The Afghanistan Withdrawal and Agent-Relative Duties
by Michael C. Dorf
A new analysis indicates that over a quarter-million Afghans who worked with or for the United States during the last two decades remain in the country. Most of them will not be evacuated in the five days between now and the self-imposed and Taliban/ISIS-K-enforced deadline for withdrawal of all U.S. forces. We will have failed those people, with catastrophic, often fatal results. More than that, we will have wronged them by violating our agent-relative moral duties towards them.
In moral philosophy, an agent-relative duty is exactly what it sounds like: a moral duty that you owe to particular individuals because of something about your relationship with them. Such duties can be usefully contrasted with agent-neutral duties, which we owe everyone.
For example, your duty not to intentionally kill people (absent justification or excuse) is agent-neutral; you are obligated to refrain from murdering everyone. By contrast, your duty to provide food and shelter for your minor children is agent-relative. You must feed and house your own children but you have fewer duties to others. Most moral theories recognize that while it would be praiseworthy for you to provide food and shelter to strangers (whether minors or adults), doing so is supererogatory; it goes beyond the call of duty.
In the last couple of weeks, we have witnessed compelling testimonials from U.S. personnel who served in Afghanistan alongside Afghans who acted as translators or otherwise assisted the U.S. mission and are now in grave peril of reprisal from the Taliban. What makes these cases so forceful, I will suggest, is that they combine multiple strong grounds for finding agent-relative duties.
Familial Obligation. Above I gave the example of obligations owed to minor dependent children as paradigmatic of agent-relative duties. More broadly, although the law in the U.S. allows parents to stop supporting their offspring when the latter reach the age of majority, I believe that most people think that family ties provide a ground for moral obligation. In other words, in many circumstances, we will feel morally obligated to our close relatives, whatever their age. Helping them is not merely supererogatory. Many people feel the same way about community members, believing that they owe more to members of their own church than to others, or more to the unhoused or otherwise in need who live in their community than they do to the unfortunate throughout the world, or more to those who live in their own country than to those who live in foreign countries. Even countries with much more generous foreign-aid policies than the U.S. act on this moral intuition: they provide foreign aid but they provide a much more robust social safety net for people living among them. I don't want to take a position on whether family, community, or national connections really do create agent-relative duties; for now it suffices to observe the widespread belief that we have agent-relative duties to family members and others with whom we have close ties.
Notably, many of the rightly anguished pleas that the U.S. government do more for Afghan translators and others similarly situated liken those Afghans to relatives. As one outspoken Marine veteran put it, "I’m especially bitter that we’ve forsaken the bonds of camaraderie and abandoned our allies." The sentiment is hardly surprising. After all, we commonly refer to those who share such bonds as "brothers in arms." And just as we would think there is special reason to rescue an actual sibling in danger, we may think that there is a special reason to rescue one who occupies the figurative moral space of a sibling.
Promises. Promises create moral obligations. I have no freestanding obligation to lend my car to Joe, but if I promise to do so and then fail to do so, I have wronged him. To be sure, the law does not enforce all promises. There must be some sort of exchange (consideration) or detrimental reliance by the promisee to render a promise legally enforceable. Meanwhile, there is a debate among scholars about the extent, if any, to which contract law does or should track the morality of promising. But wherever one lands in that debate, it seems that the Afghans who worked for the U.S. and were promised a ticket out if and when things went south, have a strong moral claim.
For one thing, many of the promises were explicit. For another, these were not gratuitous promises but promises in exchange for services rendered. In addition, Afghans who worked for the U.S. relied to their detriment on the promise of sanctuary should the Taliban return (as they have). Hence, the core elements of an enforceable legal promise are present, even if, for jurisdictional or other reasons, the promises aren't actually enforceable in court.
Debt. Evolutionary biologists sometimes contend that much of our conventional morality encapsulates the sorts of mutually beneficial rules and principles necessary for social cooperation, including strong norms of reciprocity. Our social norms and attitudes reflect this idea. If someone does something for you, you feel obligated to do something for them. That's especially true where, as in Afghanistan, they took serious risks in helping you.
* * *
Any one of the foregoing considerations standing alone would suffice to create an agent-relative obligation to the Afghans who assisted the U.S. In combination, they make an extremely powerful case, which explains why so many people (including me) who think that the U.S. withdrawal from Afghanistan is justifiable are nonetheless extremely distressed by the inadequate steps taken to ensure the safety and evacuation of more of the Afghans who assisted the U.S. during the last twenty years.
What about the other millions of Afghans who are at what I'll call average risk from the Taliban: every female; journalists; political dissidents; non-Pashtun Afghans; collectively, a clear majority of the country? Some subset of these people are at especially high risk. For example, people who operated schools for girls or journalists who were critical of the Taliban or took positions that the Taliban considers anathema. The U.S. has apparently been helping some such people exit the country if they can get to the Kabul airport, which is itself a very risky proposition. To the extent that these people undertook the activities that now put them at risk in reliance on express or tacit promises from the U.S., we also owe them an agent-relative duty of rescue; clearly, we will end up breaching that duty in many such cases.
Now I want to focus on truly average Afghans who expect to suffer under Taliban rule but not in reprisal for anything they did to aid the U.S. or even in reliance on the assumption that the Taliban would not return. They will suffer because the economy is already crashing and because the Taliban are brutal sexist theocrats, even if their rebranding of themselves as somewhat less brutal, somewhat less sexist, and somewhat less theocratic than two decades ago proves to be genuine. So average Afghans will suffer. Does that mean we owe them a special duty?
Put more precisely: Does the U.S. have any special moral duty to Afghans simply in virtue of the fact that we occupied the country for twenty years? To frame the issue still more precisely, we might ask whether the U.S. has any greater moral obligation to the Afghan people than we would have if the Taliban were now first coming to power and we had never invaded?
You might think that the answer to that question is no, but that we nonetheless have a moral obligation to mitigate the plight of average Afghans based on basic humanitarian principles in the same way that we have a moral obligation to mitigate the plight of suffering people the world over. I don't disagree with that view; I think we have some affirmative agent-neutral moral duties of aid that are not merely supererogatory, although I am very skeptical that armed invasion will usually be a good means of advancing humanitarian goals. It will be sometimes, but humanitarian armed interventions have at best a mixed record. And in some contexts they will be off the table. For example, I think that the world community has a moral obligation to aid the people suffering under the Kim regime in North Korea, but armed force would be counterproductive to the fulfillment of that obligation.
In any event, for present purposes, the ostensible moral duty to aid the people of Afghanistan in whatever way feasible simply because they are people suffering is, as I noted above, an agent-neutral duty. I'm interested in whether we have an agent-relative duty to the people of Afghanistan, one that goes beyond whatever duty we have to people suffering under an equally repressive regime in a country to which the U.S. has weaker recent ties.
On one account, the answer is a straightforward no. We can see how this idea works with a fairly callous hypothetical. Suppose the U.S. had not invaded in 2001 or that the U.S. had sent only a small force that hunted Osama bin Laden and then quickly left. In that counterfactual world, the Taliban would have been in power continuously for the last twenty years. And thus, in the callous logic, we made the people of Afghanistan better off than they otherwise would have been for the last twenty years, so far from owing them anything, we owe them less than we owe to the people of other countries who have not been suffering.
That suggestion seems wrong, however, even if we set aside all the ways in which the last twenty years in Afghanistan have been at least as difficult or more difficult for many people than they would have been in the counterfactual world. But why does it seem wrong?
The answer, I think, relates to what Professor Colb and I called the "moralized endowment effect" in our book Beating Hearts. There, we hypothesized that the psychological phenomenon of the endowment effect (and related phenomena like loss aversion)--whereby losing something one has is experienced as more costly than not obtaining it in the first place--has a moral dimension. I won't repeat our argument here, except to note that a great deal of law is based on the proposition that existing endowments have normative content.
Accordingly, I think it fair to say that by raising expectations in Afghanistan--by endowing Afghans with something even barely resembling a functioning economy and human-rights-respecting polity--we created a kind of moral claim. Thus, even Afghans who are better off now than they would have been had the U.S. never invaded in 2001 have an agent-relative claim against us for abandoning them.
Finally, in saying all of the above, I am not proposing any particular course of action, least of all further offensive military action by the U.S. How the U.S. satisfies its agent-relative moral duties to the people of Afghanistan--by accepting large numbers of refugees, sending economic aid, applying diplomatic pressure to mitigate the Taliban's policies, or, most likely, by simply forgetting about the Afghan people again and thus breaching our duty towards them--is a separate question from whether the duty exists.