A quick thought on utilitarianism, non-ideal theory, sweatshops, and distributive justice.

So, I’ve been thinking about utilitarianism and non-ideal theory. Although what I’ve come up with may be quite obvious, I’d be interested in reflections on the thought.

It seems to me that there are times when we might do best (even on utilitarian grounds) not to do what would maximize utility in non-ideal circumstances. Consider an instance in which this point may have practical bite. Some argue against ending child labor because the children we prohibit from working may suffer more for our good intentions. Child prostitution may be their second best option. But that this would be so, holding everything else fixed, does not mean we should not try to end child labor. What it shows is that we should try to end child labor and help educate the children we liberate. If one says that we do not have the resources to do this then we should reply that we can and need to find the resources — that is what justice requires. Even for a utilitarian, there are times when we should not do what might initially seem to maximize utility because doing that will only maximize utility conditional on facts that we can and should change. Perhaps there is reason to worry about doing non-ideal theory in some circumstances. Or, more precisely, that we have to be careful about what kind of non-ideal theory we are doing. Consider another example to support the point. Aid organizations spend a great deal of time and money figuring out how to allocate scarce resources. For instance, the WHO tries to prioritize health interventions to maximize the number of disability adjusted life years (or whatever) that it can save with its resources. But if the global distribution of medical resources is unjust and can be changed, the WHO might better spend its time trying to change the global distribution of medical resources.

Be Sociable, Share!
This entry was posted in Posts, Problems and tagged , , , . Bookmark the permalink.

13 Responses to A quick thought on utilitarianism, non-ideal theory, sweatshops, and distributive justice.

  1. I certainly agree that maximizing utility isn’t the end-all of morality, but I do think we ought to evaluate our actions and policies largely by the effects they produce – perhaps weighing the effects on the most vulnerable people more heavily in our deliberations. Suppose that parents in the developing world only feel the need to send their children to work because they live in poverty that they would not live in if we had a just global basic structure. Suppose further that the global basic structure is something that we can change. If we can change it, and doing so would produce the best consequences possible, then I agree that we have strong reason to do so, but I don’t see how this incompatible with the claim you deny in your first sentence. A lot will depend, though, on how the “we” in your “conditional on facts that we can and should change” should be read. “We,” in some sense of the word, can change the global basic structure. But all that I can do is make some relatively small contribution toward that change. Assume that as head of an MNE, any contribution I could make to changing the global basic structure would be negligible. Should I then nevertheless refrain from hiring child labor, on the grounds that in an better world, it wouldn’t be necessary? I don’t think so. What it is moral for me to do, I think, ultimately depends on what it is possible for me to do. If, contrary to my claim, I can make a significant contribution to changing the unjust basic structure, then I probably should. But if I can’t, then I should do the best I can within that structure to do good, and I think sometimes that will mean employing child labor. That last claim, though, is dependent on an empirical belief that we often do more good for the world’s poor through self-interested action in the market than through behavior which is directly intended to benefit them such as aid – a claim along the lines of that defended by William Easterly in his recent response to Bill Gates’ call for “creative capitalism”: http://online.wsj.com/article_print/SB120235183917849631.html.

  2. Hi Nicole,

    I’m very much in agreement with the point you make here, although I think it applies regardless of whether one is a utilitiarian or not. I think you raise an important and more general point about the role of ideal theory. I have the sense (obviously based only on my own random experiences) that there are an increasing number of political theorists who are impatient with the kind of ideal theory practiced by philosophers like Rawls or Jerry Cohen. These theorists tend to believe something like the following: ‘the assumptions of ideal theory are often so far removed from the actual circumstances we find ourselves in that they are more or less useless in terms of practical reasoning. If we want to know what we ought to do here and now, we have to begin by assuming a lot more facts that resemble the here and now. In other words, a lot of non-ideal facts (e.g. greater scarcity, history of injustice, moral apathy) have to be assumed in order to generate sound practical guidelines.’

    I think your examples help illustrate why this sort of impatience with ideal theory is misguided. We can’t be certain that action X is really the best thing to do unless we have some grasp on the kind of world we’re trying to create. This is particularly true in examples, like yours, where the ‘facts’ that might be taken as parameters in non-ideal theory are actually things about the world that we ought to change, things that shouldn’t be taken as given. I thus have a great deal of sympathy for the general form of Jerry Cohen’s ‘Facts and Principles’ argument, though unlike Cohen, I don’t believe his argument presents a fatal problem for constructivists like Rawls. What I agree with is Cohen’s insistence that we musn’t confuse fact-sensitive moral principles (do X because fact Y obtains) with those fundamental normative principles or ideals that must be fact-insensitive since they tell us, among other things, whether we should try and make it the case that fact Y obtains.

    I’ll also add a plug for my friend Zofia Stemplowska, who has a terrific paper called ‘What’s Ideal About Ideal Theory’ forthcoming in Social Theory and Practice, which makes the case for ideal theory in a convincing fashion.

  3. Is the distinction between ideal and non-ideal theory something that utilitarians should care about at all?

    I wonder because it seems to me that consequentialists should simply find the world as it is and try to figure out which of their actions helps make it better. Now, part of this may be to advocate for policies that depend for their efficacy on circumstances being somewhat different from, and perhaps better than, they are now. If those policies were implemented now, in the absence of those circumstances, the results may be disastrous. But the consequentialists are implementing those policies, they’re simply advocating them. Big difference.

  4. Nicole Hassoun says:

    Thanks for sharing your thoughts!

    Let me take each of your points in turn.

    First, Matt – Because I framed the argument in utilitarian terms I did not really intend to call into question the idea that we ought to evaluate our actions and policies largely by the effects they produce. However, I think the point pertains to individuals and holds even if I can make no changes to the global basic structure (or if there is no such thing at all). It seems that even if I cannot change any basic structure there might be other things I ought to do besides hire children. I might use my money to pay for a school for children the organization RugMark rescues from sweatshops, or I might employ the parents of at-risk youth, or I might do something else. Of course, it might turn out that on a utilitarian theory, I really should hire children, but my thought was that even utilitarians may move too quickly to that sort of conclusion because they fail to think about the presuppositions of their (non-ideal theory) arguments. Of course, there might be something wrong with utilitarianism if it does tell us to employ children but that’s a well known worry. My point is merely that there might be a theoretical reason to think twice before accepting the regular anti-sweatshop -utility maximizing arguments.

    An interesting question is why people are much more inclined to buy the pro-child-labor argument than an analogous one about child prostitution or torture. After all, one might argue that one ought to employ child prostitutes (or pay to torture children) because the next best option for these children is child slavery (or something worse). Here few people would accept the premise. Is it self-interest which motivates us to believe the economists who advocate child-labor since our economic fortunes depend upon it? Or is it that most people are further removed from child prostitution (though, if you ever visit Thailand, the argument may quickly gain relevance)?

    Next, Johnathan – thanks for the references – I don’t know that Cohen paper so I’ll have to check it out. I wonder, though, if I’ve just made his point at a more general level. It is not only facts that we must be careful of conditionalizing upon, but principles with limited application.

    Regarding the other point you make, I guess I’m inclined to take a middle-position between the ideal theory/impatience line. I think you rightly point out the virtues of ideal theory. But I think we need to test our theories against practice in order to get them right. Consider, for instance, the fact that lots of philosophers say that we should not make trade offs between meeting the needs of people below some minimal threshold. I don’t think that that principle will work in the real world. If we allow refuge camps in Rwanda some will almost certainly die who would have otherwise lives, but that doesn’t mean we shouldn’t have those camps. (It was only be reading a lot about foreign aid and refugees that I arrived at this conclusion, though I suppose my above reflections give me reason for pause here!). I also worry that (to put it in consequentialist terms) the optimization problem we must solve to figure out what is globally best (in ideal theory) may just be too difficult. But perhaps by testing our conclusions on the ground against what we come up with in ideal theory we can make some progress.

    Finally, Simon – does any of what I’ve said above respond to your worry? I guess I’m not entirely sure why it couldn’t be best (on utilitarian grounds) for utilitarians to start out by doing non-ideal theory. Couldn’t that maximize utility?

    O.K. thanks again! -Nicole

  5. Here is a story. Three adults, Ayn, Max, and Libby, are standing at different points around a shallow pond. They see three children drowning in the pond. Libby realizes that the optimal solution is for each of the adults to pull out one child, spend a few minutes tending to the child, and then walk the child home. Libby wades in and pulls out her child. While tending to her child on the shore, she notices that Max and Ayn are standing there doing nothing. She considers wading in herself to rescue the remaining two children. But Libby realizes that to do so would be “conditionalizing on injustice,” and that what *should* happen is that Ayn and Max each pull out the child closest to them. Since that is what should happen, and given that that happens, the best thing for Libby to do is to walk her child home, Libby tells Ayn and Max to each save a child, then she turns her back on the scene and walks her child home. The other two children drown.

    In this story, Ayn and Max acted terribly. But Libby also acted very badly. Since Ayn and Max were not in fact pulling out their assigned children, Libby should have waded back in and pulled them out.

    The moral of the story: Don’t choose your actions on the basis of the optimal total plan for everyone. Choose your actions on the basis of realistic assumptions about what other people will in fact do. A supplemental, empirical principle: absent specific evidence to the contrary, the most realistic assumption is that people will continue to behave in the way you see them now behaving.

    In the case at hand, here is something that is very *unlikely* to happen: Activists succeed in reducing the extent to which companies outsource jobs, *and* they also convince the first world countries to donate large amounts of money to build effective schools all over the Third World, to pay for schooling for almost all children, and to directly raise the living standard of Third World families so that children no longer have any need to work. As a result, everyone is happy and utopia is achieved.

    Here is something much more likely to happen: Activists succeed in greatly reducing outsourcing, and they fail miserably in convincing first world countries to donate large sums of money to the Third World. This slightly harms the U.S. economy while greatly damaging the economies of the Third World countries, where people are driven to work in worse conditions and are held in poverty unnecessarily for another generation, until these countries finally accumulate enough capital to build their own “sweatshop” factories.

    Lastly, here is another point about good utilitarian reasoning: You do not help people by reducing their options. If someone has bad options A and B, you may help them by creating a third option, C, that is better than A or B. But, whether or not you succeed in offering them C, you almost never help by eliminating A. If C is truly better than A and B, then in most realistic circumstances, they can be informed about C and will then choose C voluntarily. If C is not really better (contrary to your belief), or if you fail to really make C available (contrary to your intention), then the person can continue to choose A. It is much more common that C is not really better or that you fail to really make C available, than that the other person chooses A irrationally when a better option is available.

    In the present case, I would say that if you can raise money to build schools and raise the welfare of families in the Third World so that their children don’t have to work, that’s great. But in case you fail to get the funds, or in case going to school does not really turn out to be better than working, I suggest we leave as many options open to the Third World poor as possible. If and when you succeed in your plan of providing clearly better options, the poor can voluntarily take those options at that time.

  6. Nicole Hassoun says:

    Hi Mike,

    Thanks for your stimulating post! I liked your story very much and agree entirely with its conclusion. Libby behaved poorly. She shouldn’t have chosen her actions based on the optimal plan for everyone. I’d even say we (sometimes) have to act on the basis of realistic assumptions about what other people will in fact do. But, I am not so sure that we’ll agree about what is realistic. And, more importantly, I think that sometimes what we have to do is gather more information about what people will do, or we act badly. Libby could have asked the others if they would save the other children and then listened to their responses. She could have explained what the optimal action would require and tried to have convinced them to help. Though, if the others refused to help, she’d still have to save the other kids.

    I guess I worry a bit about the empirical principle you propose: “Absent specific evidence to the contrary, the most realistic assumption is that people will continue to behave in the way you see them now behaving.” Why should we assume that they won’t change in predictable ways? (If Max were standing frozen by the pond, it might be reasonable to suppose he’d continue to stand there for a while, but it wouldn’t be reasonable to suppose he’d stand there forever.) And, can’t there be times when we should try to influence others’ behavior? We should not always have to act on the basis of realistic assumptions about what other people will in fact do if that means we have to take their behavior as given. Sometimes we can influence their behavior.

    So back to the sweatshop case. I agree that activists are unlikely to “succeed in reducing the extent to which companies outsource jobs… *and* also convince the first world countries to donate large amounts of money to build effective schools all over the Third World, to pay for schooling for almost all children, and to directly raise the living standard of Third World families so that children no longer have any need to work… and utopia is achieved.” At least they probably won’t achieve all of these objectives any time soon :).

    But couldn’t I help at least as many kids attend school than the number of kids I fail to employ by donating to Rugmark (as well as buying Rugmark rugs)? Wouldn’t that be enough? And couldn’t there be a division of labor such that my responsibility is not to employ any kids and to help some kids (not necessarily those I refuse to employ)?

    I should also say that the reason I’m pretty confident in agreeing that activists will probably not succeed (any time soon) in bringing about what you call “utopia” is that it would require so much. I’m much less confident in saying that activists will succeed in greatly reducing outsourcing. Nor am I confident that they will fail to convince first world countries to donate large sums of money to the Third World (consider, for instance, how much money countries have given to try to achieve the Millennium Development Goals). I’m yet less convinced that the net result of activists actions will be to harm the U.S. economy “while greatly damaging the economies of the Third World countries, where people are driven to work in worse conditions and are held in poverty unnecessarily for another generation, until these countries finally accumulate enough capital to build their own “sweatshop” factories.”

    I guess I really do think we need empirical evidence to justify these sorts of claims. I also think that it is usually morally unacceptable to make empirical assumptions to justify things that we know would be wrong ceteris paribus. Let’s suppose you grant that ceteris paribus we should not allow sexual exploitation, dangerous working conditions, etc. and that if we allow “sweatshops” we are allowing these things. It will follow that we should not justify allowing sweatshops on the basis of unjustified empirical assumptions.

    Finally, about your thought that more options are always better for a utilitarian. I’m not sure that that is right. Couldn’t it be disrespectful (and hence lower welfare on a utilitarian theory) to offer some (bad) options? Suppose a potential employer says to you, I know you don’t like the job that I’m offering and need the money so let me make you another offer. How about I pay you to let me cover you in feces? Suppose the employer knows you won’t take the option (you don’t need the money that badly) and will dislike being made the offer. Ceteris paribus, hasn’t the employer acted wrongly (because s/he has only lowered utility by making you the offer)? I don’t know if offering people the option of sweatshop labor is like that, but it might be.

    Alright -thanks Mike! ~Nicole

  7. Thanks for your comments, Nicole. Of course, you can sometimes influence others’ behavior, and when you can, you should (according to utilitarian reasoning) include the optimal way of influencing them in your plans. In the pond example, if Libby could have convinced Ayn and Max to help, then she should have done so.

    But, in the real-world case of first- and third-world countries, I think the preceding observations are irrelevant, because we can be confident that we in fact will not greatly increase the generosity of first-world people. So our situation is more like the situation Libby would be in if she had already repeatedly pleaded with Ayn and Max to help save the children, and they had consistently ignored her pleas.

    You asked: “But couldn’t I help at least as many kids attend school [as] the number of kids I fail to employ by donating to Rugmark …?” Yes, you could do that. But here is something even better you could do: help the same number of kids attend school, *and* also help even more kids by buying from Rugmark. The idea of helping children attend school is a red herring, because it has no relation to whether you should buy from Rugmark. Sending money to aid education is good, whether or not you buy from Rugmark; and buying from Rugmark is good, whether or not you help with education. The two actions are completely independent, as far as I can see.

    My point here does not depend on any very specific predictions. It doesn’t depend, for instance, on exactly how much the U.S. economy will be hurt by reduced outsourcing. Generally speaking, I think policies that damage the economy usually cause the greatest damage (in utility, not in dollars) to the poorest people, because it is the poorest people who most need the economic growth. If Nike closed down all its factories in the Third World, I think I personally would suffer approximately zero loss. Indeed, even if the U.S. economy goes into a recession, I think I’ll be fine. So I am not arguing Rugmark’s case out of self-interest. However, none of that is very relevant.

    What is very relevant is whether the net effect of closing down ‘sweatshops’ would be positive or negative for the Third World poor. I think the arguments for why this would be a negative are well-known, so I don’t feel the need to rehearse them. (You can find the arguments and evidence discussed on sites such as http://www.capmag.com/article.asp?ID=2750.)

    If you are claiming that the net effect would be positive, then I think you need evidence to justify that claim. I don’t even know of any economic theory that would have that prediction. I do not see why your opponents would need empirical evidence but you do not.

    Perhaps your view is something like this: (a) in the absence of specific empirical evidence, we should be neutral with regard to the effects of sweatshops on utility, but (b) we have clear non-utilitarian reasons to prohibit sweatshops, so (c) in the absence of specific empirical evidence, we should prohibit sweatshops.

    If that’s the idea, I think (b) is false. I don’t know what non-utilitarian reasons we would have for prohibiting. However, I think we have a non-utilitarian reason for allowing sweatshops. This is that it is prima facie wrong to interfere with other people’s choices, and especially to interfere with the efforts of the desperately poor to improve their situation. So I would say that we would need some impressive evidence for thinking that prohibiting sweatshops would improve the situation of the poor.

    But if you’re saying that we should prohibit sweatshops even if doing so doesn’t help the poor, or actually harms the poor, then I don’t understand your moral perspective at all.

    Finally, yes there are some unusual possible cases in which you are made worse off by being offered another option. But none of those theoretical possibilities are relevant here, because none of them is realized by the third world people’s option of working for sweatshops. The people who take the jobs are not horribly insulted by the mere suggestion that they might work there; on the contrary, they are happy to accept.

  8. Added comment:

    I may be lost in the dialectic at this point. The beginning part of the dialogue seems to go like this:

    Crazy Leftist: Sweatshops are horrible. We should close them down, boycott Rugmark, etc.

    Sensible Libertarian: But they’re better than the other jobs that those workers have available.

    Nicole: Yes, but there’s a third option that’s even better than both: (1) we close down the sweatshops and (2) we provide schools and enormous development aid so that people have no need to work either in sweatshops or in other, worse jobs.

    Mike: We shouldn’t try to implement the first part of that plan, because
    (a) the second part of the plan is unlikely to happen, and if the second part doesn’t happen, then the first part will be harmful; and
    (b) even if the second part happens, the first part will be at best neutral, not beneficial.

    After that is where I get lost. The Nicole character says … one or more of these things…?:

    N1: Prove to me that doing (1) without (2) would be harmful.
    N2: Prove that (2) is unlikely to happen.
    N3: Well, (2) will probably be partially accomplished.
    N4: Mike’s claim (b) is false, because the availability of the sweatshop jobs will be insulting to the workers.
    N5: There are reasons to close sweatshops that are independent of whether doing so is harmful. (?)

    If one or more of these is right, then the next stage would be:

    Response to N1: I thought that was what you granted at the beginning of the dialogue.

    Response to N2: I’m not going to go down to the library and compile research to prove that, since this is just a blog post. But it seems plausible based on our experience over the past 50 years.

    Response to N3: Unless (2) is fully accomplished, there will still be people who would profit by working in sweatshops. And those are the people most likely to actually fill those jobs. So partial accomplishment of (2) wouldn’t make it wise to eliminate sweatshops.

    Response to N4: I don’t see how it is insulting, and I don’t think the workers do either.

    Response to N5: What reasons?

  9. Nicole Hassoun says:

    Hi Mike,

    I’ll get to this in a day or two. For now, I should just say that Rugmark is “the international nonprofit organization devoted to building schools, programs and opportunities that give children back their childhoods by ending child labor in the handmade carpet industry in South Asia ” (http://www.rugmark.org/about.php?cid=1). They educate the children they rescue from the factories they monitor. I take it that in the examples above you are supposing that they just run sweatshops that use child labor?

    Cheers, -Nicole

  10. Yes, I was confused about Rugmark. I thought it was just a rug company with sweatshops.

  11. Nicole Hassoun says:

    Hi Mike,

    Thanks for continuing the conversation. I’ll try to go through your posts point by point.

    First, you say “in the real-world case of first- and third-world countries… we can be confident that we in fact will not greatly increase the generosity of first-world people. So our situation is more like the situation Libby would be in if she had already repeatedly pleaded with Ayn and Max to help save the children, and they had consistently ignored her pleas.”

    Sticking with the analogy, I’d say that even if Libby has already pleaded with Ayn and Max, she should try to come up with another option besides abandoning the rescued kid if doing so is not optimal, maybe she could call out for help. In the real world case, perhaps we’d do best to donate to an educational organization rather than purchasing a rug (for instance) made with child labor. Perhaps we could educate 100 kids who would otherwise become prostitutes for the amount that it would take to employ a single child.

    To this kind of reply you said we’d do better to educate the same number and employ the child. I’d say, well, if you can get the money to do that you might do better to educate another 100 kids. We have a lot of other options in the real world besides spending our money on child labor made goods or donating to educational institutions.

    My point might best be put in virtue theoretic terms: Virtue requires us to think of these other options and to challenge unjust background presuppositions.

    I agree that none of this depends on specific empirical predictions though I disagree with some of your generalizations and I’m not sure I see their relevance (e.g. if our economy were hurt because we liberalized trade in agricultural goods economists predict that most of the poor would benefit, some would not. I argue in a forthcoming Public Affairs Quarterly article that if we care about the poor we have reason to try to compensate poor losers even from liberalization).

    I also agree that “What is very relevant is whether the net effect of closing down ‘sweatshops’ would be positive or negative for the Third World poor.” But I don’t think the arguments you cited here: http://www.capmag.com/article.asp?ID=2750 are convincing. What I think they show is that we have to do more for those in the factories we close down to help the people and that we should open factories that pay living wages etc. (The article just asserts, but does not argue for the claim that it would no longer be profitable to operate factories in the third world).

    I also agree that if I claimed that the net effect of closing sweatshops would be positive, I’d need evidence to justify that claim. My point is only that we should try to make it the case that closing down a sweatshop would be positive, or upgrade the factories and pay decent wages if we can. And we shouldn’t just assume that we can’t.

    For some information on the kinds of sweatshop conditions at issue here consider the following story: “According to The Observer, children who had been sold into bonded labor were producing Gap Kids clothing under conditions close to slavery in a sweatshop in New Delhi, India. Ranging in age from 8- to 15-years-old, the child workers toiled 16 hours a day, hand-sewing clothing for no pay at all. The sweatshop was smeared in filth and excrement, and the children were subject to threats and beatings” (From: http://www.sweatfree.org/hallofshame)).

    You say: “Perhaps your view is something like this: (a) in the absence of specific empirical evidence, we should be neutral with regard to the effects of sweatshops on utility, but (b) we have clear non-utilitarian reasons to prohibit sweatshops, so (c) in the absence of specific empirical evidence, we should prohibit sweatshops.”

    I’d say: (a*) in the absence of specific empirical evidence, (b*) we have clear *utilitarian* reasons to prohibit sweatshops, so (c*) in the absence of specific empirical evidence, we should prohibit sweatshops.

    I assume you’d still think (b*) is false. Though I agree “That it is prima facie wrong to interfere with other people’s choices, and especially to interfere with the efforts of the desperately poor to improve their situation.” I think we know that offering some choices is wrong (you agree in the case I provided about the feces which I think is relevantly similar to what we’re offering in the sweatshop case though what we offer there is often worse), and I’m not sure I interfere with someone’s choice when I don’t provide them with an option because I don’t offer them such a choice.

    So I hope it is clear that I’m not “saying that we should prohibit sweatshops even if doing so doesn’t help the poor, or actually harms the poor.”

    Finally, let me say a few things about your attempt to clarify the dialectic. I think we did start out with something like this:

    Crazy Leftist: Sweatshops are horrible. We should close them down.

    Sensible Libertarian: But they’re better than the other jobs that those workers have available.

    Nicole: Yes, but there’s a third option that’s even better than both: (1) we close down the sweatshops and (2) we provide schools and enormous development aid so that people have no need to work either in sweatshops or in other, worse jobs.

    Mike: We shouldn’t try to implement the first part of that plan, because
    (a) the second part of the plan is unlikely to happen, and if the second part doesn’t happen, then the first part will be harmful; and
    (b) even if the second part happens, the first part will be at best neutral, not beneficial.

    I think some confusion may have come from our switching between individual and group actions. Let’s look at the individual example:

    The proposed strategy is:

    (1) I don’t purchase from sweatshops and (2) give aid so that I do more good for people who would otherwise have to work in sweatshops or in other, worse jobs.

    I take it you’d no longer say: We shouldn’t try to implement the first part of that plan, because
    (a) the second part of the plan is unlikely to happen, and if the second part doesn’t happen, then the first part will be harmful; and
    (b) even if the second part happens, the first part will be at best neutral, not beneficial.

    I think you’d say that the second part is likely to happen and will be beneficial. Still you could say:

    I shouldn’t purchase sweatshop made clothing because
    (d) Not enough people are likely to donate to programs that educate children so that kids won’t need jobs, and (e) if kids need jobs, then not buying sweatshop made clothing will be harmful; and
    (f) even if enough people do donate, not purchasing sweatshop made clothing will be at best neutral, not beneficial.

    Then, although I might go further than this, I’d first want to say:

    N6: Prove to me that (e)
    and
    N7: Both (e) and (f) are probably false

    Furthermore, I’d want to say that you have a moral reason to prove to me that (e). This is because

    N8: You are making a rather big empirical assumption, which, if false, will have led you to act immorally.

    That is because

    N9: Offering a child the option of working in sweatshop conditions is, other things being equal, morally unacceptable. If your assumption is wrong, it is worse than offensive, worse than insulting, it lowers utility.

    Finally, I think:

    N10: Offering the child the option of sweatshop labor would, if you are wrong about the empirics, lowers utility relative to other feasible options I have (i.e. giving all my money to charity) and it will lower utility because it is a bad offer.

    I feel like I’ve gone on too long so I won’t consider the group action case unless you’d like me too.

    Thanks again Mike, -Nicole

    p.s. One more thought: the general principle underlying N8 is probably something like this:

    N11: You should try not to assume an answer to an empirical question that will lead you to do something that would be impermissible if your assumption proves false.

  12. Nicole,

    I’m not sure I understand what’s going on.

    (a) Sometimes it sounds like you’re saying, “it’s bad to rely on uncertain empirical assumptions,” but this doesn’t make sense to me, because any utilitarian reasoning relies on uncertain empirical assumptions. The claim that buying from sweatshops lowers utility is just as empirical as the claim that buying from sweatshops doesn’t lower utility. If P is empirical, then ~P is empirical. So I don’t understand how you believe you’re not relying on empirical assumptions.

    (b) Then you say you’re not making an empirical assumption because you’re only saying that we should “try to make it the case” that buying from sweatshops lowers utility (or that some other option raises utility). It seems to me that we should only try to do x if we have epistemic reason to believe that trying to do x would raise utility. So I don’t see how you hope to avoid empirical claims by just saying we should try to do something.

    But also, it seems like you’re changing the subject from what I thought we were discussing. I didn’t intend to state any claims about whether we should try to make sweatshop-buying lower utility, or whether we should or shouldn’t make certain assumptions. What I meant to say was that we should buy from sweatshops. Do you disagree with that?

    I see three possibilities: (i) You claim that we shouldn’t buy from sweatshops, for some utilitarian reason. In that case, I don’t see how you can say that that’s not empirical. (You might also endorse some other, non-empirical or less empirical claim, but the claim that you shouldn’t buy from sweatshops for utilitarian reasons would still be empirical.) (ii) You claim that we shouldn’t buy from sweatshops, for some non-utilitarian reason. But it seems like you disavowed that in the last post. (iii) You aren’t saying anything about whether we should buy from sweatshops, so we’re not disagreeing. But this would be weird.

    (c) Up til about the end of your post, I couldn’t see why you would think buying from sweatshops might lower utility. But at the end, it sounds like the explanation is this: buying from sweatshops lowers utility relative to the option of giving away all your money to charity. I think that’s true. I never meant to deny that. But then I don’t see that this as really a discussion about sweatshops. “You shouldn’t buy shoes from Nike, because you shouldn’t buy anything (beyond the bare necessities of life), because you need to send the money to Oxfam instead” seems like an entirely different argument from what I thought we were talking about. That argument might be correct. But I wouldn’t label it as an argument about sweatshops, since it applies to all products produced by anyone.

    The same goes for the remarks about Rugmark. I thought you were deciding between buying a Rugmark rug and buying an American-made rug–not that you were considering the option of going without a rug so you could send the money to Oxfam.

    (d) Lastly, about what virtue requires, I think virtue often requires us to try hard to think of better alternatives, when faced with a set of bad alternatives. But also, virtue sometimes requires us to accept unpleasant realities, rather than engaging in wishful thinking.

    Take the case of the battlefield surgeon, who has several patients in need of attention, but not enough time to attend to all of them. He must decide whom to help, and whom to let die. Suppose the surgeon decides that letting anyone die is an “unacceptable” option, so he refuses to accept that that’s the situation he’s in. He just sits there trying to think of other options. Or he divides his attention among all the patients, so that none of them receive enough care. This would not be an exercise in virtue.

    It’s not clear to me whether the reluctance to accept the buying of sweatshop products is a virtuous seeking of better options, or a vicious indulgence in wishful thinking.

  13. Nicole Hassoun says:

    Hi Mike,

    The first thing I should say is that am not, or at least I did not start out being, particularly concerned in this post with whether sweatshops are or are not justifiable (so, yes, iii is right). Nor did I think the idea I wanted to explore only applied to utilitarians (I wanted to use utilitarianism and sweatshops to illustrate the idea).

    But, since we were having a discussion about sweatshops when I mentioned this post to you originally, I think you were justifiably confused. I’ve probably gotten a bit confused myself at some points because I do happen to think there is something morally wrong with employing children. I also wanted to use the idea I was exploring to get you to agree with me that you should not support sweatshops. But to do that, I think you’d have to share my assumption that sweatshops are impermissible if the empirical case you suggest exists for them cannot be made. Do you?

    In any case, let me just try to put the idea I was originally trying to explore out there once more (in a slightly different way). I was interested in what kinds of non-ideal theories we should do and what kinds of assumptions it is permissible to make.

    I’d like to be able to say something like: (e.g. empirical) assumptions cannot be used to justify an action or policy when one grants that that action or policy would be impermissible if the (e.g. empirical) assumption does not hold.

    Now the illustration: If one grants that sweatshops would be impermissible (for utilitarian or other reasons) if there were other options, then it would be impermissible to assume that there are no other options.

    One could grant the antecedent for many reasons: One might think that we could educate the kids instead of employing them or that we could give all of our money to charity instead or that there is just something offensive about offering kids the opportunity to work in the horrifying conditions often associated with sweatshops.

    But, it doesn’t matter for my purposes why one agrees with the antecedent of the principle I wanted to explore, and I could probably come up with other examples that don’t even involve sweatshops.

    Of course, you do challenge my proposed principle and not just its antecedent with your battlefield case. If I wanted to agree with you about what the Dr. should do, I wonder if I could say that the difference between the sweatshop case and the battlefield case is that the doctor has lots of evidence that s/he faces a tragic dilemma. But I should probably just say the doctor shouldn’t assume s/he faces a tragic dilemma. Perhaps s/he could send someone to get help at the nearest hospital while treating some of the worst-off patients thereby saving them all.

    When one really is facing a tragic dilemma and it is optimal to stop searching for other options then one should stop searching for other options.

    But, in virtue theoretic terms, my idea was just that virtue requires one to search for as long as one can.

    Maybe that’s not interesting, but it might be. It might be interesting if we generally don’t search long enough. If, for instance, in sweatshop-like cases people just accept the claim that we should buy child labor made goods because if we don’t buy child labor made goods then the children will become prostitutes. Perhaps what we should do is close the sweatshops and educate the children too. Perhaps people should generally think harder about the alternatives. Sometimes I wonder if philosophers get so good at focusing on dilemmas (at conditionalizing on hard facts) that they stop questioning the assumptions. Anyway, I hope that clarifies things!

    Thanks again for the great discussion Mike, -N

Leave a Reply