Judith Lichtenberg A central question moral and political philosophers have asked in recent decades is whether well-off people have moral duties to aid those suffering severe deprivations of basic necessities, and if so how extensive these duties are. No one disputes that people have duties not to harm others; these so-called “negative” duties are about as well-established as any moral duties could be. But the very existence of “positive” duties to render aid is controversial, and even among those who concede their existence the nature and extent of such duties is disputed. A critical concern about them—perhaps the critical concern—is that once we admit duties to aid into the moral realm they threaten to take over and invade our lives: it is hard to draw a line that will prevent them from becoming relentlessly demanding. When we think of all the people in the world who suffer very serious deprivations of basic needs, and how much the reasonably affluent could do to help them, the slippery slope looms before us. Peter Singer made it clear in “Famine, Affluence, and Morality,” if it had not been clear before (and to most people I do not think it had), arguing for what seemed to many like inhumanly demanding duties of the rich to aid the poor. On the strong principle Singer defended, affluent people should give until, if they gave more, they would endure as much suffering as they would relieve for others. Even on the weaker of the two principles he defended there, “we would have to give away enough to ensure that the consumer society, dependent as it is on people spending on trivia rather than giving to famine relief, would slow down and perhaps disappear entirely.”1 But Singer was not alone, and his essay would not have produced the effect it did had it not tapped into deep concerns—on the one hand, about the extent of our responsibilities to relieve poverty and suffering; on the other, about the intrusive consequences of admitting such responsibilities for our ability to live our daily lives as we see fit.
Both common sense (of whose teachings we have, of course, reason to be deeply suspicious) as well as certain prevalent philosophical approaches tell us that we cannot be required by a reasonable morality to make very large sacrifices to our own well-being—that such requirements impermissibly infringe our autonomy and our ability to live our lives as our lives. These critiques have commonly formed part of an attack on utilitarianism and consequentialism, but they exert influence independently of these moral theories. Now one thing that gives this position—what Liam Murphy has called “the over-demandingness objection” to duties of aid or beneficence2—its persuasive power is the implicit contrast with our “negative” duties not to harm people. Although we have at most limited duties to aid people, the argument goes, we have rather strict duties not to harm them. And one thing that gives this position its persuasive power is the suggestion that not harming people is for the most part straightforward and easy. Don’t kill people, don’t rape them, don’t attack them, don’t rob them: if you follow these simple and indisputable rules you are doing what you ought to do and cannot be faulted; at least you have fulfilled your obligations.
Yet sometime in the last decade or two or three, something changed. We look at our daily lives and find—or, in many cases, others inform us in no uncertain terms—that our most humdrum activities may harm people in myriad ways we have never even thought about before. And because these activities are seamlessly woven into our normal routines, ceasing to engage in them is not at all easy—not simply a matter of refraining from things we never would have dreamed of doing in the first place, like killing and raping and robbing. Not harming people turns out to be difficult, and to require our undivided attention.
The contrast in burdensomeness between not harming people and helping them may once have been sharp, but no longer is. Although on some views minding your own business is all a person is morally required to do, even this apparently modest injunction turns out now to be almost impossible. Over the last few decades—but especially in the last few years, with economic, environmental, and electronic globalization rapidly increasing; near-consensus about the threat of severe climate change, whose effects will be felt most harshly by the world’s poorest people; knowledge that products we use every day are produced in sweatshops; and, last but not least, the growing impossibility of avoiding awareness of these phenomena—we have learned how our everyday habits and conduct contribute to harming other people near and far, now and in the future. The model of harm underlying the classic formulation of the harm principle—discrete, individual actions with observable and measurable consequences for particular individuals—and the attendant the focus on negative duties no longer suffices to explain the ways our behavior impermissibly impinges on the interests of other people.
Turn off the lights. Use compact fluorescent bulbs (even if they produce an ugly light). Drive a small, fuel-efficient car. Drive less. Take public transportation. Don’t fly unless you really need to (no more trips to international conferences, no more exotic vacations). Turn down the thermostat in the winter. Turn off the air-conditioning in the summer (too bad if you live in the humid south, east, or midwest). Make sure your appliances are energy-efficient. Take cooler showers. Eat local (except sometimes; find out when3). Don’t eat factory-farmed meat; producing it is not energy-efficient (leaving aside harms to animals). Don’t buy Chilean sea bass, or salmon, or… (fill in the blank, depending on which sea food is overfished at any given time). Don’t drink bottled water—the energy costs of producing and transporting it are wasteful (leaving aside that only 14 percent of bottles are recycled). Don’t use plastic bags (or paper bags either!).4 Recycle. Compost. Don’t use chemical fertilizers on your lawn; better still, get rid of your lawn.5 In this new world in which we find ourselves, “each bite we eat, each item we discard, each e-mail message we send, and each purchase we make entails a conversion of fossil-fuel carbon to carbon dioxide,” with possible deleterious consequences for others and for the globe.6
Apart from the environmental consequences of our actions, which disproportionately affect poor people, other kinds of harms also loom. Don’t buy clothing made in sweatshops. (Find out which those are.) Was your oriental rug knotted by eight-year-olds? (Find out.) Do you own stock in a company that exploits its workers? (Find out.) Is the coltan in your cell phone fueling wars in the Congo? Leif Wenar explains how western consumers may “buy stolen goods when they buy gasoline and magazines, clothing and cosmetics, cell phones and laptops, perfume and jewelry.”7 These harms result from flaws in the international system of global commerce, which allows corrupt dictators in resource-rich countries to profit hugely at the expense of their impoverished citizens.
Every bite we eat! Every purchase we make! To not do these things, to know what not to do, to know what to do instead, can encroach on our autonomy at least as oppressively any duties of aid or beneficence. Thomas Pogge conceives of “human rights narrowly as imposing only negative duties,” in order to keep his argument for human rights “widely acceptable.”8 This rejection of positive rights as elements of human rights may seem surprising; one might believe that a progressive, humanistic philosophy of human rights would embrace the responsibility to protect the vulnerable even when it is not our fault that they are vulnerable. Leaving that issue aside, however, the question is whether or to what extent the “wide acceptability” of negative duties depends on the assumption that they do not make significant demands on us—and whether they will remain widely acceptable once we realize just how challenging negative duties can be.
Causality and psychology
Central to the classical picture of harm on which the primacy of negative duties depends is “the idea that individuals are primarily responsible for the harm which their actions are sufficient to produce without the intervention of others or of extraordinary natural events…”9 Two elements are important. One is that an individual’s action is sufficient, without the acts or interventions of other people, to cause harm.10 The other is that the harmful effects a person’s action produces are generally near and immediate. My fist comes into contact with your nose (and breaks it); my car runs you over (and crushes your leg).
This causal picture less accurately reflects the mode of individual agency increasingly prevalent in the world today than it does to classic torts, for example. In the cases I focus on here—what we might call the New Harms—no individual’s action is the cause of harm; it would be more accurate to say that individual actions make causal contributions to an overall effect that may be very large and significant. Samuel Scheffler has described concomitant changes in what he calls the phenomenology of agency in these New Harms. Individuals may not be aware of the contribution their act makes, they have little or no control over the larger processes, it is difficult to get information about these processes, and equally difficult to avoid participating in them.11 Psychologically or phenomenologically, “[t]he primacy of near effects over remote effects means that we tend to experience our causal influence as inversely related to spatial and temporal distance.”12Our immediate influence on our surroundings seems real to us in a way that our remoter effects do not. And since with the New Harms there are no palpable, immediate, visible effects of our actions, we are likely to feel no regret, no guilt, no shame, and no drive to act differently.
These psychological states—or perhaps we should say the absence of them—resemble our mindsets when we do not aid those whom we could aid. As Scheffler puts it, “we experience our omissions as omissions only in special contexts.”13 If I fail to jump into the pond to save the drowning child before me, or if I do not intervene when I see a mugging on the subway, I am likely to “experience” the omission. But I will not ordinarily experience my failure to aid starving children a half a world away as an omission, much less as a failure. Ordinarily, I will have no experience.
Lacking the relevant psychological states, people do not “feel” they are doing anything wrong when they contribute to the production of New Harms; in the absence of this self-recognition, changing behavior is correspondingly more difficult. I return to these points toward the end of the paper.
In the contemporary world, not harming people may be every bit as arduous and demanding as aiding them, perhaps more so. Since a central objection to acknowledging duties to aid is their demandingness, the question is what the demandingness of negative duties implies about their stringency, or about their stringency relative to positive duties. I shall argue that the comparable demandingness of positive and negative duties to remediate the New Harms undermines the claim that negative duties are far more stringent than positive duties. But I will also argue that moralists have good practical reasons to drop the forbidding language of ‘duties’ altogether.
First, however, we need to consider some apparent asymmetries between harming and not aiding (or between not harming and aiding). Any such asymmetries could show that the analogy does not hold or is less apt than it appears, and could affect the conclusions we should draw about our duties not to harm or to aid. I will argue that despite some differences the analogy holds.
I see two possible kinds of asymmetries between harming and not helping—or, similarly, between negative and positive duties—that pull in opposite directions. One asymmetry derives from the commonsense belief that duties to avoid harm take priority over duties to aid. Call this view the Moral Priority of Avoiding Harm Over Helping. The other asymmetry suggests that, in the kinds of global cases we are considering, people have more reason to aid others than they do to avoid harming them, because, acting as individuals, they can in fact be more effective in rendering aid than they can in avoiding harm. Call this view the Greater Effectiveness of Helping Over Not Harming. I consider each of these possible asymmetries in turn.
First possible asymmetry: The Moral Priority of Avoiding Harm Over Helping
The first asymmetry derives from the widespread belief that duties not to harm are more stringent than duties to aid. One basis for this belief, I think, is just the point under discussion here: the idea that duties to aid, unlike duties not to harm, are unacceptably demanding. But this is not the only reason for it. One who harms another makes that person worse off than she would have been had the agent not done what he did; one who fails to aid does not make someone worse off in this way. In light of this difference, some assert a kind of existential claim: you are liable for making the world worse than it would have been had you not acted—more dramatically, had you never been born—in a way you are not liable for failing to making the world better. Something like this point seems to underlies the view that, as Scheffler puts it, “individuals have a special responsibility for what they themselves do, as opposed to what they merely fail to prevent.”14 This outlook comes in various strengths: in the strongest version, you have no responsibility for not making the world better; in weaker versions, you are responsible, but not to the same degree as if you had made someone worse off. A view of this kind is central to Robert Nozick’s claim that the state may prohibit people from harming others, but may not require them to aid others.15
There is much room for disagreement about this fundamental existential claim, especially in its strongest forms: that people are not morally required to aid others, or that the state may never force them to do so. Yet one version of the asymmetry claim seems difficult to deny. In at least one sense harming is worse than not aiding, because having harmed a person always provides a reason to rectify her plight over and above any other reasons one has. Thus, imagine the proverbial drowning child in the pond. Most people believe the bystander ought to wade in to save the child. Yet few would deny that the reason to intervene intensifies if the bystander is no mere bystander but (accidentally or intentionally) has pushed the child into the water.
In this sense, then, it seems incontrovertible that harming is worse than not aiding (or, in other words, that negative duties are stronger than positive): however strong the reasons to alleviate a person’s suffering, a person has an additional reason to do so if she has had some role in bringing that suffering about. Thus, other things being equal, duties not to harm are more demanding than duties to aid—not in the sense that they are necessarily more onerous, but in the sense that they are more obligatory.
It should be noted that nothing in this argument tells us how much stronger negative duties are than positive duties. True, having reasons x and y for acting provides more push than having reason x alone. But if x is itself a very strong reason, then y may not add much additional force to it. The proponent of the Moral Priority of Harming Over Not Helping simply begs the question if she assumes that reasons for helping (x) are weak relative to reasons for not harming (y). One reason for that assumption is precisely the intuition that x is much more demanding than y; if it turns out that they are similarly demanding, the justification for that assumption weakens.
In any case, this asymmetry supports my overall argument rather than undermining it. To the extent that we accept the Moral Priority of Not Harming Over Helping, we have reason to take our negative duties in the New Harms very seriously.
Second possible asymmetry: The Greater Effectiveness of Helping Over Not Harming But another factor may pull in the other direction. In the cases before us, it might be argued, the effects of an individual alone refraining from the acts in question may be negligible or nil, while on the other hand it may seem that she could through her aid single-handedly make a significant difference to someone’s well-being.
So the following argument might be made. “You say duties to avoid harm are just as demanding (or more so) in the world today as duties to aid. But there’s an important difference. For a given unit of effort or money, a person can be more certain that her aid (say $100 sent to Oxfam) will help someone than she can be sure that the equivalent amount (for example, $100 saved in carbon emissions) will avoid harm. So, other things being equal, she has more reason to give aid than to avoid engaging in harmful practices.” If this argument is sound, then duties not to harm might not be as demanding as they seem. For, it might be said, people do not have duties to refrain from engaging in harmful practices unless doing so would make a difference to the outcome. And acting alone, individuals do not have good reason to think they can make a difference.
To appreciate the force of this objection, we must explore two possible lines of argument. (1) The first, taken up in the remainder of this section, explores the causal claim that individual attempts to aid in such cases are more efficacious than individual attempts to avoid harm. (2) The second, explored in the next section, asks whether making a difference to the outcome is the only relevant consideration. It examines the rejoinder that it is wrong to be complicit in harm by participating in harmful activities irrespective of whether one’s own conduct makes a difference.
Do I—living in a safe American suburb far from the frontlines of global poverty—have more reason to give $100 to Oxfam or Doctors Without Borders than I do to cut my carbon emissions by $100 or to refrain from buying $100 worth of products made in sweatshops—on the grounds that the former acts are more likely to make a difference than the latter?16
To answer this question, we need facts both about the efficacy of aid and the efficacy of refraining from harm, neither of which are easy to come by. Begin with the question of the efficacy of my aid dollars. In recent years, critics—including many former insiders in the world of international aid—have challenged the effectiveness of humanitarian assistance, in the form both of disaster relief and long-term efforts to improve well-being among the world’s poorest people. The titles of their books speak for themselves: The Road to Hell, Lords of Poverty, Famine Crimes, Condemned to Repeat?, Aid as Obstacle,The Dark Sides of Virtue, The White Man’s Burden.17 As Garrett Cullity explains, the central charge is that large-scale aid programs “damage the local economy and pauperize the ‘target population’. …The effect is to create aid-dependent economies in which the task of developing economic self-sufficiency has been made much harder than it was before.”18 Aid programs can disrupt traditional institutions, undermine incentives to work, erode recipients’ self-respect, and encourage corruption by local governments. According to some critics these problems plague not only institutions like the World Bank and the International Monetary Fund, but also NGOs like Oxfam and Save the Children. Michael Maren, one of aid’s harshest critics, claims that such organizations prey on the pity of donors and spend an inordinate amount of time and money jockeying with competing organizations for position and media attention, while their young and privileged employees live high on the hog amid the poverty they are supposedly there to fix.19
Despite these allegations, which undermine the Greater Effectiveness of Helping Over Not Harming principle, throwing up our hands and concluding that we can do nothing to improve conditions of poverty, disease, and ignorance is not justified. Most of the authors of the above-cited books have proposals for alleviating world poverty, even while they are harshly critical of many existing approaches. Other dedicated groups and individuals—Nobel-Prize winning microfinance guru Mohammed Yunus and New York Times columnist Nicholas Kristof are among those who have received much attention in the media—have made important strides in improving our understanding and disseminating information about the most effective approaches. Contributions that support vaccination programs protecting people against dread diseases, for example, can hardly fail to do good. MORE EXAMPLES? Still, if we hope to do more than assuage our consciences with mere gestures, we will have to work to figure out how to invest our resources, and we must realize that as single individuals our contributions, even if generous, will be very limited.
What about refraining from contributing to harm? How efficacious is it likely to be? Here it may seem that the probability of making a difference is even smaller than in the case of individual aid. If I am careful in how I allocate my donations it seems probable that I can actually help a small number of people, but it is hard to imagine that my solitary refusal to use plastic grocery bags will make any contribution to slowing climate change or that my refraining from buying sweatshop-made clothes will alleviate worker exploitation even a little bit.
But is this right? We have reason to distrust the intuition that our behavior, because it constitutes only a tiny fraction of the whole effect, makes no difference. In the first place, since aggregate effects are a function of individual actions (carried out within a framework of institutions and policies), it would seem that tiny individual changes will have at least tiny effects on the outcome.20 Moreover, people effect change in other ways than by direct reductions in harm. Driving a fuel-efficient car or carrying reusable bags to the supermarket are publicly visible acts; through them a person can set an example or fuel a trend that others may imitate, whether out of conviction or conformity. The power of fashion, the desire for approval and avoidance of shame, pride in living up to one’s principles, the effects of “tipping points”—via such psychological and social phenomena actions with negligible direct effects can nevertheless produce widespread changes in behavior over the longer run. And the belief that one’s own conduct makes a difference is a potent and probably adaptive human trait.
This discussion is inconclusive, because the facts needed to resolve it are hard to come by. It is possible that, per unit of human effort (measured in dollars, or some other way) expended, we are more likely to make a difference by giving aid than we are by refraining from contributing to harm. But we cannot be sure.
It is difficult to entirely wall off the efficacy question, especially when we include indirect effects of our behavior such as setting an example or fueling a trend. Yet over and above any direct or indirect material effects a person’s actions may have, there are other reasons to refrain from participating in harmful activities.
What difference does making a difference make?
In reflecting on such reasons it appears that the harmful activities we are mainly concerned with divide into two categories. In one are the kinds of environmental harms epitomized by the threat of climate change. Such harms are essentially aggregative: there is nothing intrinsically harmful to the environment or other people in burning fossil fuels; the harms depend on the joint effects of many people’s actions. By contrast, other kinds of harms—buying products whose manufacture exploits workers or whose sale deprives owners of their rightful property—are not essentially aggregative; they involve actions that, we might say, are mala in se. Although the commercial practices under scrutiny would not exist without the participation of large numbers of people, each individual act of theft or exploitation exhibits a wrong-making quality.
The distinction between essentially aggregative and malum in se harms affects the moral reasons for refraining from participation in harmful activities. Consider a person who refuses to eat meat because he believes killing animals is morally wrong or at least that the conditions under which most animals raised for food are unconscionable.21 I think most people would agree that, assuming the truth of these premises, the vegetarian’s decision to abstain from eating meat is appropriate, even if his eating or not eating meat does not significantly affect the lives and well-being of any animals, and irrespective of whether other people eat meat or not. Although I have no control over what other people do, the agent may say, I can at least control what I do; and I choose not to contribute to these wrongs. This reasoning might be applied both to malum in se and aggregative harms, but intuitively at least it seems to hold more sway in the former case, where what we do is intrinsically wrong-making.
In the case of aggregative harms the reason to refrain from complicit acts is different. An agent employing the Categorical Imperative will perform only those actions she would be willing to allow everyone to perform. Since allowing everyone to consume energy at the rate consumed by the average American leads to disaster, she concludes that it is wrong or unfair for her to consume at that rate. The vegetarian’s decision not to eat meat, by contrast, is not dependent in this way on fairness (unless it is fairness to the animals themselves).
Recall that the point of this discussion was to consider a possible asymmetry between not harming on the one hand and aiding on the other that would undermine the argument for strong negative duties. The objection can be put in the form of an argument:
1. In the kinds of cases we are considering, an individual acting alone can be more certain that her aid will be effective than that her refraining from harm will be effective.
2. Therefore, per unit of effort (measured in dollars or some other way), an individual has more reason to give aid than to refrain from harming.
3. Therefore, considered as a function of the over-demandingness objection, duties to refrain from harm are weaker than duties to give aid.
4. And, in absolute terms, since refraining from harm is not effective in these cases, the duty to avoid harm is weak or nonexistent.
My first response to this argument was to question the empirical claim in the first step. The second response is to ask whether causal efficacy in bringing about a desired result is the sole criterion by which to judge whether one ought to refrain from participating in harmful activities. I suggested another moral reason one might give for avoiding harm: the importance of doing the right thing, irrespective of whether it is effective. In the case of aggregative harms, doing the right thing involves an appeal to the unfairness of acting inconsistently with how one thinks others ought to act.
We possess rich linguistic resources to describe what is objectionable about such conduct apart from its direct effects. We can talk about the expressive or symbolic meaning of a person’s conduct, about personal integrity, or about “participating in” or being “complicit” in harmful activities. Yet these approaches—all of which bypass questions about the efficacy of a person’s actions—raise other questions about the distinction between duties (or reasons) not to harm and duties (or reasons) to aid.