The First Method: Rejecting a Principle

Sometimes a dilemma arises because acting one way violates one principle, but acting another way violates another principle. For example, when UNHCR was deciding whether to help with repatriation, it perhaps needed to follow a principle of voluntary consent. This principle is often presented as:

For an example of this principle outside the area of migration, imagine that a boxer consents to take part in a boxing match. If the boxer provides consent because someone will shoot them if they do not provide it, they have not given voluntary consent to take part. Others should therefore not agree to fight this person, or to even make it possible for the boxing match to take place, such as by renting out a boxing ring.

If this principle is right, then when refugees are told they will be detained or forced into camps if they do not repatriate to unsafe countries, their consent to repatriate is not voluntary, and so others should not help them repatriate.

However, in some cases the voluntary consent principle may conflict with another principle:

The above principle is often applied to individuals initially fleeing their home countries: if individuals risk being killed in their home countries due to persecution, and so wish to save their lives by escaping to a neighboring country where they risk being detained, it seems permissible for an aid organization to help them escape to the neighboring country. This is true even though their escape—and so their consent to assistance with escaping—is involuntary. For example, it seems permissible for an aid organization to arrange transport and travel documents for Gazan residents trying to escape to Egypt, even if further injustice in Egypt awaits, if the Gazans will be safer in Egypt and this is what they want from the options the aid agency can provide. Helping refugees in this way is a type of humanitarian assistance. Similarly, when refugees wish to repatriate because of injustice in their countries of asylum, it seems permissible to help them repatriate if this is what they want from the options available, assuming repatriation is better for them compared to staying where they are.

We therefore face a conflict between the principle of voluntary consent and the principle of humanitarian assistance.

One way this conflict can be avoided is by demonstrating that one of the principles is false. For example, if the principle Voluntary Consent is false because valid consent can in fact be involuntary, there need not be a dilemma: providing repatriation is permissible because refugees can give their valid consent despite being coerced, and is also permissible because it is a form of humanitarian assistance.

Here is one reason to suppose Voluntary Consent is false. It begins with the premise that there are (at least) two kinds of coercion:

1. 1. Coercion where one party wrongly threatens to harm a second party unless they consent to a third party acting a particular way, but the third party does not need to act this way to help the second party avoid harm.

2. 2. Coercion where one party wrongly threatens to harm a second party unless a third party acts in a particular way, and the third party must act this way to help the second party avoid harm.

The case of the boxer is the first sort: To avoid being harmed, the boxer needs to consent to fighting, in the sense of providing someone consent for the match, but they will not be harmed if nobody actually serves as an opponent. When a person is wrongly threatened with harm unless they provide consent for some action, but will not be harmed if nobody partakes in this action, then their consent to the third party to engage in this action is invalid: The boxer’s consent to box is invalid. This is because their consent does not signal to others that they actually want to fight: if the boxer is only saying they consent to avoid being harmed, and this harm can be avoided if they provide consent and there is no actual opponent, the boxer clearly just wants to say they consent without anyone actually fighting them. In contrast, a person’s consent can be valid with the second type of coercion. For example, if someone threatens to shoot the boxer unless they actually box—that is, actually show up at the ring and fight someone else—it is more plausible that the boxer can give valid consent to the person they fight if this person cannot help them in any other way. That is because the boxer’s consent signals to the innocent opponent that they prefer boxing to getting shot, and boxing is the only way the innocent opponent can prevent them getting shot. If the boxer’s consent is valid despite being involuntary, then it is not true that consent is invalid whenever involuntary. If so, the principle Voluntary Consent should be rejected.Footnote ⁷

If it should be rejected, we need to reformulate it. A reformulated variant might disambiguate between the two types of coercion and conclude that the second type of coercion is compatible with valid consent. If it is, then when aid agencies help refugees repatriate—acting on refugees’ involuntary consent—the consent they obtain can be valid if refugees want to actually repatriate given the risks of remaining. If refugees’ consent is valid, and it is also permissible to help them repatriate because it provides humanitarian help in accordance with the Humanitarian Assistance principle, then there is no conflict between principles.Footnote ⁸

Another possibility is that the original Voluntary Consent principle is correct,Footnote ⁹ but we should reject Humanitarian Assistance: organizations should not always help victims escape injustice when this leads to the victims facing another injustice instead, even if the victims want the help provided. Or perhaps organizations should help victims in these cases—Humanitarian Assistance is correct—but they commit a pro tanto wrong because consent is invalid.

There are more possibilities. For now, I hope only to have illustrated that we can examine whether a dilemma exists by evaluating if the principles that conflict are plausible. If they are not, there may be no conflict between plausible principles after all.

The Second Method: Utility

Even when there is a conflict between two plausible principles, it may be possible to find a third adjudicating principle that establishes what ought to be done.

Here is an example that also involves repatriation, but with money: Money is sometimes paid to refugees and other migrants agreeing to repatriate home. For example, Australia paid Afghan refugees to repatriate in 2003, and Japan paid Brazilian migrants to repatriate shortly after.Footnote ¹⁰ These cases raise a dilemma because, on the one hand, the money helps refugees and other migrants who wish to return home. On the other hand, the money can express just how much migrants are not wanted, sending an offensive message.

The clearest example of this dilemma is in Sweden, which from 2026 will start paying migrants wishing to repatriate $34,000, a policy promoted and endorsed by the Sweden Democrats party. This party was previously neo-Nazi, and while it has rebranded itself as conservative, its members still hold some objectionable views. Let us assume (for the purposes of this argument) that the views really are objectionable and that some parliamentarians endorse the policy for racist and xenophobic reasons.Footnote ¹¹ Such policymakers obviously do not think of themselves as facing a dilemma—they just want migrants gone—but other policymakers or citizens might: those who wish to welcome migrants might not be sure whether to support or oppose the policy. Supporting the policy could signal a dislike for migrants, but opposing the policy—such as by actively lobbying for it to be overturned—could mean migrants do not receive money they wish to receive so that they can live in countries to which they want to return.Footnote ¹² The latter consideration seems especially important if we assume—which I will for simplicity—that migrants are returning to safe countries. If return is safe, there is a conflict between the wrong of not supporting the objectionable motives behind the policy and the right of helping individuals who want to safely return home.

This dilemma could be expressed as a conflict between two principles. Here is one:

For example, imagine someone cries to refugees, “Go home!” without necessarily deporting them, as when U.K. foreign secretary Theresa May in 2013 paid for billboards throughout the country telling migrants they were not wanted.Footnote ¹⁴ If May acted wrongly, then according to the Wrongful Message principle it would have been wrong for others to express support for her message, such as by saying, “Theresa May is right to express support for migrants leaving.” When states pay migrants to repatriate, they are often implying or stating that they want migrants to leave. If it is wrong to express this message, others should not show support for a policy that implicitly expresses this wrongful message.

Now, not all will agree with this conclusion. Some might think that if a policymaker endorses a policy that implies a dislike for migrants, then others commit no wrong in supporting the policy if their reasons for supporting the policy are not based on a dislike of migrants. However, I want to put aside this possibility here: let us assume that the principle is correct and implies there is something wrong about supporting a policy of paying migrants to repatriate.

Let us also assume that it conflicts with this other principle:

If we assume that migrants obtaining the money to repatriate home are fulfilling their right to return home—a right they cannot really obtain without the funds to do so—then the money could be of value. This seems especially relevant for refugees with a moral and legal right to return to the countries from which they fled,Footnote ¹⁵ and who can exercise this right with the money received.

To be clear, the above claim is not that the anti-immigrant policymakers who want migrants to leave are doing something that is morally virtuous or even morally permissible if the money helps: the claim is that those who are not motivated by wrongful sentiments themselves seem to have strong moral reasons to not block the policy if it helps refugees and other migrants repatriate home.

If there really is a dilemma in this case—that is, a conflict between two principles of equal importance that both stand—there is a question of how to resolve this dilemma.

One way to resolve this dilemma is to consider whether a third principle can adjudicate between the two that conflict.

Here is one contender for such a principle:

Welfare has value,Footnote ¹⁶ I assume, and while always maximizing welfare can lead to grievous wrongs—for instance, it is wrong to murder a person to harvest their organs to save five other livesFootnote ¹⁷—in this case there is no such counterintuitive implication. For example, it seems intuitively compelling to conclude that the policy should be stopped if, though migrants who obtain the money benefit, the money reduces aggregate welfare for all migrants because of various negative effects. Negative effects could include reducing refugees’ ability to integrate because more Swedish citizens view them as outsiders, and because all migrants feel alienated from society by the policy. In fact, a recent Swedish government report concluded just that, stating that the harms to migrants would be substantial and so the policy should be abandoned.Footnote ¹⁸ That is the right course of action if the Utility as Tiebreaker principle is correct.

However, as it stands, the principle is not quite correct. Clearly, some welfare should not be counted in determining what ought to be done. For example, imagine a refugee taking the money to repatriate would likely be killed once home, and their organs used to save another five refugees’ lives. If this (unrealistic)Footnote ¹⁹ scenario somehow led to greater aggregate utility compared to blocking the policy, this would not be a reason to support the policy. So, considering aggregate utility is sometimes wrong.

The above is just a particular example of the more general problem of appealing to aggregate utility. Perhaps there is a principle of aggregate utility that avoids counterintuitive implications, and some have certainly tried to find one.Footnote ²⁰ However, if we cannot find one that is compelling, we need to look for an alternative adjudicating principle.

The Third Method: Democracy

Another principle to adjudicate between two that conflict is democracy.

Democracy often provides a solution for what should be done when reasonable disagreement arises.Footnote ²¹ For an example unrelated to immigration, individuals might disagree as to whether to legalize some type of recreational drug, with some supporting legalization (with good arguments as to why) and others opposing legalization (with good arguments as to why). When this happens, some claim the majority should decide what to do. We could make the same claim when there is a dilemma regarding migration: democracy should determine what a person or organization ought to do.

That is because, when there is a dilemma, I assume there are two equally important values that cannot both be promoted, such that it is unclear what one has an all-things-considered reason to do. If that is right, one way to determine what to do is to change the circumstances so that one of the actions or policies promotes an extra value not found in the other action or policy. Putting the matter up for a vote has this effect: Whatever policy wins promotes democratic value, tipping the balance to endorsing this policy. In other words, if voting on which of two actions to take or policies to implement means there is democratic value in pursuing one and not another, then even if there are two equally important values that conflict, at least we can conclude that there is an additional value (that of democracy) that points toward one action or policy and not the other.

The reason democracy adds this type of value is partly that it gives people power. For democracy on the state level—as when citizens vote in elections and plebiscites—this is important because the state has power over citizens, including being able to tax and imprison them, and so it is important that citizens have power over their government via their vote.Footnote ²² Organizations are different—they cannot tax and imprison—but they still have power over people’s lives.Footnote ²³ For example, human rights organizations have power over migrants’ lives by determining how and when to campaign on their behalf. If organizations have power over migrants’ lives, and democracy has value by giving those subject to power some power in turn, it follows that giving migrants some power over what the organization does has value. If so, then when an organization faces a dilemma and is not certain what to do, it can organize a vote that all registered migrants can take part in, helping determine which of two conflicting actions to take. For example, Swedish human rights organizations wishing to help migrants might be unsure whether to endorse the new payment policy, and can determine what to do by asking all registered migrants whether to endorse the policy or not.

Organizations also have a more subtle kind of power: They can impact whether people commit wrongs. In particular, if an organization commits a wrong in its actions, it might implicate its members, including those working and volunteering for the organization. This is true if organizations are comprised of their members, such that if the organization commits a wrong each member does as well. For example, if an organization campaigns for a policy it morally should not campaign for, this could mean each member engages in wrongdoing. Or perhaps when an organization commits a wrong only some members are implicated, such as those who are aware or ought to have been aware of the organization’s actions. Regardless, at least some organizations in some circumstances implicate some organizational members when committing a wrong.Footnote ²⁴ If the organization’s wrong implicates these members, the organization has power over whether these members commit a wrong. Moreover, if members are implicated in this way and their lives are worse as a result—for example, they might feel guilty if their organization spends money or campaigns wrongfully—then the organization has the power to make members’ lives go worse.Footnote ²⁵

If an organization has power over members’ lives in this way, members should have power to influence an organization’s decision-making as well. They can if their vote impacts what the organization does. If so, then perhaps not only migrants should vote on what the organization does, but some or all members as well. If most migrants and members want the organization to support a given policy, then supporting this policy progresses a value of democracy, potentially tipping the balance in favor of supporting this policy when a dilemma arises.

This still leaves important questions unanswered, such as whether members’ votes should hold the same weight as migrants’ votes; perhaps the power organizations have over migrants is greater than the power they have over members (especially if members are particularly well off) and so migrants’ votes should hold greater weight. There also remain pragmatic questions, such as how members and migrants would cast votes in practice. For example, it is not clear how to determine who are migrants, and so who has a right to vote. Perhaps the organization could widely publicize a call, via partner migrant organizations, for any migrants wishing to vote to send proof of their migrant status. To ensure privacy, proof could be brought in person or over a video call. Or, if there are problems with such a method—such as unauthorized migrants being nervous of informing an organization of their legal status—then the organization could use an honor system and let all those identifying as migrants vote. While there might be risks of nonmigrants posing as migrants, these risks could be lower than any risks that asking for proof might pose.

More analysis is necessary to consider the ethics of how to organize democratic decision-making among migrants, but for now my point is only that if democracy has value, it is one method for resolving difficult dilemmas.

The Fourth Method: Random Selection

Another method for resolving dilemmas is random selection.

To see how this solution can be applied, consider a dilemma faced by UNHCR when deciding which refugees to resettle from low-income countries (like Kenya and Jordan) to wealthier countries (like the United States and Australia).Footnote ²⁶ The agency often prioritizes refugees who are most in need, including those facing gender-based violence (GBV), medical complications, and malnutrition. Because all refugees with these needs cannot be resettled—given that wealthy countries refuse to accept more than a very small percentage of the world’s refugeesFootnote ²⁷—a dilemma arises in choosing who to select. For example, when resettling those fleeing GBV will protect women’s rights, whereas resettling those with disabilities will protect disability rights, it is not clear what ought to be done when the two groups cannot both be resettled.Footnote ²⁸

Because of this dilemma, UNHCR asks refugees to come in for multiple interviews, just to be sure it is selecting the refugees who are the most badly off.Footnote ²⁹ However, refugees are then forced to tell their stories again and again, sometimes training themselves to tell their stories in the most compelling manner.Footnote ³⁰ This can cause trauma, partly because refugees are forced to regularly recall in detail what they experienced, and partly because they cannot decide for themselves how to present their life narratives, forced to distort these narratives to fit the rubric of what UNHCR requires for resettlement.Footnote ³¹ Their narratives can also be an inaccurate way of determining who ought to be resettled, as some refugees are better than others at explaining and remembering their lives and histories. In fact, it may be that those who struggle most to tell or remember their stories are particularly vulnerable.Footnote ³² Moreover, there is something wrong about a system that requires refugees to compete between each other to show they suffer the most, where the winners are determined by relatively privileged UNHCR protection officers who decide if they think the refugees they interview are suffering enough or not. This can lead to distrust between refugees,Footnote ³³ and place them in a position where they are dominated by the protection officers in a certain sense.Footnote ³⁴ While some domination might be inevitable, the domination seems to continue if refugees must prove again and again just how vulnerable they are, often begging to be selected.Footnote ³⁵

Given these problems—of trauma, inaccuracy, and domination—interviewing refugees multiple times is not always a good solution for determining who to resettle when many refugees seem equally badly off.

Here is another solution: Use a lottery. More specifically, UNHCR could first determine who is sufficiently in need to be a contender for resettlement, and then run a lottery to determine who is ultimately selected.

The idea of running a lottery to determine selection is not new. For example, the United States runs a green card lottery, where anyone in the world can pay a fee to try to win a visa to live in the country.Footnote ³⁶ Indeed, many refugees enter the green card lottery to try to escape difficult conditions in transit countries.Footnote ³⁷ Lotteries are also run for the provision of aid for refugees and other at-risk populations.Footnote ³⁸ For example, when the organization GiveDirectly first gave out cash grants to families living in Kenya, only families randomly selected were given grants, so that the organization could compare the outcomes of those randomly selected to the outcomes of those not given aid.Footnote ³⁹ Lotteries are also used in some democracies when citizens are randomly selected to take part in committees that give recommendations to the government.Footnote ⁴⁰ Lotteries are even used for allocating medical treatments, such as when during the COVID-19 pandemic some hospitals randomly distributed finite doses of lifesaving medicine.Footnote ⁴¹

Lotteries can also be a good solution when facing dilemmas of who to resettle. This is partly because, unlike UNHCR interviewing refugees multiple times, a lottery does not force refugees to repeat their stories, or disadvantage those who are not good at telling compelling stories. Importantly, the refugees who lose a lottery at least had an equal opportunity to win, given that a lottery does not discriminate against those who are not as talented at telling their stories. A lottery, in other words, serves as a type of justification to refugees who are not selected: they were not selected because a transparent and random method was used, rather than because they did not demonstrate sufficient vulnerability to protection officers.Footnote ⁴² Importantly, this transparent and random method can be defended on egalitarian grounds: each has an equal chance of being selected, even if some are selected and others are not.Footnote ⁴³

If each has an equal chance of being selected, there may no longer be a conflict between two values. For example, if one refugee faces gender-based violence, and another disability-based immobility, selecting the first refugee and not the second via a lottery does not necessarily lead to the value of women’s rights being progressed at the expense of disability rights: Disability rights are respected in the sense that the refugee with a disability had an equal chance of being resettled (even if they were not). If there is no longer a conflict between two values, then the lottery serves as a solution to the dilemma.

It is also worth noting that even if one thinks there is still a conflict between two values with a lottery—maybe if one refugee is randomly selected and another is not, there remains a conflict between the value of protecting the first refugee and not the second—a lottery at least provides an answer as to what ought to be done. That is because a lottery promotes equality, given that each person has an equal chance of being selected.Footnote ⁴⁴ If a lottery promotes equality, and equality is fair, then fairness is promoted to an extent. If so, a lottery will mean we can make this claim about some refugees: they are both badly off and selecting them is fair. Selecting these refugees is therefore justified on an all-things-considered basis.

Importantly, if it is justified partly on grounds of fairness, it is an improvement over some lottery-like elements currently in place. For example, under a U.K. policy Afghan refugees can more easily qualify for resettlement if a humanitarian organization refers their case to the U.K. government, and this has been criticized as “like a lottery”: refugees who just happen to meet someone in an organization who happens to refer them for resettlement to the U.K. can get resettled, while others cannot, and this is unfair.Footnote ⁴⁵ If today’s resettlement is metaphorically like a lottery in an unfair sense, then creating an actual lottery—where every refugee has an equal chance of being selected, rather than prioritizing those with connections—is an improvement.