Introduction

Externalities are costs imposed during production or consumption not incurred by the producing or consuming parties. The classic example is the capitalist who builds a factory that pollutes a river, a cost that must be borne by those who drink the river’s water. This is a negative externality, but externalities can be positive as well. This happens, for instance, when an apiarist’s bees wander over to a neighbor’s garden and pollinate her flowers, resulting in a beautiful bloom.

Pecuniary externalities are costs imposed on third parties mediated through the price system. Pecuniary externalities are thus a proper subset of all externalities. For example, suppose you make pizzas for a living, and a new pizzeria opens down the street. Because the new shop lures away some of your customers with their cheap and tasty pies, your revenue and hence profits decrease. This is a pecuniary externality—the new pizzeria’s productive activities impose a cost on you mediated through the price system. Pecuniary externalities can be positive as well. An entrepreneur might invent a new, more efficient way of using copper; because your business relies on copper, this lowers your costs of production and so (all else equal) increases your profits.

All except the most ardent libertarians believe significant non-pecuniary externalities must be regulated. What about pecuniary externalities? Among economists, pecuniary externalities are not typically a cause for concern and, in fact, are part of a well-functioning market (e.g., Holcombe & Sobel, 2001). They put the “destruction” in Joseph Schumpeter’s (2008) famous “creative destruction” mantra. Among moral philosophers, views are mixed. Some hold pecuniary externalities are not cause for moral concern (e.g., Thomson, 1986, p. 160; Schmidtz, 2011, pp. 604–605). Others disagree (e.g., Sen, 1985, pp. 5–6; Hausman, 1992, pp. 103–105; Olsaretti, 2004, p. 120). What all these philosophers have in common, though, is just how little time they spend analyzing pecuniary externalities. Typically, pecuniary externalities receive no more than a paragraph or two worth of attention.

Two recent papers have changed this. Both Richard Endörfer (2022) and Hayden Wilkinson (2022) place pecuniary externalities at center stage. To the best of our knowledge, their papers represent the first sustained investigations into the moral status of pecuniary externalities. Though their arguments differ significantly, both conclude that pecuniary externalities are cause for moral concern. Endörfer thinks the state (in some cases) should regulate pecuniary externalities just as it does ordinary externalities, such as pollution. Wilkinson thinks individuals (in some cases) have decisive reason to alter their consumption behavior to impose or avoid imposing pecuniary externalities.

We join the consensus view among economists that pecuniary externalities are not cause for moral concern. To this end, the current paper critiques the recent arguments of Endörfer and Wilkinson. The structure of the paper is simple. The next section presents and then criticizes Endörfer’s argument (§2). From here we explicate Wilkinson’s argument and then show why it fails (§3). To be clear, we are not arguing that it is impossible to justify the regulation of pecuniary externalities, either with the power of the state (Endörfer’s position) or through individual consumption decisions (Wilkinson’s position). Rather, in showing that both Endörfer’s and Wilkinson’s arguments fail, we are shifting the burden back to those who favor regulation. Until a new argument emerges, pecuniary externalities should be regarded as a morally neutral component of the market process.

Endörfer’s Argument

In On Liberty (chap. 1, para. 9), John Stuart Mill famously introduces the Harm Principle (HP), which says that “the only purpose for which power can be rightfully exercised over any member of a civilized community, against his will, is to prevent harm to others.” Endörfer’s central claim is that pecuniary externalities are subsumed under the HP and thus subject to state regulation. We disagree. We shall argue that, on the most plausible interpretations of the HP, pecuniary externalities are not subsumed under the principle. Although perhaps there are other reasons to regulate pecuniary externalities with the coercive power of the state (we are agnostic on this point), the HP gives no such reasons.

Endörfer (2022, p. 223) interprets Mill’s HP as saying that the “state is pro tanto justified in coercively interfering with a citizen’s conduct if her conduct is harmful or likely to be harmful to other citizens.”Footnote 1 If A harms B or will likely harm B, then that gives the state a reason to coercively interfere with A’s conduct, according to the HP. Note the HP as stated here does not require the state to coercively interfere with A’s conduct. The state’s reason to regulate A’s conduct is pro tanto, which means competing considerations can override it. Just what considerations can override it we discuss below.

To say that pecuniary externalities fall under the purview of the HP means the state has reason to use its coercive power to regulate them. Why think pecuniary externalities fall under the purview of the HP? It must be because they constitute harms. Endörfer does not clearly define harm. The closest we get to a definition is when he defines “market harms” (his term for pecuniary externalities) as “severe welfare losses mediated by price shifts in a competitive market borne by the losers of market competition” (Endörfer, 2022, pp. 221–222). If harms are defined as severe welfare losses, then many pecuniary externalities will count as harms and thus be subsumed under the HP. Consider the following two cases, both of which come from Endörfer (2022, p. 224).

Pollution Pharmaceutical company PollMed Inc. produces several kinds of drugs in their factory complex, which is located close to a river that is not owned by the company. PollMed Inc. manages to keep production costs low by simply dumping the toxic waste generated in the production process into the river. Due to the pollution of the river, the residents of the downstream village, who drink water from the river, suffer considerable liver damage.

Ann’s Apples Ann sells apples at the weekly market. It is her only source of income. One day, CheapPears Inc. opens up a stand right next to her, offering high-quality pears at a much lower price than Ann’s apples. Consumers now exclusively buy pears from CheapPears Inc., while Ann is forced to shut down her apple stand and has no means to maintain her income.Footnote 2

Pollution is an example of a non-pecuniary externality, and Ann’s Apples is an example of a pecuniary externality. Most people agree that PollMed Inc.’s actions are harmful and should be subject to state regulation. Many are hesitant to say the same about CheapPears Inc.’s actions, however. If harms are just severe welfare losses, though, then there is not much difference between the two cases. The villagers suffer severe welfare losses in Pollution and are thus harmed; Ann suffers a severe welfare loss in the latter case, so she is harmed too. The HP applies to both cases. The state has pro tanto reasons to coercively regulate the respective conduct in both cases.

In response to this, one might argue that there is something about pecuniary externalities in particular that exempts them from the HP. That is, there is something special about pecuniary externalities such that, even though they are harms, the state does not have a reason to regulate them. Endörfer spends the bulk of his paper considering four such exculpatory considerations: that pecuniary externalities generate efficient allocations of goods, that pecuniary externalities result from just exchanges of justly held property rights, that market participants consent to having pecuniary externalities imposed on them, and that the policies of the welfare state will protect against harms inflicted by pecuniary externalities. Endörfer convincingly argues that all these considerations fail to exempt pecuniary externalities from the purview of the HP. So, his conclusion stands: the HP applies to pecuniary externalities, making them the subject of regulatory concern.

We agree with Endörfer that if pecuniary externalities are harms, then there is no principled way to exempt them from the HP and thus possible state regulation. We get off the boat earlier than that. We deny pecuniary externalities are harms, at least the kinds of harms typically subsumed under the HP.

To begin our argument, we note that Endörfer’s definition of harm is inconsistent with Mill’s.Footnote 3 Endörfer, recall, defines harm as severe welfare losses. According to one popular interpretation, for an action to be harmful according to Mill it “must actually violate or threaten imminent violation of those important interests of others in which they have a right” (Brink, 1992, p. 85). Others offer a similar interpretation (e.g., Jacobson, 2000, p. 302; Fuchs, 2006, pp. 147–150; Rawls, 2007, p. 291; Donner, 2009, p. 161). In support of it, consider the following passage:

The acts of an individual may be hurtful to others, or wanting in due consideration for their welfare, without going the length of violating any of their constituted rights. The offender may then be justly punished by opinion, though not by law (Mill, On Liberty, chap. 4, para. 3) (emphasis ours).

Mill suggests here that severe welfare losses—actions that are, in his words, “hurtful” or “wanting in due consideration” of others’ welfare—are not sufficient to constitute harm. For an action to be harmful it must set back someone’s interests, interests that they have a right to have respected. If A punches B, then A sets back B’s interests, and it is plausible to think B has a right to have his interest in bodily integrity respected; so, A harms B. If A practices a religion that B finds distasteful, it might cause B great consternation and thus set back his interests, but intuitively it does not seem plausible that B has a right to have this interest—an interest in not having religions he finds distasteful practiced around him—respected. Although A may cause B severe welfare losses, A does not harm B.

There are textual reasons to think Endörfer has the wrong definition of harm, but this may be dismissed as irrelevant. Perhaps Mill got the wrong definition of harm and Endörfer’s definition is superior. We do not think so. To see why, consider a case.

Sad Swifties Country and pop music megastar Taylor Swift is burnt out and wants to retire from the music industry forever. She will no longer record any new material. She will no longer perform any concerts. In fact, she desires to remove herself from the public eye completely and become a recluse. Of course, Swift’s dedicated fans—known as “Swifties”—will be devastated by this. If Swift follows through with her retirement plans, Swifties around the world will suffer severe welfare losses.

According to Endörfer’s definition of harm, Swift’s actions constitute a harm, for she imposes severe welfare losses on her fans. By the HP, the state has pro tanto reason to coercively force Swift to continue making music in perpetuity. That the state would be justified in forcing someone to work in an occupation against their will—what amounts to slavery—is deeply counterintuitive. This is an objection to Endörfer’s definition of harm. Endörfer might respond that the state’s reason to regulate Swift’s occupational choices is only pro tanto; in practice, it will almost always be defeated by competing reasons, so in practice the state will rarely (if ever) force Swift to continue making music.

There are two issues with this response. First, the fact that a moral theory generates repugnant reasons can constitute an objection to the theory, even if these reasons are almost never decisive. Some object to utilitarianism on these grounds. Utilitarianism says slavery is justified so long as it maximizes happiness. In response, defenders of utilitarianism argue that, given plausible empirical assumptions, slavery is unlikely to maximize happiness (Hare, 1978, p. 118). Critics respond: the rejoinder misses the point. If utilitarianism rejects slavery only because the facts turn out a certain way, then that’s a problem with the theory (Rawls, 1971, pp. 158–159). Regardless of its impact on overall happiness, slavery is wrong, and any plausible moral theory must say so. Analogously, one might think we never have reason to coercively force Swift (or anyone else for that matter) to continue making music against her will. A moral theory that says otherwise cannot be rescued by claiming that, most of the time, other considerations will defeat the reasons we have to force Swift into slavery. Whether the state should force Swift to continue making music against her will should not be up for discussion. A moral theory that opens the door to this discussion—by saying that the state may be justified in forcing Swift into slavery, depending on what other reasons there are—is inherently flawed.

Second, we ought not be so confident that the state’s pro tanto reason to force Swift to continue making music will be so easily defeated. A lot depends on what moral considerations are used to supplement the HP. Mill seemed to think that whether the state should act on its pro tanto reason to coercively restrict harmful conduct depends on a utilitarian calculus (Brink, 2022, Sect. 3.6). For instance, he writes:

As soon as any part of a person’s conduct affects prejudicially the interests of others, society has jurisdiction over it, and the question whether the general welfare will or will not be promoted by interfering with it, becomes open to discussion (Mill, On Liberty, chap. 4, para. 3) (emphasis ours).

If this is how the HP is supplemented and if we adopt Endörfer’s definition of harm, then the state’s reason to force Swift to continue making music may become decisive. If the state does not act on its reason, then hundreds of millions of fans will see their welfare greatly reduced, while Swift gets some respite from life on the road. If the state acts on its reason, then hundreds of millions of fans are filled with joy, while only Swift is miserable. On at least the Millian way of supplementing the HP, the pro tanto reason in Sad Swifties may be decisive.

There are many reasons, then, to reject Endörfer’s definition of harm. What should we replace it with? In what follows we shall examine the two most prominent accounts of harm in the literature, offered by Joel Feinberg and Joseph Raz (Holtug, 2002). Indeed, Endörfer (2022, p. 221) cites Feinberg and Raz as fellow travelers who embrace the HP. He does not, however, look at how they actually define harm. Feinberg’s (1984, p. 15) account of harm is explicitly intended as a reconstruction of Mill’s. Raz (1986, pp. 412, 420) understands his account as going beyond Mill’s, but it is still prominent in the literature. We shall now argue that on both Feinberg’s and Raz’s accounts of harm, pecuniary externalities are not harms. Not only does Endörfer adopt an implausible conception of harm but also, if he adopts a more defensible conception, his argument collapses.

Feinberg’s Conception of Harm

Let’s begin with Feinberg’s conception of harm. He writes: “the term ‘harm’ as it is used in the harm principle refers to those states of set-back interest that are the consequence of wrongful acts or omissions by others” (Feinberg, 1984, p. 215).Footnote 4 We can plausibly understand Endörfer’s severe welfare losses as setback interests. Thus, Feinberg’s conception of harm adds to Endörfer’s conception. Setback interests (severe welfare losses) are necessary but not sufficient for harm; to be sufficient for harm, these setbacks must be wrongful. Much hangs on how wrongfulness is understood. According to Feinberg, a wrongful act is one that is a violation of a right. Person A wrongfully sets back B’s interests when “B’s set-back interest is one that he has a right to have respected” (Feinberg, 1984, p. 108). Thus, to harm another is to set back their interests in violation of their rights. As we noted above, many think this is roughly how Mill understood “harm.”

Feinberg’s account of harm gets the right answer in Sad Swifties. Swift no doubt sets back Swifties’ interests when she retires, but this is not wrongful, because she does not violate any Swifties’ rights in doing so. Swifties do not have a right to have their favorite artist perform and create music for them in perpetuity. So, Swift does not harm her fans when she retires. Hence, the state has no reason to coercively regulate Swift’s occupation choices.

The question now is whether pecuniary externalities count as harms according to Feinberg’s conception of harm. No doubt pecuniary externalities set back persons’ interests. But are any rights violated in the process? Let us think through this question by returning to Endörfer’s Ann’s Apples case. CheapPears Inc. floods the market with inexpensive fruit, putting Ann out of business. Her interests have no doubt been set back, but what right of hers has been violated?

It cannot be that Ann has a right to not face competition from other businesses. Such a right would violate universalizability or reversibility, a condition many believe to be a formal constraint on the concept of right (e.g., Hare, 1963, pp. 89–90; Rawls, 1971, p. 132; Baier, 1995, p. 280). This condition says moral “principles are universal in application. They must hold for everyone in virtue of their being moral persons” (Rawls, 1971, p. 132). Thus, if Ann has a right to not face competition from other businesses, then everyone else must have that right. This includes CheapPears Inc., whose right to not face competition is violated by Ann’s apple stand. Ann’s right to not face competition thus implies that Ann is wrong for running her business in the first place. This, we take it, is a reductio.

Nor can it be that Ann has a right to own and operate a successful business. Many are skeptical there is any right to own and operate a business in the first place (e.g., Rawls, 2001, p. 114). Others believe people do have a right to own and operate a business (e.g., Tomasi, 2012, pp. 68–84; Freiman & Thrasher, 2019). Those who believe there is such a right, however, do not claim that persons have a right to own and operate a successful business. People have the right to try, but not the right to succeed. Indeed, consider the implications of such a right. The right to a successful business would imply that the state must either force persons to shop at businesses they do not want to shop at or subsidize every business that cannot turn a profit. Clearly, this is absurd.

The most plausible response to “what right of Ann’s has been violated by CheapPears Inc.?” is her right to employment. Some assert that such a right exists (e.g., Elster, 1988; Greene, 2019; Schaff, 2017; Tcherneva, 2020). The right to employment, though, does not imply the right to your most preferred type of employment. If that were true, then a world realizing this right would be occupied by very hungry rock stars, artists, and professional athletes. At best, the right to employment implies the right to employment that one finds reasonably satisfying. So construed, CheapPears Inc. does not violate Ann’s right when they put her out of business insofar as there are other employment opportunities available for Ann that she would find reasonably satisfying. Under normal labor market conditions (i.e., no recession or depression), we believe this will be the case.

Moreover, if there is a right to employment, the guarantor of such a right would not be private individuals and firms (such as CheapPears Inc.), but the state. Pavlina Tcherneva (2020) argues that the right to employment should be discharged through public works projects as part of a Green New Deal initiative. Thus, if Ann’s right to employment is violated because her apple stand goes out of business and there are no other reasonably satisfying employment opportunities available, then it is the state, and not CheapPears Inc.’s pecuniary externality, that violates her right and thus harms her. This is because it is the state, and not CheapPears Inc., who is duty bound to supply her a job.

Summing up, Feinberg defines harm as setbacks to interests in violation of rights. If this is how we define harm, then pecuniary externalities are not harms. Pecuniary externalities set back interests, but they do not violate rights.

Raz’s Conception of Harm

Raz provides a different account of harm. According to Raz, “to harm a person is to diminish his prospects, to affect adversely his possibilities” (Raz, 1986, p. 414). Rather than a setback to interests, Raz understands harm as a setback to autonomy, where autonomy is understood as “the ability to choose between an adequate range of valuable options, while in possession of the appropriate capacities to make such choices and while sufficiently independent of others” (Stanton-Ife, 2022, §5.3). For A to harm B thus implies that A renders B’s choice options inadequate, reduces B’s capacities to make choices, or renders B objectionably dependent on A or some other entity.

Raz’s account of harm gets the right answer in Sad Swifties. Swift’s retirement is no doubt disappointing, but she leaves her fans an adequate range of valuable musical options to choose from.Footnote 5 Moreover, her choice to retire does not reduce Swifties’ decision-making capacities, nor does it render them dependent on others in any way. According to Raz’s conception of harm, Swift does not harm her fans when she retires, so the state has no reason to coercively interfere with her occupational choices. The question now becomes: what does Raz’s account of harm say about pecuniary externalities? Is Ann harmed by CheapPears Inc. on Raz’s account?

We think it is obvious that putting Ann’s apple stand out of business does not reduce her decision-making capacities, nor does it render her objectionably dependent on any other entity. The open question is whether it renders her occupational choice options inadequate. What, for Raz, constitutes adequate choice options? He offers several criteria. Adequate choice options “should include options with long term pervasive consequences as well as short term options of little consequence, and a fair spread in between” (Raz, 1986, p. 374). Moreover, adequate choice options include a “variety of options available … A choice between hundreds of identical and identically situated houses is no choice, compared with a choice between a town flat and a suburban house, for instance” (Raz, 1986, p. 375). Finally, adequate choice options must allow an individual to “develop all his abilities, as well as to concentrate on some of them. One is not autonomous if one cannot choose a life of self-realization” (Raz, 1986, p. 376). On our reading, for Ann to have an adequate range of choice options requires that the options she confront are (i) consequential in terms of their impact on her life, (ii) of sufficient variety, and (iii) allow her to develop her skills and abilities.

In normal labor market conditions, we believe these criteria will typically be satisfied. Suppose that after going out of business Ann can either apply her expertise of the fruit market by working for CheapPears Inc., apply her expertise in sales more generally by working in a marketing department at a large firm, or chart a new path entirely by going back to school and getting a degree in a new field, such as medicine or data analytics. Whichever option she chooses will certainly be consequential in terms of how it shapes and affects her life, satisfying desideratum (i). It’s also clear that there is variety among these options, satisfying desideratum (ii). Moreover, these options seem like they allow Ann to develop various skills and capabilities she might have, satisfying desideratum (iii). In this case—and we do not think it is at all unrealistic—going out of business does not set back Ann’s autonomy. The pecuniary externality thus does not harm her, according to Raz’s theory.

Suppose for the sake of argument, however, that CheapPears Inc.’s pecuniary externality leaves Ann with an inadequate range of occupational choice options. Perhaps there is an economic depression and her only option is to pick up trash on the side of the road. Clearly there is no variety from which to choose, and this occupation does not seem to allow Ann to develop her skills and talents. According to Raz’s theory, Ann has been harmed.

While it’s clear that Ann has been harmed, it’s not clear that CheapPears Inc.’s pecuniary externality is what harmed her. According to Raz (1986, p. 416), persons can be harmed not only by actions but also by omissions as well. Moreover, Raz makes clear that it is the job of the state to ensure persons have access to an adequate range of valuable choice options. He writes: “Governments are subject to autonomy-based duties to provide the conditions of autonomy for people who lack them. These extend beyond the duty to prevent loss of autonomy” (Raz, 1986, p. 415). As another example, “The autonomy principle permits and even requires governments to create morally valuable opportunities, and to eliminate repugnant ones” (Raz, 1986, p. 417). On our understanding of Raz’s theory, if Ann is harmed because she is left with an inadequate range of choice options, the harm is not done by CheapPears Inc.’s pecuniary externality, but by the state who harms her by omission. In particular, the state harms Ann by failing to provide an adequate range of employment options for her.

Endörfer might respond that the state does not have the sole responsibility to ensure Ann has an adequate range of choice options. Private individuals, firms, and other organizations bear some of this responsibility as well. Even if this is true, it is still not clear that CheapPears Inc.’s pecuniary externality is what harms Ann. Since Raz admits autonomy-reducing omissions constitute harms, why not say that employers harm Ann by not extending her sufficiently attractive job offers? Why not say the local community harms Ann by buying CheapPears Inc.’s produce instead of hers? Ann is harmed on Raz’s view because she lacks an adequate range of choice options once her business closes, but it’s not clear why this should be pinned on CheapPears Inc.’s pecuniary externality specifically. Why she does not have an adequate range of choice options is overdetermined.

Summing up, Raz defines harm as setbacks to autonomy. Actions (or omissions) are harmful when they leave persons with inadequate choice options, reduce their decision-making capacities, or render them objectionably dependent on others. Most of the time, pecuniary externalities do not do these things. In rare cases, pecuniary externalities may result in inadequate occupational choice options. If they do, it’s not clear the pecuniary externality is what harms the individual, for other parties (such as the state) have a duty to provide adequate choice options. If these other parties fail to discharge this duty, then they are the ones most plausibly responsible for the harm.

To sum up our overall criticism of Endörfer’s argument, Endörfer’s definition of harm is implausible, because it is too capacious. It wrongly generates harms in cases such as Sad Swifties. If we turn to prominent accounts of harm in the literature—those offered by Feinberg and Raz respectively—it becomes clear that pecuniary externalities are not harms. If pecuniary externalities are not harms, then they are not subsumed under the HP. This is in contradiction to Endörfer’s thesis.

Wilkinson’s Argument

Wilkinson (2022, p. 205) understands “market harms” as synonymous with negative pecuniary externalities and “market benefits” as synonymous with positive pecuniary externalities. For ease of exposition, we follow his terminology throughout this section. Wilkinson (2022, p. 205) asks two questions in his paper: first, if an agent’s action imposes a market harm (benefit), does that give the agent a reason to avoid (perform) the relevant action? And second, are these reasons ever decisive, in that agents should avoid (perform) an action because it imposes a market harm (benefit)? His answer to both questions is “yes” (Wilkinson, 2022, p. 222). For the sake of argument we grant that market harms and benefits give agents reasons to avoid or perform certain actions. We target the decisiveness claim. We argue that market harms and benefits very rarely give agents decisive reason to alter their consumption behavior. If we are right about this, then Wilkinson’s argument has little practical relevance. Even if he is right that market harms and benefits give us reasons for action, in very few cases will these reasons actually alter our conduct.

Why think market harms and benefits give agents reasons to alter their consumption behavior? When you impose a market harm on someone—for instance, by contributing to demand and thus raising the price of a good they consume—you reduce their “material wellbeing and/or their overall welfare” (Wilkinson, 2022, p. 222). The fact that imposing a market harm reduces someone’s material wellbeing and/or their overall welfare gives you pro tanto reason to not perform actions that impose market harms. When you impose a market benefit on someone—for instance, by contributing to demand and thus raising the price of a good they produce—you improve their material wellbeing and/or their welfare. The fact that imposing a market benefit improves someone’s material wellbeing and/or their overall welfare gives you pro tanto reason to perform actions that impose market benefits.

Our consumption decisions impose both market harms and benefits. If you purchase good x, you contribute to demand and raise x’s price. This inflicts market harms on other buyers of x (because they must pay more for it) and market benefits on sellers of x (because they make more in profits). Thus, in determining what products to buy or eschew, we need to weigh the relative strength of these conflicting reasons. And this requires us to look at who market harms and benefits accrue to. Wilkinson writes:

I suggest that if we ever have moral reason to avoid imposing market harms (absent other considerations), we do when those harms are imposed on the poor to the benefit of the rich. Likewise, if we ever have moral reason to provide market benefits, we do when those benefits involve taking money from the rich and giving it to the poor (Wilkinson, 2022, p. 218).

A few examples will make Wilkinson’s position clear. Some foods are produced by the poor and consumed by the rich, such as avocados, bananas, cashews, cocoa beans, coffee, tea, and quinoa (Wilkinson, 2022, p. 219). By buying these foods, one contributes to demand, which raises their price. By raising their price, one takes money out of the pockets of rich consumers and puts money in the pockets of poor producers. Here, the pro tanto reason to impose market benefits on poor producers outweighs the pro tanto reason to avoid imposing market harms on rich consumers. Other goods—such as lentils, rice, soybeans, and wheat—are produced by the rich and consumed by the poor (Wilkinson, 2022, p. 218). By buying these goods, one contributes to demand, which raises their price. This takes money out of the pockets of poor consumers and puts money into the pockets of rich producers. In this case, the pro tanto reason to avoid imposing market harms on poor consumers outweighs the pro tanto reason to impose market benefits on rich producers. Thus, whether the market harm (benefit) outweighs the market benefit (harm) of a specific consumption decision depends on the overall distributional impact of that specific decision.

Market harms and benefits give us reasons, but are these reasons ever decisive? That is, do these reasons ever determine what we all-things-considered ought to do? According to Wilkinson (2022, p. 222), the answer is “yes.” The easiest way to demonstrate this is with a highly contrived case where you must eat something and have only two options: quinoa (which is produced by the poor and consumed by the rich) and wheat (which is produced by the rich and consumed by the poor). Suppose further that both options cost the same and you are indifferent between them. In this case, you have decisive reason to choose the quinoa over the wheat, because the market harms and benefits associated with purchasing quinoa have a more desirable distributional impact than the market harms and benefits associated with purchasing wheat and, by hypothesis, there are no countervailing reasons (e.g., based on price or preference) to choose wheat over quinoa. Thus, “if the agent is indifferent (or nearly so) between the two options … then it is overwhelmingly plausible that market harms and benefits are morally relevant” (Wilkinson, 2022, p. 226). Call these kinds of cases—where you are indifferent between choice options, so the only relevant considerations informing your choice stem from the market harms and benefits your choice will impose—indifference cases. Market harms and benefits are decisive in indifference cases, Wilkinson argues.

Outside of indifference cases there may be countervailing reasons that outweigh the distributional impact of market harms and benefits. Suppose you really hate quinoa and really like wheat. The distributional impact of the market harms and benefits gives you reason to choose quinoa over wheat, but your preference gives you reason to choose wheat over quinoa. Which side wins? Wilkinson does not say. That ultimately depends on other moral commitments that he does not want to take a stand on: for instance, whether there are agent-centered prerogatives and how strong these prerogatives are (Wilkinson, 2022, p. 225). But for Wilkinson’s thesis to be interesting and relevant, it must be that market harms and benefits give people decisive reasons in a sufficiently large set of non-indifference cases. This is because indifference cases are (in our estimation) extraordinarily rare in the real world; surely most people have some preference for quinoa over wheat or wheat over quinoa, or surely there will be some price difference between the two. For Wilkinson’s argument to have any practical relevance, market harms and benefits cannot only reign supreme in cases where other considerations are silent, for in practice there are few such cases. This does not mean they must be decisive in all non-indifference cases. Rather, our claim is that market harms and benefits must be decisive in an interestingly large set of such cases.

We do not think Wilkinson’s argument succeeds. We advance two claims: first, the reasons that market harms (benefits) give us to avoid (perform) certain actions are weaker than they initially appear. If we are right about this, then the number of non-indifference cases in which market harms and benefits are decisive shrinks, putting Wilkinson’s thesis in jeopardy of irrelevancy. Second, not only do we have reasons to impose market benefits, but we also have reasons to not impose them. If we are right about this, then the number of non-indifference cases in which market harms and benefits are decisive shrinks even further. Moreover, the distributional impact of market harms and benefits will not even give us decisive reasons for action in all indifference cases. Both criticisms suggest that market harms and benefits are very often not morally relevant.

Both criticisms are grounded in the same error in economic reasoning that, we believe, Wilkinson makes. Wilkinson errs by only focusing on the short-run distributional consequences of market harms and benefits, ignoring how changing prices impact the distribution of wealth in the medium to long term. For example, if you decide to add wheat to your diet, Wilkinson is concerned your added demand may raise the price of wheat, which results in greater profits for rich wheat producers and higher prices for poor wheat consumers. We do not deny that changing prices may have this kind of impact. Changing prices, however, not only impact how much consumers pay and how much producers earn but also where entrepreneurs allocate their effort and capital. As economist Israel Kirzner puts it, “market prices” are “spontaneously generated flashing red lights alerting market participants to the possibility of pure entrepreneurial profit or the danger of loss” (Kirzner, 2018, p. 30).Footnote 6 How do changing prices impact where entrepreneurs allocate their effort and capital? There are two ways, each corresponding to a criticism of Wilkinson’s argument.

Changing Prices and Market Entry/Exit

Changing prices incentivize suppliers to enter or leave specific markets (Baumol & Blinder, 1991, pp. 402–406; Gwartney & Stroup, 1990, pp. 300–303; Krugman & Wells, 2005, pp. 288; Mankiw, 2024, pp. 390–395). If the price for good x is rising, then entrepreneurs are incentivized to produce more x. This is because producing more x allows entrepreneurs to capture extranormal profits. If the price for good y is falling, then current producers of y are incentivized to allocate capital away from y, as their capital may now have more profitable uses. Continuing to produce y can result in losses when compared to y’s opportunity costs.

The fact that changing prices induce suppliers to enter or leave certain markets suggests we should expect market harms and benefits to often be temporary. Suppose many people follow Wilkinson’s advice and start buying quinoa over wheat. No doubt poor quinoa farmers will receive a temporary benefit and rich quinoa consumers a temporary harm. In equilibrium, however, higher quinoa prices will induce new suppliers to enter the market. If their climate allows it, entrepreneurial farmers and landowners will switch from whatever they are currently doing with their land to grow quinoa and capture extranormal profits. The market having been flooded with additional supply, the price of quinoa will drop. The price having dropped to normal levels, whatever market harms and benefits were bestowed on quinoa producers and consumers will dissipate.

Indeed, Wilkinson (2022, p. 203) opens his paper noting how, due to its sharp rise in popularity, the price of quinoa increased dramatically from 2005 to 2015. But Wilkinson fails to mention that this demand-driven price shock ultimately resulted in more suppliers and lower prices. Emma McDonell (forthcoming) calls this the “quinoa bust.” According to McDonell (2018), a pound of quinoa sold for $0.25 (USD) in 2000; the demand-driven shock drove it up to around $4.00 (USD) per pound, but by the end of 2014 it had fallen to around $0.60 (USD) per pound. Jeremy Cherfas (2016) writes that “The cost of quinoa started to fall in February 2014 and sank as fast as it had risen. By late 2015 the cost of quinoa was back where it was in 2012, before the price increases accelerated dramatically.” Today the price of quinoa has fallen back to roughly what it was in 2008 in nominal terms (Andrango et al., 2020, Fig. 1). This is mainly because new suppliers entered the market. Once unheard of, quinoa is now grown in the United States, in places like Oregon, Washington, and Colorado (Bland, 2012). This is all to be expected from basic microeconomic theory.

The fact that market harms and benefits will often be temporary causes problems for Wilkinson’s argument. More specifically, the transient nature of market harms and benefits implies that the reasons market harms (benefits) give us to avoid (perform) certain actions are weaker than they initially appear. We illustrate this with a thought experiment. You have the option of pressing two buttons. If you press the first button, you give a poor Bolivian an extra $100 next year. If you press the second button, you give a poor Bolivian an extra $100 per year for the next fifty years. Clearly, you have reason to press both buttons. Yet, it’s clear that your reason to push the second button is far stronger than your reason to push the first. This is because the benefit you give to the poor Bolivian when you press the first button is short lived compared to the benefit you give the poor Bolivian when you press the second button. All things equal, the more temporary the benefit we can bestow on someone, the weaker is our reason to bestow the benefit.

So, because market harms and benefits are transient, our reasons to impose them are weaker than initially estimated. How does this affect Wilkinson’s conclusions? It doesn’t change the fact that, in indifference cases, the distributional impact of market harms and benefits give agents decisive reasons. This is because, by hypothesis, there are no countervailing reasons in indifference cases. Although market harms and benefits give us weaker reasons than once thought, weak reasons still beat no reasons.

However, if market harms and benefits give us weaker reasons than initially thought, it reduces the number of non-indifference cases in which market harms and benefits are decisive. Countervailing considerations are present in these cases. For instance, while the distributional impact of market harms and benefits says you should buy quinoa over wheat, you might prefer wheat over quinoa, or perhaps wheat is far cheaper than quinoa. As already mentioned, Wilkinson does not say precisely when market harms and benefits are decisive in these sorts of cases, but they must be decisive in a sufficiently large set of them for his thesis to be interesting, for real people in the real world rarely find themselves in indifference cases. Yet, if the reasons that market harms and benefits give us are relatively weak due to their transient nature, the implication is that countervailing reasons need not be all that strong to defeat them.

An example will make this point clear. Suppose you can’t stand quinoa, but love wheat. Suppose also that quinoa is a lot more expensive than wheat. The distributional impact of market harms and benefits says you should buy quinoa over wheat. Your palate and wallet say you should buy wheat over quinoa. Suppose there is something deeply wrong with the market for quinoa, rendering the supply of the grain highly inelastic. Supply cannot meet growing demand. Thus, if you continually buy quinoa over wheat, then you continually contribute a higher price and continually transfer money from the pockets of rich consumers to poor producers. Under these conditions, it’s plausible to think the long-term distributional benefits outweigh your palate and wallet.

Now suppose the supply of quinoa is perfectly elastic or nearly so, in that increased demand and higher prices result in new suppliers entering the market, which then drives prices back down to initial levels or somewhere thereabouts. In this scenario it is less clear market harms and benefits are decisive. While poor Bolivian farmers might receive a wealth transfer for a year or two, this rent will eventually dissipate, as new suppliers enter the market and bid down prices. It’s not crazy to think that, in cases like this, market harms and benefits are outweighed, and your palate and wallet will win the day.

To sum up our first criticism of Wilkinson, once we account for the fact that changing prices induce suppliers to enter or leave specific markets, the reasons market harms and benefits give us appear weaker than initially thought, because market harms and benefits are (in well-functioning markets where supply is sufficiently elastic) transient. This is troubling, because (all else equal) it reduces the number of non-indifference cases in which market harms and benefits give us decisive reasons. If this set of cases gets too small, then Wilkinson’s argument becomes uninteresting and irrelevant.

Rising Prices and Substitutes

There is a second way in which changing prices influence where entrepreneurs allocate their talent and capital, corresponding to a second problem with Wilkinson’s argument. Rising prices not only encourage new suppliers to enter specific markets, but they also incentivize entrepreneurs to innovate substitutes (Zamagni, 1987, p. 201; Samuleson and Nordhaus 1998, pp. 87–88; Mankiw, 2024, pp. 66–67). Substitutes are goods and services that can be used in place of each other to satisfy particular needs or wants. Typical examples include coffee and tea, butter and margarine, cigarettes and vapes, and public transportation and private cars.

Rising prices for good x incentivize entrepreneurs to not only replenish the supply of x but also to find substitutes for x. This is especially likely if the supply of x is, for some reason, highly inelastic, so prices remain high for an extended period of time. As an example of this, in England in the 1500 s wood was the primary source of fuel. Due to overharvesting a shortage ensued, which dramatically raised the price of wood, wood became so scarce, and prices rose so much, that Parliament banned the building of new foundries in 1580 and passed a law in 1588 that prohibited cutting down trees for industrial purposes (Simon, 1996, p. 169). Although at the time “the use of coal in place of charcoal had been known, there were technical difficulties” that prevented its widespread adoption (Simon, 1996, p. 169). However, the “wood shortage exerted pressure that led to the development of coal as well as blowing machines to be used in smelting, a keystone in the upcoming Industrial Revolution” (Simon, 1996, p. 169). The high price of wood incentivized innovation that led to the adoption of coal.

As another example, whale oil was the primary illuminant in the 1800s, but prices rose in the 1840s and then even higher during the Civil War. These rising prices “provided incentive for enterprising people to discover and produce substitutes” (Simon, 1994, p. 27). Initial substitutes included oil from rapeseed, olives, linseed, pine trees, and coal. The ultimate substitute, however, was crude oil, which was used as an illuminant in the form of kerosene and, it turns out, had many other important uses as well (Simon, 1994, p. 27). The high price of whale oil incentivized innovation that led to the adoption of crude.

There is some evidence that the market for quinoa may soon be replete with substitutes. Consider two different types of substitutes. One type of substitute comes in the form of new and improved varieties of quinoa. In the United States, three new varieties have recently been bred through a collaborative venture by researchers at Washington State University and Brigham Young University (Truscott, 2023). The three strands are superior in that they can grow in a broader range of conditions and have higher yields. This may shift quinoa production out of places where it currently benefits poor producers (such as Bolivia) and into places where it benefits rich producers (such as the United States and Europe), as agriculture tends to be more productive in richer countries. The second form of substitute is a genuine alternative that possesses similar properties, such as sorghum––a tiny, beige, gluten-free whole grain that has been described by some as “the new quinoa” (Largeman-Roth, 2016). Sorghum is “the cheapest of the ancient grains to cultivate, … grows well in dry conditions where the soil isn’t so great,” and has a similar nutrition profile to quinoa (Largeman-Roth, 2016). High quinoa prices may induce entrepreneurs to advertise sorghum or similar grains to rich Westerners as a healthy, tasty, and cheaper alternative to their expensive quinoa habit.

The fact that rising prices incentivize the search for substitutes causes problems for Wilkinson’s argument. Suppose you are trying to help a specific group of producers by bestowing a market benefit on them via higher prices. Microeconomic theory tells us that you are also incentivizing entrepreneurs to innovate substitute products that, if successful, may cause grave harm to the initial group you were trying to help. An example will make this clear. Suppose you live in 1500 s England. Suppose timber producers are poor and timber consumers are rich. You use as much timber as you can to drive the price up, transferring wealth from rich timber consumers to poor timber producers. Because timber takes a long time to produce, supply is replenished slowly, and so the price of timber remains high. While this is good news for poor timber farmers in the short run, it incentivizes entrepreneurs and engineers to refine the use of coal as a fuel source in the long run. They eventually succeed. Not only is coal a substitute for timber, but it is also superior. Everyone eventually switches to coal. The market for timber collapses. In the short run high prices made poor timber farmers better off, but in the long run it led to their demise.

Wilkinson argues that we always have pro tanto reason to bestow market benefits, because doing so improves people’s material wellbeing. We now add that we have pro tanto reason to not bestow market benefits on producers specifically, because doing so incentivizes competitors to search for substitutes which, if successful, may reduce the wellbeing of the producers we were trying to help. Is one of these reasons stronger than the other? That depends on the likelihood that higher prices will result in the successful innovation of a substitute good. If it is very unlikely that this will happen, then perhaps the higher prices for producers outweighs the possibility of obsolescence. At some point, however, the likelihood that higher prices will result in the successful innovation of substitute goods becomes large enough such that the possibility of obsolescence will outweigh higher prices for producers in the short run. How to assign probabilities here is a difficult question we sidestep. Our point is that there will be some cases in which the reason to not bestow a market benefit outweighs the reason to bestow it.

Higher prices incentivize a race for substitutes which gives us a reason to not bestow market benefits on producers. How does this affect Wilkinson’s overall argument? It does so in two ways.

First, it further reduces the number of non-indifference cases in which the distributional impact of market harms and benefits give us decisive reasons for action. Let’s return to a non-indifference case considered above. You hate quinoa, love wheat, and wheat is far cheaper than quinoa. Your palate and wallet tell you to choose wheat over quinoa. The distributional impact of market harms and benefits says to choose quinoa over wheat. Above we argued that the reason to buy quinoa is weaker than it initially appears because, if the supply of quinoa is elastic, quinoa producers will only receive a temporary benefit from higher prices. This makes it more likely that your palate and wallet win the day. To this we add a new reason, higher prices for quinoa producers incentivize entrepreneurs to innovate substitute grains; if they are successful, this could devastate quinoa farmers. This makes it even more likely that your palate and wallet win the day.

Second, the incentive effects higher prices have on the search for substitutes may render the distributional impact of market harms and benefits indecisive in some indifference cases. This depends on the likelihood of higher prices inducing successful innovation. Consider once again Wilkinson’s indifference case: you can buy wheat or quinoa; both cost the same and you are indifferent between them. Because there are no countervailing reasons, Wilkinson says the distributional impact of market harms and benefits are decisive: you should buy the quinoa. Incentivizing the race for substitutes introduces a countervailing reason. By buying quinoa you not only transfer money from the pockets of rich quinoa consumers to poor quinoa producers, but you also incentivize entrepreneurs to seek substitutes that could devastate poor quinoa producers. Whether this countervailing reason wins the day depends on the probability we assign to this happening. But there is some probability assignment such that the distributional impact of market harms and benefits won’t even be decisive in the easy cases.

To sum up our second criticism of Wilkinson, once we account for the fact that rising prices incentivize the search for substitutes, it becomes clear that we have reason to not bestow market benefits on producers. This reason further reduces the number of non-indifference cases in which market harms and benefits give us decisive reasons, placing Wilkinson’s argument in even greater jeopardy of irrelevancy. Moreover, the reason to not bestow market benefits on producers means that the distributional impact of market harms and benefits won’t even be decisive in all indifference cases.

To conclude our criticism in this section, we do not dispute Wilkinson’s claim that market harms and benefits give us reasons to avoid or perform certain actions. Our claim is that, once we account for the equilibrium effects of changing prices—that is, how changing prices affect the distribution of wealth in the medium to long term via incentivizing entrepreneurial activity—these reasons will be decisive in very few cases. Wilkinson’s argument has little practical relevance. Go ahead, buy the wheat.