Welcome to our discussion of Hayden Wilkinson’s “Market Harms and Market Benefits”! Kian Mintz-Woo has graciously provided a critical précis, which appears below. Hayden will offer an initial comment, and then all are encouraged to join in the discussion.
* * *
Some philosophers have influentially attempted to guide choices in charitable giving so as to increase (or even maximize) the impartial benefit of that giving. In short, if we choose to give in ways that are more effective, then those choices can do more good. In Hayden Wilkinson’s exciting new paper, Wilkinson extends similar reasoning to everyday supermarket purchases. The gist of the piece is: if we can choose some staple foods in an informed manner—information the paper includes!—there are weighty (perhaps decisive) moral reasons to select the foods that will contribute to more just global distributions.
In line with the remit of Philosophy and Public Affairs, this paper offers something to both the interested layperson and the trained philosopher. I recall the quinoa panic amongst my friends which Wilkinson discusses; the concern was that the trendiness of quinoa would make it unaffordable to those in Bolivia and Peru who have eaten it for a long time. At the time, I felt a vague unease that there were surely more complex effects than just that some were unable to afford quinoa. Wilkinson’s paper helpfully provides a framework for understanding the twin effects of price changes, effects in terms of both those who buy (or demand) the food and those who sell (or supply) the food. In other words, the extra profit to those selling the quinoa after westerners demand more means a higher quality of life and greater welfare to those selling. This kind of discussion is helpful to anyone wishing to be an informed consumer.
Before I get to some objections, I will sketch the argument. Wilkinson points out that there are three possible outcomes when one purchases a good: it is purchased and there is still inventory left over with no other action by the seller (small benefit to the seller, no other major effects); it is purchased and there is no inventory left over with no other action by the seller (small disbenefit to someone else who would have liked to purchase the good, no other major effects); or it is purchased and the seller increases the price—perhaps slightly more likely when no inventory is left over, but possible regardless of inventory levels (small subsequent disbenefit to each of the future purchasers, significant subsequent benefit to the seller). The issue is how often (and how much) a given purchase can make a difference to the price of that good. Some philosophers have been skeptical that individual actions can be morally relevant difference makers in contexts with a large number of actors whereas other philosophers have argued that some actions make a difference in expectation, so there are morally relevant expected harms caused by individuals.
Wilkinson takes the latter side, arguing that these price changes do occur at some thresholds where supply curves meet demand curves. While it can be objected that these curves are actually step functions (where the ‘curves’ are actually just steps with horizontal sections punctuated by vertical climbs), Wilkinson makes the sensible point that in global markets, most foods are traded by so many people that all their individual step functions will combine to make international functions that are much closer to a smooth curve than for individual demand or supply functions. In other words, although Wilkinson takes the expected impact to be morally important, in the cases of globally traded foods like quinoa we should think that the actual impact is often the expected impact (or close to it).
The curvature of these supply and demand functions is given by what economists call the demand and supply elasticities: the percentage change in demand or supply in response to a percentage change in price. When the elasticity is higher (more elastic), a change in price leads to larger shifts in how much people buy or sell that good (more precisely, how much is produced or consumed). When the elasticity is lower (more inelastic), a change in price leads to smaller shifts in how much people buy or sell that good. One thing economists do is try to estimate these elasticities empirically by looking at how much buying and selling is done in response to historical price shifts.
One of the valuable contributions of this paper is bringing together some estimates of elasticities for goods that many people will find in their supermarket everyday (wheat, quinoa, coffee, bananas, etc.) and another contribution is to bring together some idea (at a very rough national level) of how those elasticities would affect globally poorer versus globally wealthier people. The idea is that if the foods are predominantly consumed in wealthy countries, wealthy countries benefit if the price goes down; if the foods are predominantly made or grown in poorer countries, poor countries benefit if the price goes up). When all of these values are small (or ‘marginal’), we can approximate the effects of increasing the demand for these foods in a simple equation that depends on just these elasticities and the amount you spend. When we put all of these points together, they give us a rough idea of which food price increases are good for the global poor (Wilkson reports that this applies to foods like bananas, quinoa, tea and coffee) and which are good for the global rich (Wilkinson reports this applies to foods like lentils, rice, soybeans and wheat). Wilkinson is fairly convincing that this conclusion is robust in that it could be justified for theorists of different types (I would be interested in conversation below about whether those from different theoretical starting points are convinced). To me, the most valuable contribution of the paper is in bringing some of these concepts into philosophical discussion—that isn’t to say others have not, but it is to say that they are still uncommon and I hope this paper leads more people to think in these kinds of quantitative and empirical terms in order to draw out important practical conclusions.
I am largely in agreement with Wilkinson’s points, and I don’t think it very productive to critique the coarseness of these claims (obviously, wealthy and poor countries alike include significant heterogeneity); I think the steelman of the project is an invitation for others to provide more empirical evidence and expand the scope of goods to try to allow for more informed choices with this paper as a strong first pass. This conception of the project is very similar to the attempts to systematically compare charities in order to improve choices about charitable giving. However, there are three points I want to press on in this precis. The first relates to reasons and rights; the second asks what whether these reasons can be motivational in light of observed preferences; and the third presses the role of prerogatives.
The first point relates to the way Wilkinson situates the paper. Since there is relatively little philosophical discussion of market harms, he considers John Stuart Mill and Judith Jarvis Thomson. They say something like it infringes no right to take an action that leads to market harms. While I understand that few have discussed market harms in the way Wilkinson does, I do not think that his claims really refute these: Wilkinson’s conclusion is that there is a weighty (perhaps decisive) reason in favor of buying foods that have the effect of transferring resources from the globally wealthier to the globally poorer. That does not establish any rights violations in failing to do so; it is at least deontically consistent with their claims. I don’t think that Wilkinson has to establish that they are wrong (and he may be happy to admit this); however, since he sets these up as his interlocutors, it could seem to a reader that he intends to.
The second point has to do with motivation and internal and external reasons. The simple version is that Wilkinson argues that there are reasons to choose items that transfer resources from the globally wealthier to the globally poorer, but this seems to be an external reason in Williams’ sense. An internal reason is one that an individual can be motivated to do given their antecedent goals or motivational sets; an external reason is one that has no necessary connection to anyone’s motivational sets.
In particular, many actual people reveal through their behavior that they would not have altruistic motivational sets. For this reason, few would be moved by this argument. I suppose that the primary audience of Wilkinson’s piece are wealthy consumers (or at least wealthy in a global context). But this audience reveals its preferences through their actions. Most of these preferences seem to be self-interested, or to only apply to those nearby. Many people do not donate in generally altruistic ways. Most of those who do donate do so in ways that reflect their personal projects (e.g. diseases their friends or family members have had, arts giving that is important in certain social spaces, funding wealthy universities from which they graduated as opposed to poorer colleges that might have greater need), not in terms of improving the world impartially. Let us grant that if they bought $10 worth of bananas, that action would shift roughly the same amount in expectation globally from some who have more to some who have less. But given that people do not demonstrate interest in helping those around the world, this does not seem very motivationally relevant. So it seems that, at the very least, there are a few more steps needed to make this fact an internal reason for them. Perhaps Wilkinson is satisfied with merely providing evidence for an external reason, but I would guess that he wants more, so I would be interested in hearing what he thinks about how to change people’s motivational sets (or if he agrees that most people are not altruistic).
The final point is perhaps the most substantive. Wilkinson considers an objection (the first in his list of objections) that there are agent-centered prerogatives, asking where what principled basis could they have (and attempting to reductio ad absurdum that people’s eating habits could be morally innocuous by considering food grown by slaves). While I do not myself agree with common-sense morality, I do think that this morality does hold shopping, clothing and eating habits (and perhaps choices about having children) to be in a private sphere untouched by morality, perhaps excepting extreme circumstances. I suspect that many, perhaps the bulk, of philosophers might also endorse some version of this view.
Wilkinson responds that prerogatives are most plausible when it is extremely demanding to avoid the action in question, but I do not think common-sense morality adopts that position. For instance, suppose that you know that everyone in your office likes red and if you wear a red shirt everyone will have a somewhat more pleasant day; I believe the common-sense view is that you are not obliged to wear red (or really even to take into account everyone’s preferences about what you wear)—you can wear exactly what you want. This permissibility is the common-sense view even though it is not extremely demanding for you to wear red (or to refrain from wearing red).
Similarly, even when it is not demanding for you to select bananas over apples or quinoa over wheat, I believe common-sense morality does not oblige you to consider the market harms (or benefits) these choices could make. These personal choices are outside the sphere of moral concern.
Wilkinson attempts to rebut this point by saying that if such choices were outside the sphere of morality, then it would be morally permissible to eat factory farmed meat or to buy products that were made with slave labor. But from a common-sense point of view, this is an unconvincing slippery slope. The market harm one is contributing to in Wilkinson’s case is something along the lines of $10-50 being distributed from the globally wealthy to the globally poor. The fact that we could imagine decisions with sufficiently important side effects that duties of beneficence or non-maleficence apply does not undermine these choices being within our personal sphere free of moral concern. There are several sophisticated ways that common-sense morality could justify this position. For instance, as when making most other eating and purchasing decisions, those global transfers might be good side effects, but they are not intended and they are legitimate to disregard. Or the fact that there is intervening agency in the marketplace means that you are not morally responsible for these outcomes or effects.
A small anecdote: a distant family relative once said that, while she was comfortable buying meat from the butcher and eating it, she would be somewhat critical of anyone who married a butcher. The butcher, I believe she meant, is doing something morally suspect or betraying a bad character, but the fact that she pays him to do it did not tie her close enough to that suspicious or vicious activity to be morally responsible. In line with this kind of reasoning, it might be good to transfer $20 from someone (or someones) who need it less to those who need it more, but common-sense might suggest that you do not get the moral credit if that transfer is mediated by the market—even when you incentivized that transfer to happen yourself.
With those three critical points, I welcome discussion both by the author and by anyone else on this excellent paper.
Many thanks to Kian for this excellent summary of the paper and points of critique, as well as to David Faraci and Peter Jaworksi for organising this discussion!
I agree that each of the three points that Kian raises has some bite, but I’ll briefly comment on each.
The first point was that I don’t establish that any rights violation occurs when one imposes a market harm, and thereby don’t show that the claims made by either Mill or Thomson are false. It’s true—I don’t show that a rights violation takes place (at least not if we don’t each have a general right against being greatly harmed). The case for conferring market benefits and avoiding market harms is stronger if we are concerned with the wellbeing of those affected, or with distributive justice, or with virtue and vice, or with whether a norm allowing such behaviour can be reasonably rejected, or some such other moral consideration; less so if we are concerned only with avoiding rights violations. (Maybe such a case could be made, but I at least don’t make it.)
As for what Mill and Thomson say, it is not entirely clear whether they’d genuinely disagree with me. Thomson, at least, says nothing specific about the permissibility of imposing market harms. But I interpret Mill as suggesting, somewhat implicitly, that it is generally permissible to impose them (all else being equal). After all, he says that it’s “better for the general interest of mankind, that persons should pursue their objects undeterred by this sort of consequences.” Given his broader (utilitarian) views on the connection between betterness and permissibility, I think this is evidence enough that he considers it permissible to impose market harms, all else being equal.
The second point was about the sort of reasons I present for worrying about market harms and benefits. To many agents, those reasons may be merely external, not internal ones that will have any effect on their motivational sets. In the abstract, I do think that critiques like this are important—it seems desirable that papers in applied ethics have some effect on the behaviour of real-world agents. But I suspect that this paper’s prospect of doing so is quite good. Other forms of ‘ethical consumerism’ are already widely practiced among wealthy consumers—e.g., boycotting animal products to avoid harming farmed animals, attempting to ‘buy local’, foregoing carbon-intensive products to avoid harming future generations, and boycotting sweatshop-produced goods to improve working conditions for the poor. And there are at least some individuals who are (at least partly) motivated by impartial altruism to take drastic actions—the effective altruism community especially. It might be true, as Kian suggests, that *most* people are not altruistic. But even if only a *small portion* of consumers are so motivated, and so the paper gives them the required internal reasons, that could easily still mean tens of millions of dollars transferred to the global poor.
The third point was about agent-centred prerogatives. I claimed in the paper that it’s only plausible that we have such prerogatives to impose market harms (or forego market benefits) when it would be costly to do otherwise. But you might think that there are certain classes of decisions—of what to buy, of what to wear, of what to eat—which fall within a private sphere, in which we have general prerogatives to do as we wish, at least in non-extreme circumstances. Such prerogatives seem fishy to me, but others might find them far more plausible. (If any readers do endorse them, I’d be keen to hear how they’re justified, and whether you find the following convincing!)
If we do endorse them, there’s the question of how we identify the extreme circumstances where they are overridden. If the destruction or survival of humanity rides on which shirt you wear or which fruit you buy, it seems clear that no such prerogative permits you to cause destruction. But is this because the stakes are enormous in the *absolute* sense, or because the stakes are enormous *relative* to some feature of the action? (Perhaps relative to the costs you bear in taking that action.) The latter seems much more plausible. For instance, suppose you accept that eating large quantities of meat is wrong due to the harms thereby imposed on farmed animals—that those prerogatives to buy and eat as you wish are overridden due to stakes involved. But if you eat just the tiniest amount of meat, the absolute stakes will be small. But it would be a strange moral view that permits you to eat tiny amounts of meat but not large quantities (especially since eating many small quantities eventually adds up to a large quantity!). So, it seems far more plausible that it is not the *absolute* stakes but the *relative* stakes that determine whether such prerogatives are overridden. And, if you buy even a small quantity of wheat or of quinoa, the relative stakes are high! Relative to the amount you spend, quite a lot of money hangs in the balance for the global poor. And relative to the cost imposed on you in switching from one good to the other, the benefits are enormous (at least in the cases I focus on).
Then there’s the related issue that common sense suggests that we are morally permitted to do harm (or fail to benefit) when doing so is mediated by the market, or by the voluntary action of another agent. I think common sense gets it wrong here too. We can imagine a wealthy person paying an assassin to kill an enemy of theirs, thereby distancing themselves from the vicious activity. It seems overwhelmingly plausible that they act still wrongly—that any prerogative they have to buy and sell as they wish is overridden—and, indeed, that the intervention of the transaction and other agent make no difference at all. (You might think that this is merely because the absolute stakes in such a case are so high; again, it seems impermissible to pay an assassin to even slightly harm someone.)
On that last point in particular, I don’t expect my responses to convince everyone. Nor do I think all possible moral views will agree that we are morally required to buy more quinoa and bananas. But, at the very least, I think we do have moral reason to do so according to a wide range of plausible views. And that, by itself, is a pretty surprising result.
Thank you for this novel and very interesting intervention in the consumer ethics literature, Hayden.
Following up on Kian’s point about prerogatives. I think quite a lot rests in your argument on what you call your “expectational” view that, if I understand rightly, creating a 1 in 1000 chance of doing a harm of value 1000 is as bad as doing a harm of value 1 for sure.
After all the difference between risk and direct harm is one big difference (beside intentionality) between the person paying the assassin (in your response to Kian) and you and I buying wheat bread. We might be tempted to treat unintentional, expected harm might be treated very differently from intentional, actual harm.
You consider what might happen when you relax the expectational view. You claim it doesn’t affect your argument because “the most plausible alternatives to [the expectational view] assign greater importance to reducing the probabilities of worst outcomes.” (section VI.G)
I’m not sure this is correct, for your purposes, which is to discuss individual morality, not individual prudence or public policy. You cite Buchak, Arrow and Williamson and Bottomley, and note that their views could be applied to our effects on others, as well as our effects on ourselves, and that you know no philosopher that defends “risk seeking”. I’m not an expert in the morality of risk by any means, but the range of relevant alternative views to the Expectational view seem much more diverse to me than you suggest.
I would say that common-sense morality holds that we *can* discount creating very small chances of great harm. Since almost everything interesting we do has some risk of being part of a disaster, to count expected harm as equally bad as a similar quantum of sure-thing harm would probably be paralyzing. An extreme case: assume that (due to the role of technology in graphics processing cards in driving AI research) there is a tiny chance that my purchasing a graphics card will be the trigger for an AI apocalypse. Should we actually treat that expected harm as a reason against buying the card? I would say plausible views (perhaps influenced by common sense intuitions) would say there is some threshold of risk that we can just ignore.
Relatedly, Garret Cullity has an interesting argument in his 2019 Monist paper “Climate Harms”. Because increases in ambient noise appear to raise the risk of violent assault, we might think [on an expectational view] that the expected harm of contributing to a violent assault should be weighed equally with other moral reasons in our decision whether or not to mow our lawn. But, Cullity argues, treating expected outcomes as equal to (some fraction of an) a corresponding actual outcome ignores that typically, when we cause [small] actual outcomes, we are more often directly responsible than when we happen to be the unfortunate trigger for some unlikely but bad actual outcome. Dale Jamieson makes a similar argument in Reason in a Dark time.
I think Cullity and Jamieson would ask whether you, as a consumer, are responsible for the bad outcome that occurs if and when your purchase of wheat bread causes bread prices to rise for the global poor. That’s a much more complicated question than just running the math on expected outcomes, I think, and raises deeper questions about the appropriate response to global income and wealth inequality.
But in the broadest sense, I’m interested to hear more from you about why you are confident that “the most plausible alternatives to [the expectational view] assign greater *moral* importance to reducing the probabilities of worst outcomes.
Thanks for joining the conversation, Ewan! I’ll let Hayden answer some of the more substantive points, but I thought I’d add two potentially helpful details.
First, with respect to simple expected utility (probability weighted outcomes), Hayden definitely rejects your position that we can discount very small probabilities to zero and endorses simple expected utility. In fact, he has done so in print (Ethics, vol 132, n. 2, https://doi.org/10.1086/716869, doi:10.1086/716869 ), defending the ‘fanatical’ view that you should always multiply the value of the outcome by the probability of the outcome, however small that probability is. Basically, he argues that discounting very small probabilities to zero leads to several even more objectionable or implausible conclusions.
Second, in terms of relaxing the view, I assume he’s thinking of alternative shapes or patterns of justice in distributive justice, where the dominant ones (e.g. prioritarianism, sufficientarianism, egalitarianism) all overweight the harms to the worst-off compared to utilitarianism. (There is some complexity here because the relevant good being distributed varies a bit between these theories so there is some talking past each other, but assuming the numeraires can be held fixed, I think that his claim is plausible.)
Thanks for this fascinating paper, Hayden, and thanks, Kian, for alerting me to it and to this discussion!
I find the central argument convincing. I wonder though if the much more urgent ethical problem is the role of corporations in shaping supply and demand rather than the behaviour of individuals. You have convinced me that, other things being equal, that there is a moral onus on individuals to choose goods that contribute to the more beneficial wealth transfer. But does this framing of the ethical issues elide the massive global corporations who devote vast resources to ensuring that other things are not equal, that some choices are significantly easier and more appealing than others, and it is not clear that these corporations recognise any inherent value in the transfer of wealth from rich to poor. Indeed, in practical terms, they seem committed to the opposite.
You point out that, for many people, the harm of having to make the ethical purchasing decision is likely to be neglibible as the necessary information is readily available. If you are a well-off person with a PhD shopping in Whole Foods, the ethical choice might indeed be relatively easy, but I’m not sure that’s equally true if you’re a working class Irish person shopping in Aldi, surrounded by marketing material telling you how you’re supporting Irish jobs by buying this yoghurt or that steak.
In On Liberty, Mill is clear that manufacturers and sellers have no right to unfettered activity qua manufacturer or seller, only those they already have qua private individuals. Presumably then they also don’t have moral responsibilities over and above those they have as individuals. So the duty to make profit for shareholders, say, is a legal duty and cannot trump the moral duty to process and market products in the least harmful and/or most benefical way. If so, individuals working for businesses that act as go-betweens for producers and consumers are also subject to the same moral responsibility you (rightly) attribute to individual consumers. If they fulfilled this responsibility, it would become much easier (or even unnecessary) for consumers to fulfill their parallel responsibility.
As I said, not an objection to the central argument, but I’d be curious to hear your thoughts. Thank you!
Thanks for your thoughtful contribution, Mags, which I think would reflect the views of many philosophers.
One possibility is that Hayden is identifying a way that individual action might go beyond what we would expect. Another possibility is that the question of non-individual (social/political/etc.) action is meant to be beyond the scope of this paper. A final possibility is that the discussion of what social actions are needed is mature but this discussion of individual reasons for buying in certain ways is very young, and he is trying to seed a new discussion, so let a thousand flowers bloom. (I suspect he would endorse each of these points.)
One other small thing: it’s not even as easy as you say in the easy case. In particular, there might be moral reasons against buying at Whole Foods as well. For instance, if you are pro-union, then Whole Foods’ current ( https://www.businessinsider.com/whole-foods-tracks-unionization-risk-with-heat-map-2020-1?op=1&r=US&IR=T ) and continuous ‘post-union’ (i.e. anti-union) ( https://www.theguardian.com/business/2018/sep/27/amazon-whole-foods-training-video-union-busting-efforts-staff ) stances might not align with your values. If you are for public healthcare, Whole Foods’ stance against public healthcare ( https://www.nbcnews.com/business/business-news/whole-foods-ceo-regrets-comparing-obamacare-fascism-flna1B8022914 ) and against worker healthcare (e.g. https://www.cnbc.com/2021/01/04/whole-foods-ceo-john-mackey-best-solution-is-not-to-need-health-care.html ) might also be objectionable.
Thanks, Ewan!
There are two separate problems here, I think. One is whether intentional and unintentional harms should be treated the same. The other is whether expected and actual harms should be treated the same.
On the first, I think it’s true that market harms won’t be much of a worry under a view by which unintentional harms don’t bear on the permissibility of an action. If you accept (a permissive version of) the Doctrine of Double Effect, for instance, then the fact that buying wheat harms the poor need not give you reason against buying it. But, that said, I think that views like this are generally implausible. They would say the same about actions with other sorts of harmful side effects (e.g., increasing carbon emissions).
On the issue of how expected versus actual harms, I agree that they very plausibly come apart wrt their moral significance. (Buchak and co describe views on which that’s true.) For instance, it would make sense to give less weight to tiny probabilities of enormous benefit. But I don’t think it’s plausible to discount or ignore *all* tiny probabilities, including tiny probabilities of enormous harm. On ignoring tiny probabilities, I’d recommend my paper on fanaticism that Kian linked above (“In defence of fanaticism”), as well as a recent paper by Beckstead & Thomas (“A paradox for tiny probabilities and enormous values”).
In short, here’s the sort of case that makes me think that we can’t just ignore sufficiently tiny probabilities. You’ve got two options: 1) bring about a pretty good outcome (with no benefit or harm to anyone) for sure; and 2) bring about that same pretty good outcome with probability 0.99999, and another outcome in which you impose enormous, gratuitous harm with probability 0.00001. (If you think 0.00001 isn’t sufficiently small, replace it with as small a probability as you like.) Intuitively, I think, Option 2 is better (perhaps not *much* better, but at least somewhat better); and, if you must decide between them, you’d be required to choose Option 2 if there was no cost to you in doing so. If common sense morality says otherwise, it seems like a case of common sense getting it wrong.
In the case of buying a graphics card or slightly raising the ambient noise, you might think it’s much less plausible that we’re required to avoid the risk of harm. But that could be explained by the cost to yourself of foregoing the better graphics card, or of not enjoying some music in the evening. And perhaps that cost to yourself is greater, in expectation, than the harm you thereby impose. (Or perhaps it’s lower, but you weigh costs to yourself more heavily.) I think it’s much more plausible that what’s happening is those cases is that the risk of harm is being outweighed by another moral consideration, not that it has no weight at all. And as long as the small probability of harm has *some* weight then, in cases where the other consideration isn’t present–e.g., Option 1 versus Option 2, or the decisions I describe in the paper–it’ll give you decisive reason to choose one way or the other.
Thanks, Margaret!
Basically, I agree! The stakes are much higher for corporations making decisions about what goods to produce and how to influence demand via, say, marketing. The reasoning in this paper applies in much the same way to show that corporations impose pretty enormous harms and benefits when they get involved in or influence the market for any good that’s produced in a poor country and consumed in a rich country, or vice versa. If a corporation releases a new breakfast cereal that’s wheat-based rather than quinoa-based, then (in expectation) they transfer money from poor to rich. Or if they spend money on advertising a wheat-based product, they do the same. Or if they open a banana farm and thereby push down the price of bananas, etc. By doing any of those things, they transfer far more than any individual consumer choosing their groceries each week. So, as a matter of business ethics, I think there’s a clear case to be made that they often act wrongly (and don’t at all escape moral responsibility).
There’s the other issue of how obligations differ between consumers in different situations (the wealthy Whole Foods shopper versus the working class Aldi one). It seems pretty plausible that the relatively poor shopper isn’t obliged to buy the quinoa, since it will often be costly (in terms of wellbeing) for them to do so. Likewise if a shopper doesn’t have access to this information about market harms/benefits, or information about how the supermarket’s marketing material is misleading (both empirically and normatively). I suppose, at least in part, it’s our job as ethicists and educators to change that!
Thanks Mags, for your thoughtful comment. Two quick thoughts.
I suppose Hayden might be willing to grant that societal or social or political responses to global unjust distributions are not dealt with. Even so, I see at least two conciliatory replies to this point he could offer: (1) he was trying to show something non-obvious about the benefits and harms of individual actions; and (2) social and political responses to global maldistribution are well-established in the literature, so he was trying to help seed a new literature about consumer ethics–so let a thousand flowers bloom.
One small point about your easy consumer ethics case: the wealthy person buying at Whole Foods. This could also be challenged, and there could be moral reasons to refrain from shopping there. For instance, if you care about unionization, it might be disconcerting to know that it is “post-union” (read: anti-union) ( search the Guardian for “Whole Foods union busting training video”, since links seem to make this unpostable ). Or if you care about healthcare provision, Whole Foods has been against public healthcare ( search for “Whole Foods CEO regrets comparing Obamacare to facism” ) as well as healthcare for employees ( search for “Whole Foods CEO: best solution is not to need health care” ). So even your easy case is less clear than one might expect.
Thanks Kian for that addition. I’ll have to look at the Ethics paper! In the meantime, Hayden, I’d like to ask you about another point, about demandingness, which echoes Margaret’s point above.
I agree choosing the quinoa and bananas in your scenarios as an act is not very demanding. On pp 28-29, you raise the important potential objection that figuring out which goods one has most moral reason to buy might be demanding (call this epistemic demandingness). You respond that your view is not very epistemically demanding – it only requires consumers to read your paper! :). Like Margaret, I think the demands on consumers might be more epistemically demanding than you think, even for a relatively well informed consumer with significant options. This is especially so if we care about the magnitude of the impact and not just its direction, which I think we should.
First are general worries: Is the formula a good model? Given that this is published in a philosophy journal, a conscientious and practically minded reader might wonder whether it has been reviewed by economists or not, and why economists are not making a similar argument (or are they?). Are there long-run unintended consequences of increasing sales of goods primarily grown in poor countries and ultimately consumed in rich countries?
There are more specific concerns about the size of the impacts. Are the elasticities truly representative of actual markets? You note that for bananas, only 12% of A will reach the grower, with much of the rest soaked up by actors in the supply chain, many of who will not be very poor, or in poor countries. Is this 12% higher or lower than for other goods? And for the market harms (rather than benefits) a similar question arises: how much does my purchase of a loaf of bread actually influence the price of a loaf of bread in Egypt, given that they are linked mainly by the global price of unprocessed wheat, which is a fraction of the cost of the bread.
Why would a consumer care about the size of the impact? Your footnote 37 suggests why. Once we start making ethical considerations part of our deliberation about what to buy, a huge range of factors start to seem important. You mention the presence of forced labor in the supply chain of coffee and cocoa as a reason we should not try to use your approach for those goods. But forced labor is surely just one consideration conscientious consumers want to make. If conscientious consumers are considering market harms and benefits and complicity in forced labor, why would they stop there? Shouldn’t they also consider carbon footprints, water footprints, nitrogen footprints, whether goods are union-made, and so on? What about as Margaret raises, the status of the company distributing the product? Consumers might wonder whether they have problematic political views or histories or not?
In order to balance the market harms and benefits from these more standard ethical consumerism concerns, I could imagine a conscientious consumer wanting to be much more sure about the size of the expected market harm/benefit impact in these cases than you see them to be. (It’s perhaps rare that a conscientious consumer will be actually indifferent between the wheat and the quinoa, or bananas or nothing, since probably some of these ethical considerations push in other directions as well.)
The discussion leads to a broader challenge that bolsters some of the other objections you consider. Allowing market harms to influence our purchasing decision is not demanding, but allowing all relevant ethical considerations to influence our purchasing decisions is very epistemically demanding. Holly Lawford-Smith makes this point nicely [https://onlinelibrary.wiley.com/doi/abs/10.1111/misp.12051] and I’ve discussed it as well [https://doi.org/10.1017/beq.2020.13]. Given this level of epistemic demandingness, Hussain’s idea that the market is not the appropriate place to try to resolve these issues seems to have more weight behind it. Outside the market, activists and political reformers can specialize on one issue, and hopefully make high quality decisions. In the market, we purchase so many different items each with a great range of potential ethical flaws, it could be appropriate to bracket out ethical concerns, including market harms and benefits.
Of course, this epistemic demandingness problem arises for any form of consumer ethics where a purchasing decision is the main locus of action, but I’m interested to hear what you think of it. Is there a good reason consumers should only care about forced labour and market harms when trying to be an ethical consumer? Or are you sanguine about the ability of consumers to be able to weigh a wide range of ethical considerations tolerably well without spending hours of research for each product?
Thanks again for the thought provoking piece!
To save Hayden time, I can give an idea of what Ewan argues in his piece linked above: the complexities of the supply chain suggest that it is overwhelming for individual consumers to determine which or to what extent their purchases are morally objectionable or contribute to morally problematic outcomes. Ewan concludes that conscious consumption is possible, but not by considering individual consumption choices. Instead, he suggests that conscious consumption could be enacted by (for instance) using the time and resources saved and distributing them to effective charities–maybe those that reflect the morally problematic outcomes that the consumption choices might contribute to.
Thanks Ewan for the great comments!
Yes, it certainly does seem like a lot of work for a consumer to figure out the full list of harms and benefits their purchases impose. The market harms and benefits are really just a small part of that. (And even those are likely much more complicated than I’ve described them, if the consumer can go and find more detailed information about the firms operating in banana production, the distribution of wealth among growers, etc, etc.) I agree that, yes, conscientious consumers should be considering the full list of impacts, including the environmental impacts and the status of the company selling the product. It quickly becomes very complicated.
And yes, that does make it awfully demanding for consumers. In an ideal world, I think, we’d have economists and philosophers doing this work for them and publishing easily digestible advice. But in our actual world, I’m not sure what we should demand of consumers. The situation reminds me of cases of ‘cluelessness’, especially as discussed in a recent paper by MacAskill and Mogensen (“The paralysis argument”). I suspect that at least some moral views (e.g., expectational utilitarianism) will be able to deal with the situation: in the absence of information, they can treat all goods as equally likely to impose the same harms and benefits, and so these unknown considerations would all cancel out between options; and they would only be required to seek out new information when it’s relatively low cost to do so. But, under other views, the story will be much more complicated. Perhaps optimistically, I suspect that any reasonable moral view would say that, if you know about some harms/benefits but not about others (and you can’t easily find out), then you should just make your choice based on the harms/benefits you know about. But perhaps not!
On the formula for the amount of money transferred, yes, about half a dozen economists looked over it and its derivation before I submitted it to PPA. It’s accurate, provided that the market we’re looking at isn’t too monopolistic (or monopsonistic). And we should be uncertain about the estimates for elasticity.
Sadly, most economists don’t seem to be interested in figuring out the effects of individual purchasing decisions. I think it might just be the nature of the discipline that they’re much more interested in policy interventions and predicting large-scale trends; less so in providing advice to individuals. One notable exception is Phil Trammell, a colleague of mine, who’s done some recent work on ethical consumerism, prompted by my paper. (I’m having trouble posting links, but one of his drafts–titled “Ethical consumerism”–can be found online.) He finds some interesting long-run effects of purchasing one good rather than another–basically, the more of X you buy, the better producers get at producing X. But I don’t know of any work looking at the broader effects on economic development of buying goods produced by the poor. I can only assume that they’d be positive!
On the elasticities, it’s not clear how accurate they are for actual markets. Most of the figures cited are from more than a decade ago, and generally it’s quite hard to work out long-run elasticities. So we should definitely be very uncertain there. (All the more reason for a risk-averse agent to avoid buying wheat–it could be even more harmful!) Likewise, we should be very uncertain about how much of the cost of our supermarket purchases filters through to growers. For that, I couldn’t even find numbers for most of the goods listed!
And, on whether consumers should only care about forced labour and market harms when making purchases, I think no! The case for caring about it in that setting is especially strong, since they’re plausibly imposing harms themselves. But if agents have any obligation or interest in doing good more broadly, intervening in markets and/or campaigning against forced labour seems like a good way to go! (Very speculatively, I think it might be a good option for impact-driven philanthropists to put their money into wheat/soybean/lentil farms, or for political activists to focus on removing tariffs from banana imports.)