, but unwilling to return the favor. This approach, I show, makes better sense of our intuitions. However, as I reveal in Section IV, it suggests significant revisions to our understanding of fair play. Accepting it requires upending long-standing claims about our responsibility to vote, pay taxes, and undertake other civic acts. In Section V, I defend this view against objections. The reciprocity approach, I argue, provides a more parsimonious, grounded, and instructive account of fair play, one that better explains and distinguishes cases of wrongdoing and predatory demands. Underlying defenses of fair play is a concern for fairness. Beneficiaries are said to treat benefactors unfairly when they fail to participate in the scheme from which they have benefited. Rawls, for example, writes, “We are not to gain from the cooperative labor of others without doing our fair share.”4 Klosko holds that non-cooperation is problematic because “th[e] situation is unfair.”5 Murphy regards lawbreakers as acting wrongly because they enjoy an “unfair profit.”6 Just what defines something as unfair, and the features in virtue of which an act qualifies as such, have remained largely unspecified in discussions of fair play. In recent scholarship, only Garrett Cullity and Isabella Trifan have sought to offer a systematic account.7 Fairness, they argue, is a matter of appropriate impartiality. A judge ought not to prefer one convicted felon over another in sentencing because of the former's religion, a professor should not grade her spouse, all the participants in a soccer game should be held to the same offside rule. Free-riders, they argue, act wrongly because they violate this standard. In refusing to repay the benefits that they have received, such actors permit themselves a privilege they would deny to others. I think this focus on impartiality captures something important about transactional fairness. The request that others treat us fairly is importantly comparative. In this way it is distinct from the demand that we receive what we want or even otherwise merit. The worry is lest we be treated differently to our detriment. Of course, not all differential treatments fail this standard. If I enter the children's race, I will not be awarded a medal even if I run faster than my toddler opponent. The question is what qualifies an act as instantiating an unjustified privilege? When these features are present, a person who avoids contributing unjustifiably exempts herself from rules that she would apply to others. But those who do so in cases that lack these desiderata exhibit no such bias: they would take everyone to have good reason to act in the same way. This account, Cullity argues, succeeds in our aim: it distinguishes cases where fair play applies from those to which it does not, permitting us to identify wrongful demands. Situations like Snow, he suggests, violate the third requirement. To take Neda to have to pay the elves, “would mean holding everyone liable to pay for all unsolicited benefits that are worth their cost.” This, he contends, would be destructively inefficient.9 Here is the problem. Park can be read as having the same form as Snow: a group of people provide an unsolicited benefit and demand reasonable payment (in the former case, that Miguel join the patrol; in the latter that Neda pay the bill). If this behavior violates the generalization requirement in the latter case, why not in the former? Enterprising Scientists. Everyone in our town runs the risk of contracting a mild chronic illness because the water supply is tainted with a dangerous chemical. One day, I am delighted to find that the water is safe to drink, thanks to a group of scientists who passed through our town and implemented a water-purifying mechanism overnight. I am less delighted when I receive the (reasonably priced) bill.10 Like the neighborhood watch, the scientists provide a good (enhanced health) that improves everyone's lives. Framed this way, the case passes the test. Contributing both to this and to any similar future efforts would make each person better off. At the same time, the scientists did precisely what the elves were doing—providing drive-by benefits and hoping to get paid. According to Cullity, such behavior flunks the test. The account thus fails to provide conclusive guidance: which features should we treat as salient for the purposes of generalization? The Free Rider's Preference (FRP). I prefer that others pay for this valuable collective good that I can enjoy for free and for which I would be prepared to pay, in the conditions under which it is offered, if I had to.11 They share this attitude, she holds, so long as—and only so long as—their preferences are ordered: Trifan takes this view to provide the clarity we seek. On this reading both Miguel and the townsfolk act wrongly if they refuse to pay up. Each shares with their benefactors the right point of view: the desire that the park and the water supply be made safe without their having to contribute to the process of making it so (in the case of the scientists because—or so Trifan posits—they share in a duty to keep consumers healthy). In contrast, the elves are not interested in the benefit that they provide (or so we are to presume). What they want is not to have the sidewalk shoveled: it is to be paid.13 Consequently, Neda violates no rule of impartiality in failing to pay her bill. Garden. You go out of town for a few weeks and forget about your garden. Mary notices the plants wilting away. She steps up and waters the garden daily. Mary does not own any plants. This case violates the FRP as Trifan understood it. Mary does not desire the benefit of having you water the plants that she does not have. According to the shared preferences account, you thus have no obligation to repay her efforts. But that seems silly: there doesn't seem anything predatory if Mary expects you to return the favor with cat-sitting, cookies, a ride to the airport. Indeed, the stakes here precisely echo fair-play scholars' complaint: if you do nothing to Mary's benefit, you enjoy an unfair profit. You are better off and she worse off because of work that she did to your good. Far from being fair, if Mary were (for example) to ask you to check in on her cat while she was away, the declaration that “I owe you nothing because you don't have a garden” would ring almost bizarrely churlish. If Cullity's explanation is unduly vague, Trifan's thus proves unduly narrow, cutting out cases that seem to generate real considerations of fairness. There is, however, a way to revise the view that avoids this mistake. We can capture the sense that Mary's efforts deserve reciprocation by recasting the FRP. As formulated, the approach presumes that persons are similarly situated only insofar as they share an interest in particular goods—watered flower garden, washed cars. This vastly limits the set of cases to which fair play applies. Our tastes and aims are diverse. You love dark chocolate; I hate it. You care about arias; I only like Norwegian death metal. But we need not view ourselves as operating in such distinct spheres. Despite our differing views on Zumba classes or DIY home renovation, there is something we share—a broader mutual concern for a world where people provide benefits that enrich our lives. My life goes better when you go out of your way to return my lost wallet; your life goes better when I refill the printer paper or drop off cookies in the breakroom. Each of us equally desires that others not take advantage of our own such efforts. You do not want your work on my behalf to go unreciprocated; I want the same for my efforts. In this way, Mary and you find yourselves in the same position despite your divergent hobbies. Both of you have an interest in other people advancing your projects. While you need help with your faltering garden, Mary may find herself requiring assistance with her car or somebody to take in the mail before it is stolen. At a more general level, your preferences thus align precisely as Trifan would desire. Each of you share the preference ordering (for simplicity's sake, here and going forward, I will focus on the first two preferences, since Trifan takes them to encompass all the elements relevant for the FRP) that: This view rescues our intuitions about fairness. If, given this preference ordering, you are willing to have Mary work on your behalf without doing the same, you do precisely that which Cullity and Trifan found to violate impartiality: make an unjustified exception of yourself by holding others to standards you would not yourself employ. The challenge is that this formulation of the FRP no longer answers our original puzzle in what we took to be the right way. While Garden, Enterprising Scientists, and Park all qualify as demanding a return, so too does Snow. After all, the elves really do work for Neda's benefit at a cost that all-things-considered she is willing to pay. The solution lies in further revising and refining the FRP. There are, I think, two improvements required: the first to the way the approach conceptualizes costs, the second to the way it conceptualizes benefits. Let's start with the former. Recall that our aim is to identify the ways in which parties must be similarly situated for non-contributions to count as violations of impartiality. As currently formulated, the FRP takes parties to have this status so long as they both share a preference for receiving a benefit at the price of “bearing the benefits-producing costs,” though they would prefer to receive it for free. This focus on the cost of producing the benefit is, presumably, itself intended to reflect a concern for fairness. After all, the thing that is said to trigger the obligation is the fact that one party has taken on those very costs to the advantage of the other. To equalize the situation, we are to infer, the beneficiary should join the benefactor in paying the price by joining in the production of the good from which they have benefited. It is worth noting that, although this repayment structure is baked into the FRP and often taken for granted by scholars of fair play, it is not, in fact, something that Trifan seems to accept, though she offers no explanation. After all, her view is not that having benefited from the work of the scientists, the townspeople are duty-bound to study water safety and take up a practice of roving do-gooding, but rather that they ought to pay the bill with which they have been presented. In light of this, there are two ways to read the FRP. On one reading, the cost that beneficiaries must be willing to bear involves joining the processes required to produce the relevant goods, as Hart and Rawls suggest. On another, beneficiaries are duty-bound to pay a reasonable price for the good that they have received, as determined by the benefactors.14 The two readings are linked, since benefactors presumably play some part in determining the structure of the benefits-producing practice. For our purposes, selecting between these readings is irrelevant: both violate impartiality. In either case, benefactors reserve a privilege for themselves that they would deny to beneficiaries. The issue is this. Benefactors choose the good that they will provide at their own discretion. Your neighborhood watch, your city's road builders, your garden-watering neighbor do not—in many cases cannot reasonably—ask whether you would prefer they build park trails or lay new train lines. This is in part what distinguishes fair play from consent or contract. All that matters in generating the obligation is that the resulting good is genuinely beneficial and the cost not too high. By contrast, beneficiaries are said to enjoy no such freedom. They are required either (on the first reading) to contribute to the production of the good they received (selected at the benefactor's discretion) or (on the second) to pay the price (set by the benefactor). In this way the parties are not similarly situated: benefactors exercise discretion over how they advantage beneficiaries, beneficiaries lack this privilege. Baking. Many of the residents of Jane's apartment building decide to create a scheme for repainting the hallway walls. Jane knows that she will benefit from the improved decor but refuses to participate because she hates the smell of paint. Instead, she bakes cakes for all the participating residents and leaves them in the common area for them to enjoy, which they very much do.15 According to the classic account of fair play reflected in the FRP, Jane acts wrongly. Despite her distaste for the smell, she would (we can imagine) prefer assisting in repainting to leaving the walls undone. She is thus said to possess the free-rider's mindset. But, Jane could say, she is not making an exception of herself. She is doing precisely what her benefactors are doing—contributing to their good at her own discretion. In this way her behavior is consistent and impartial: she would take everyone to have reason to act in the same manner. Inaction. Many of the residents of Jane's apartment building decide to create a scheme for repainting the hallway walls. Jane knows she will benefit from the improved decor and would prefer the cost of having to contribute to or otherwise repay the efforts to not receiving the good. However, she refuses to join in or do anything to benefit the residents who take on this labor. Here Jane does make an unjustified exception of herself. She wants others to labor to her good while she does nothing to benefit them. Taking the wrongness of unfair play to consist in violations of impartiality thus supports a revised version of the FRP, one that takes a more open-ended approach to payment: This version of the FRP maps more neatly onto the cases that we have explored, with greater explanatory power. On this reading, Jane acts appropriately in Baking, but behaves unfairly in Inaction when she does nothing to her neighbors' benefit. Miguel's neighbors equally have cause for complaint when he benefits from the neighborhood watch, but does nothing for them. Unlike shared preferences, this approach makes sense of the intuition that Mary's work watering your garden triggers a real concern for fairness. And unlike that account, it can sensibly explain what went wrong in Snow. The problem isn't that the elves didn't want sidewalks shoveled (perhaps they did!). Rather, the issue is that the elves took themselves to have the right to dictate the terms of repayment. Their error lay in presenting a bill. By doing so, they took themselves to have the right to impose obligations to act in particular ways over their beneficiaries—a right they did not take their beneficiaries in turn to have over their behavior. This violates impartiality. Critics might raise two objections to mutual exchange: that it is unjustifiably partial, and that it fails to reflect persons' actual shared preference-rankings. Let's consider each in turn. First, detractors might deny that Jane and her neighbors occupy the same position in Baking. While the neighbors are bound to abide by the painting scheme, Jane has retained for herself the right to repay their efforts in whatever form she desires: cakes, plumbing, a communal garden. She thus enjoys a freedom that they lack. But, notice, if the neighbors are bound to the painting scheme, it is only because they committed to doing so.16 Like Jane, they had the freedom to choose a different path. Even if we were to treat this loss of freedom as a cost that must be included in our calculation, surely it too can be repaid by taking on additional burdens.17 Imagine that Jane builds her neighbors a new laundry room, plants a rose garden, and bakes them weekly treats. Something would seem very off if Jane's neighbors were to complain she acted unfairly, while happily eating her cakes among the flowers she laboriously nurtures. Still, opponents might insist that mutual exchange fails to capture participants' and non-cooperators' actual preferences. Both, they might argue, prefer receiving goods at the cost of being bound to (depending on your preferred reading) adopt the same practices as those who benefited them or pay producers' reasonable demands. Consequently, a person like Jane who acts otherwise grants themselves an unjustified exception. As we will see in Section IV, I think the revised FRP can answer the real concern at the heart of this objection—the worry that we will not be able to work together to produce valued goods. Nonetheless, it is important to see that this concern is not foundational to our account of impartial action, as this critique would have us believe. The idea that people wish to be bound to act as their benefactors demand is both unrealistic and utterly unappealing. Why would I take it to be a given that, simply because somebody has done something to my benefit, at their own discretion, they get to tell me exactly how to behave? The distastefulness of this notion is precisely what lends force to many famous objections to fair play.18 It seems unfair that one party could come in this manner to exercise such control over another—an intuition that makes sense only if this is not a realistic measure of the relevant shared preferences. Of course, I have reason to consider my benefactors' requests—I am well inclined towards those who benefit me—but that is quite different from being obliged to obey their dictates. We can make better sense of the notion that people desire to be bound to replicate the benefit-producing behavior from which they have profited, especially in the kind of cases long favored by advocates of fair play—complex cooperative ventures such as those necessary to advance public health or safety. In such instances, participants and beneficiaries might each fear that others exercising discretion will vitiate the relevant benefit: too few people will show up for jury duty, vote, serve in the military. Consequently, they may prefer others to be bound to act as the scheme requires, even at the cost of being so required in turn and despite their lack of control over the initial selection of the relevant good or the precise manner of its provision.19 But the claim that this captures our preferences seems mistaken on several grounds. We have already seen one example. It is commonplace to feel we have moral reason to act on others' behalf differently from how they acted to our benefit: we rely on them when our car breaks down, they on us when they need to carry a couch up on the stairs. If these exchanges are subsumed into a single practice, the restrictions lose all relevant purpose—they do not hold people to advancing specific benefits. If they do not, then this approach either altogether rejects an ordinary feature of our moral experience or demands that we develop a distinct further theory to explain such cases. But perhaps we can succeed at the latter. Critics might argue that this preference ordering applies to a very specific set of cases—large-scale interactions which lend themselves to free-riding, because we need enough—but not everyone—to undertake a task if we are to succeed at producing the relevant good. Such circumstances are routine: we need enough people to serve in the military, vote, go carbon neutral. However, it does not make sense to conclude on these grounds that all of us prefer that everyone be bound to take up the relevant task. Indeed, such a demand would be absurd. It would require that everyone be farmers, dentists, school-crossing guards, sewer-system experts.20 That isn't just unattractive—it is impossible. What we really want is not for everyone to take up such roles, but for everyone to be attentive and responsive to whether enough people are doing the necessary work.21 What we would like even more is that those who do not themselves undertake the relevant acts (having satisfied themselves that sufficiently many are doing so) perform further valuable labor to our benefit. When they do so, we each end up both equally situated and better off—all have paid a price necessary to produce goods and all have benefited from such work. Fair play is not a leveling-down principle. The demand is not simply that others suffer. I don't really want other people to make their way to the polls when the right candidate's election is secured—I want them to do something actually productive to help bring about good social conditions. Miguel's neighbors don't really desire a 400th person on nightly patrol in the medium-sized park—what they want is a sufficient number of people at the job, and others promoting the good through further, more efficacious, means—improving the lighting on local streets, planting a flower garden. The challenges of coordination give us a reason to be humble and careful in our assessments of where our participation is required, but not a reason to accept that we be duty bound to undertake innumerable actions that add no value. This is especially true because public goods are rarely one-shot deals. We receive regular feedback as to whether enough people are going on patrol or using solar panels. Finally, this do-as-I-do approach overlooks an important worry for impartiality: the fact that persons are routinely not similarly situated. It is more costly for a single mother working two jobs to get to the polls than a tenured professor on sabbatical. It seems fundamentally unfair to ask they do the same task in anticipation of the same benefit when the burdens involved differ so dramatically. One person might reasonably balk at being asked to take on an unequal cost. Now, proponents might say that schemes with this feature fail the morality standard and thus do not activate concerns for fairness. Still, this would in practice apply to almost any realistic large-scale cooperative scheme. Such practices are slow-moving and often inattentive to individual circumstances in a way that makes it unlikely they will ever, in themselves, fully succeed at laying out rules that divide work in a genuinely equitable fashion. It is thus reasonable to conclude that people might share a preference for taking on a fair share of the labor of working to others' good, without being willing to undertake precisely the same acts as others, or bow to the dictates of the scheme in which they have participated. Although these objections fail, they do reflect an important way in which mutual exchange proves discordant with our intuitions. Earlier, we treated Park and Scientists as clear instances of wrongful free-riding. Indeed, our central task was explaining what distinguished these cases from inappropriate demands like Snow. That is precisely what the FRP was intended to do. But our revised approach classifies both as predatory demands. Neither the participants in the watch nor the scientists have the right to specify the form in which their efforts are repaid. Both behave no better than the elves when they require the townsfolk to pay up, Miguel to show up. There is a sense in which I think that this is correct. None of these benefactors has a right to specify the way beneficiaries' debts are satisfied. Park, Scientists, and Snow thus share the same moral mistake. Nonetheless, I think more can be said to explain our sense that the former cases generate concerns for fairness absent in the latter. I will say more to reconstruct Park in Section IV. For now, let us focus primarily on Scientists. Understanding how their claims differ from the elves' pushes us to further refine the FRP. Earlier I suggested that each of us shares an interest in “others laboring on our behalf.” Read that way, the elves and the scientists have the same merit. Both work to others' benefit—in ways that we have reason to believe constitute a net gain to recipients. However, our interests are not quite so straightforward. To see this, we need to situate fair play in the broader moral landscape. Freelance do-gooding is not the only way we seek to advance our projects or call on others to do the same. In many cases, institutions beyond others' kindness—the market, contract, consensual negotiation—better track our interests, are more effective and efficient means of benefit. They provide what we want at lower cost with greater control over our circumstances. Each of us thus shares an interest in the success of these institutions. Actions that undermine them are not to our benefit. Still, these institutions have their gaps and limits, in both ideal and non-ideal conditions. There are spaces where non-market actions do better at advancing what we care about, for any number of reasons. There is no company that notices my forgotten flower garden; markets do poorly at preventing pollution; in many places no affordable childcare services for the quite poor exist, nor reasonably priced green spaces. Fair play is suited to these interstitial gaps, to the spaces where other institutions and practices fail to best advance our interests. When people act to fill in these gaps, they provide a great benefit. Conversely, they threaten our well-being when they act to undermine institutions better suited to promoting the good, even when their immediate action technically provides a benefit. This suggests a further revision to the FRP. Contributors and would-be free-riders are similarly situated insofar as they share the following attitudinal ordering: Unlike shared preferences or mutual exchange, this approach successfully captures and explains our deeply held views about fairness in a way that makes sense of the felt difference between Snow and Scientists. Consider the fact that discussions of fair play typically focus on public goods, benefits defined (among other things) by jointness in supply such that the provision of the good to one person entails furnishing it to all.22 The reciprocity account explains this emphasis. These goods are poorly suited to consent- or market-based approaches. It is simply too hard to negotiate terms, communicate agreement, or limit benefits to those who assent. The same