Setting the stage

Some assumptions will help to frame our inquiry. Start with the assumption that a country suffering an unprovoked armed invasion has multiple moral grounds for fighting back. The just war literature highlights three such (non-mutually-exclusive) grounds.Footnote ⁵ First, defending citizens’ individual rights. Second, sustaining the national community as a self-governing entity. Third, upholding the international norm of non-aggression. I put aside complicated cases where the grounds for defensive war are not entirely clear. Instead, I focus on cases of the Ukraine sort, where the justifiability of defensive war is plainly over-determined.

Other key assumptions concern the individual duty to fight. Throughout this article, I assume that if (and only if) all standard just war requirements are met, fighting-age citizens have a moral duty to participate in a just defensive war as a way of doing their fair share in sustaining just institutions. Outside of special cases – such as those where medical and/or family circumstances clearly remove the duty to fight – those citizens who fail to participate are unfairly free-riding on others’ protective efforts. This assumption, to be clear, allows for the (intuitive) thought that citizens’ duty to fight in a just defensive war is suspended if their democratic government surrenders or negotiates peace. Yet the individual duty to fight is not itself derived from a governmental decision. For example, even if Putin’s invasion had temporarily succeeded in neutralizing Ukraine’s elected government (leaving, in its stead, a puppet government or simply a power vacuum) ordinary Ukrainians would still have had a moral duty to fight the invaders (and to revive their democracy), so long as the fight had a reasonable chance of success.Footnote ⁶

Internal conscription, in turn, is a tragically familiar means to which a democratic government might resort to enforce its citizens’ moral duty to fight. More precisely, I assume that although conscription, as John Rawls put it, is ‘a drastic interference with the basic liberties of equal citizenship’, internal conscription can be morally permissible when it is a necessary means for achieving the ‘security of liberty’ (Rawls Reference Rawls1999, 333–334). The security of (individual) liberty is surely at stake in the relevant way when conscription is essential for safeguarding individual citizens’ basic rights. But the same is true when internal conscription is vital for collective liberty – for safeguarding the national community’s continued existence as a sovereign entity, and/or for upholding the international norm of non-aggression, without which international society might very well devolve into ‘universal tyranny’ (Walzer Reference Walzer2000, 59).

Our question is how these internal conscription assumptions bear on the justifiability of foreign countries using border controls to push an invaded country’s citizens to participate in armed resistance. Foreign governments can engage in such external conscription directly – by announcing that they will bar draft-eligible citizens of Country A from entering their jurisdiction for the sake of sustaining A’s war effort. But a given foreign government, B, can also pursue external conscription indirectly, by pressuring other countries (C, D, E, …) to close their borders to the same end. The following arguments are meant to apply to both these variants of external conscription.Footnote ⁷

External conscription as law-enforcement: the simpler cases

External conscription may initially trigger considerable alarm. This is especially true if we imagine a major international power that pursues a particularly ambitious form of external conscription, combining both the direct and indirect variants just noted. Imagine, for instance, that as the war in Ukraine started, the United States not only closed its own borders to draft-eligible Ukrainians but also tried to deploy its considerable geopolitical clout to incentivize virtually all other foreign countries to similarly deny entry to such Ukrainians. Even if all of these border closures featured commonsensical exemptions of the sort found in international law (for instance, for pacifistsFootnote ⁹ ), it might very well have seemed alarming for the United States to effectively trap Ukrainians inside their own country to get them to fight.

However, we can begin to temper this alarm by noting two crucial points. The first has to do with a straightforward moral symmetry between internal and external conscription. Even a foreign government that goes out of its way to ‘trap’ an invaded country’s citizens inside their own borders to push them to fight is no more guilty of such trapping than a domestic government that announces martial law and prohibits its conscription-eligible citizens from leaving.Footnote ¹⁰ Yet hardly anyone denies that the latter ‘trapping’ can be justifiable under emergency circumstances. Why think that the former is so morally different?

The second point is that, morally speaking, external conscription is unremarkable when it is a straightforward legal corollary of internal conscription. Suppose, for instance, that following the Ukrainian government’s announcement of internal conscription, many foreign governments publicly announced that they would bar draft-eligible Ukrainians from entering their respective jurisdictions to help Ukraine uphold its conscription laws. Arguably, such a scenario would not be very different from a scenario where all foreign governments announce that they will bar the entry of Ukrainians who have been charged by the Ukrainian government with, say, tax evasion. The latter case merely reflects a commonsensical commitment on the part of the relevant governments, not to enable another country’s citizens to break that country’s laws. But a government that allows foreign draft evaders to enter its jurisdiction is enabling the violation of their country’s laws in much the same way.

This point arguably means that external conscription triggers genuine philosophical complexity only when it is independent of any internal conscription laws. In what follows, I am going to focus on cases of this sort, where we cannot defend external conscription simply by appealing to legal obligations imposed by the (prospective) conscripts’ own government.Footnote ¹¹ This focus obviously narrows the scope of the argument. Even so, the universe of relevant cases remains large enough to be of both philosophical and practical interest. If nothing else, this is because the ‘fog of war’ is often sufficiently thick to require not only prompt action from each relevant government but also action in the absence of any certainty regarding other governments’ (quickly evolving) positions.Footnote ¹² Moreover, even if such urgent policy decisions amidst deep uncertainty are disproportionately concentrated at the very beginning of an armed conflict, it is also true that these decisions carry outsized practical weight. If nothing else, that is because war – and particularly, collective resistance in war – is so often path-dependent.

A prime example of (awareness of) this path-dependence was President Zelensky’s extremely public decision to stay in Kyiv at the very start of the Russian invasion instead of accepting the American offer of a secure exile. When Zelensky famously proclaimed ‘The fight is here; I need ammunition, not a ride’,Footnote ¹³ he was not merely illustrating an actor’s penchant for memorable lines. He also recognized that heading abroad at the very beginning of the war was not akin to visiting foreign capitals later in the conflict. This was because the earliest stages of the battle of Kyiv mattered disproportionately for sustaining and bolstering Ukrainians’ collective belief in the possibility of a successful defence of their country. That belief, in turn, strongly depended on the probability that each citizen could rationally assign to other citizens joining and staying in the fight. Zelensky’s example – and possibly even his announcement of martial law – may have meaningfully impacted that probability of collective action ‘only’ during a specific time window at the very onset of the conflict. But to the extent that he did shape that initial period, however brief, Zelensky arguably altered the course of the war. The same, I would suggest, might very well be true for external conscription.

Whose authority is it anyway?

Equipped with these preparatory remarks, we can now consider the first account attempting to justify the normative discrepancy between internal and external conscription. On this account, the discrepancy has to do with moral authority. Internal conscription can be justified because a democratic government can have the moral authority to enforce its own citizens’ moral duty to fight in a just war. But foreign governments lack this kind of authority: it is simply not the business of foreign governments to try to compel another country’s citizens to fight another country’s war. Call this the No-Authority view.

The main problem with this view is the fact that it ignores the normative impact of emergencies. Internal conscription, as we have already emphasized, can only be justified in emergency circumstances. But emergencies are, by definition, humanity’s business. This is because, at least outside of threats to other species, emergencies feature a threat (often, a large-scale threat) to the most basic human interests. And the most basic human interests, in turn – such as the interest in physical survival and freedom from violence – are properly a subject of universal concern. In turn, since just defensive wars of the sort in which we are interested here are fought (at least in part) to protect the most basic human interests of innocent populations targeted by an aggressor, these wars too are properly a subject of universal concern. It follows that every such war is, in some non-trivial moral sense, humanity’s war: every such war is fought, at least in part, for reasons that humanity as a whole ought not just to recognize but to respond to through non-trivial action.

Admittedly, to say that every emergency is in some sense a universal emergency is to make a large claim. I doubt I can give a complete defence of this claim here. I do, however, hope to show that this claim has a great deal of plausibility, such that the burden lies with its opponents to show its falsity.

One way to evince the universality of human emergencies is to consider the well-known idea that large-scale threats to human rights are properly a subject of unique ‘international concern’ (see, for example, Beitz Reference Beitz2009, Ch. 7). Because human rights protect the most basic human interests, macro-level threats to these rights constitute an emergency, which uniquely licenses extraordinary international action. This thought is especially familiar in those contexts where the government in whose jurisdiction the emergency takes place is unable and/or unwilling to uphold its own citizens’ human rights. Under such conditions, it is often thought fitting for international actors to use their coercive capacities as an emergency ‘backup’ (Song Reference Song2015).

A proponent of the No-Authority view is likely to object that these standard international involvement cases differ from external conscription since the former typically aim to coerce powerful domestic actors who have culpably caused large-scale emergencies (murderous rulers, for instance). External conscription, by contrast, features international coercion of innocent civilians. Underlying this objection is (presumably) the following thought. Even in emergencies, coercion aimed at getting innocent people to respond as they ought to may only be carried out by agents who have a special relationship with the coerced. That is why compatriots might be justified in coercing their fellow citizens to fight in at least some just wars, while strangers cannot be so justified.

This position, however, encounters at least two major difficulties. First, the immediate priority in emergencies ought to be addressing the emergency as effectively as possible, rather than assigning culpability for the emergency or punishing the (most) culpable. Accordingly, if it makes moral sense, in the context of international emergencies, to single out the most powerful domestic wrongdoers as targets of international coercion, the reason is not their distinct culpability per se. Rather, coercing the culpable is uniquely appropriate when (and because) doing so is likely to be the most effective response to the emergency – for instance, leading an internally repressive regime to stop its violence against a minority of its citizens. But if effectiveness rather than culpability is the fundamental moral concern in an (international) emergency, then there cannot be categorical constraint against international coercion that aims to get the innocent to do their duty in an emergency.

Second, when we are trying to ascertain different agents’ moral duties in the face of a given problem, it can often be important to distinguish between those who have and those who have not (culpably) caused the problem. But this distinction is far less salient in (1) emergency circumstances where (2) we already know the content of different agents’ respective duties, and where our concern is simply with enforcing these duties. In these specific contexts, culpability is not a morally relevant factor – and neither is the absence of any special relationship between the duty-bearers and the agent(s) coercing them into doing their duty.

Suppose, by way of illustration, that you witness an ongoing robbery in a park and rush in to help. Much stronger than the robber, you force him to flee with only a fraction of the victim’s possessions. Among these stolen possessions, however, is the victim’s inhaler, which was urgently missed because her ordeal induced an acute asthma attack. Another park visitor, a passer-by, has an inhaler on hand, which he does not need at the moment, yet he refuses to give it to the victim despite your pleas. So, you have to resort to threats, telling the passer-by that you will name and shame him before the entire world if he refuses to do what he (plainly) ought to do. The robber surely couldn’t invoke the fact that you are a total stranger to deny the permissibility of your coercing him into abiding by his duties towards the victim. So why should the passer-by’s case be any different when you’re coercing him into doing right by the victim? The passer-by didn’t cause the victim’s predicament in any way, whereas the robber did. But why think that this distinction is morally salient here?

Now, to be sure, the passer-by’s moral duty to offer help in this case is far less onerous than citizens’ moral duty to perform military service in the face of an armed invasion. However, in emergency settings, even onerous duties can permissibly be enforced by agents who have no special relationship with the duty-bearer(s), for much the same effectiveness reasons noted above. Consider the following escalation of medical duties:

The physician’s duty to save a life on board is clearly quite onerous. Still, so long as the burdens he faces are not so heavy so as to altogether remove this duty, you remain justified in enforcing it.

But there is a further, deeper point that should be noted in response to the onerousness issue. The No-Authority view cannot coherently appeal to the burdens of military service to salvage the normative distinction between internal and external conscription. After all, these burdens remain the same regardless of whether conscription is enforced (entirely) from the inside, by one’s government, or whether it is enforced (at least in part) from the outside by foreign governments.Footnote ¹⁴

Bearing this crucial point in mind, consider a further example, inspired by an actual case from the Russian invasion of Ukraine. Two months after the start of the invasion, the Washington Post reported the story of a decentralized network of Belarusians who had been sabotaging the railways essential to the Russian onslaught:

That these Belarusians have gone beyond the call of moral duty is evident, not least because they have operated within a brutal Kremlin-controlled dictatorship, which (predictably) inflicted appalling retribution on any apprehended ‘railway partisans’ (Sly Reference Sly2022). Yet we can modify the Belarusian tale so as to bring it within duty’s call. Consider:

In this scenario, the foreign workers have limited reason for fear if they hinder the Russian war effort. So, they arguably have a moral duty to do so. And this duty, in turn, is enforceable by anyone capable of enforcing it, given the emergency stakes. The most senior foreign hire may not share a nationality (or any other special relationship) with the other foreign workers whom he is threatening. But, morally speaking, that is immaterial, just as it is immaterial whether you share any special relationship with the doctor whom you might compel to provide assistance in a medical emergency. Here as there, the underlying reason – and the underlying normative lesson – is the same: every emergency is humanity’s emergency, and, as such, is humanity’s business.Footnote ¹⁵

Before proceeding, I should clarify the scope of this argument. The argument should not be taken as providing an indiscriminating moral license for any agent to enforce others’ emergency duties. This is mainly because, even within emergency contexts, there are typically weighty reasons to leave such enforcement in the hands of relevant state authorities – if these authorities have the will and the capacity to pursue an adequate response. Of course, the destruction or disappearance of a functioning, reasonably well-ordered state itself constitutes a major emergency. But whenever a macro-level emergency leaves the relevant well-ordered state intact, both other states and non-state actors should have a strong presumption in favour of deferring to that state in shaping a response.Footnote ¹⁶ This presumption means, among other things, that outsiders should be reluctant to second-guess the judgement of a reasonably well-ordered state that is capable of enforcing conscription but chooses not to enact it, or that opts for a more limited conscription scheme than may seem optimal from the outside, due to its own calculations as to how to best sustain the war effort.Footnote ¹⁷

Fairness

Let us now consider another account that attempts to explain the problem with external conscription without impugning internal conscription. This account takes the idea of ‘humanity’s emergency’ in a different direction. Here the thought is that its distinct dependence on the universality of emergencies renders external conscription far less fair and morally coherent than internal conscription (which need not be grounded in any such universality). External conscription (the complaint goes) singles out the citizens of a particular country, effectively assigning to them a special duty to stay within the country’s territory to fight a predatory aggressor. But if the universality of emergencies is what grounds foreign governments’ authority to engage in external conscription, then there can be no such special duty incumbent upon the victim country’s citizens. Rather, there can only be a universal duty to oppose the aggressor.

Perhaps the best way to assess this argument is to consider its ‘brain drain’ analogue. Multiple philosophers have debated whether brain drain can justify immigration restrictions, paying particular attention to the large-scale migration of medical personnel from developing countries with acute healthcare needs.Footnote ¹⁸ In turn, one important question within this debate has been what – if anything – distinguishes the moral duties of local medical professionals in the face of national health crises, as compared to the duties of their foreign counterparts. If we are too cosmopolitan to insist that ties of nationality as such generate a special duty to stay in one’s country, why single out a physician or nurse from a given country as having a particularly weighty or stringent duty to be ‘on the ground’ in that country (at least if they didn’t receive their training at their compatriots’ expense)? If every large-scale health crisis is truly ‘humanity’s business’, why not think that many medical professionals from other countries face essentially the same moral duty to spend time where they are most needed?Footnote ¹⁹

The juxtaposition with the medical brain drain helps to make the fairness account more familiar. But this juxtaposition also shows the account’s limitations. This is because there is at least one pivotal difference between medical brain drain on the one hand, and, on the other hand, collective resistance in war. This difference stems from the path-dependent nature of war, emphasized in section 2. Locals matter disproportionately in a war against an invader when compared to a problem such as the medical brain drain. To be sure, in the latter context too, the mass-scale departure of locals could certainly derail the collective effort, leaving foreigners with more work if and when they join the cause. But when the collective goal is to thwart an armed invader, mass-scale departure by locals represents a far more acute danger since it threatens to doom the collective effort entirely; at the limit, leaving foreigners with no cause at all for which to fight by the time they (inevitably belatedly) arrive. To know that a sizeable number of local physicians are leaving a developing country might well be disheartening. But if an equal or even larger number of foreign physicians is genuinely committed to replacing the departing local medical staff, there is real reason to hope that any (extra) deficits in the provision of medical care on the ground will be only temporary. That is much less likely to be the case when thinking about an invaded nation.

Consider our central Ukraine example once more. Imagine that, just as the Russian military was bearing down on Kyiv, a hundred thousand more of the city’s draft-eligible residents had fled abroad. This kind of development would have hardly been offset by news that an equal number of foreign volunteers were making real steps to travel to Ukraine to join the fight, since these foreigners’ contributions might well have happened too late to really matter. To be sure, staving off a large-scale military crisis can be just as universal a moral goal as staving off a large-scale health crisis (a point to which I return in the final section). Still, in the former case much more than in the latter, path dependence generates a unique role for locals.Footnote ²⁰

A hybrid view?

I now want to examine a possible hybrid view, combining claims about fairness and authority. The gist of this view can be summarized as follows. It might be true that every emergency is a universal emergency and it might also be true that, in the emergency context of resistance to an unjust invasion, there are reasons to focus on mobilizing the invaded country’s population. Nonetheless, foreign societies still lack the authority to enforce that population’s moral duty to fight, so long as these societies refuse to incur their own fair share of the costs involved in the war effort.

Let us grant, at least arguendo, that a country cannot justifiably pursue external conscription if it refuses to undertake costs that it ought to be willing to incur for the sake of the same just war. So, for instance, a country which ought to be sending tanks and extensive munitions to an unjustly invaded nation, but which instead limits itself to sending helmets (think of Germany at the onset of the Ukraine war), lacks the moral authority to pursue external conscription of that nation’s citizens. Yet, even if granted, this point fails to distinguish external from internal conscription. Rather, it merely shows that both kinds of conscription can be impermissible if pursued in manifestly unfair ways. So the hybrid view must feature some further element if it is to set internal and external conscription apart.

Arguably the most practically salient (and theoretically interesting) such element is uncertainty about the enforcer’s future conduct. As we already saw, policy decisions about external conscription matter disproportionately at the very start of a conflict. But, at least during that initial stage, a government pursuing external conscription may very well be uncertain about its willingness and/or ability to shoulder its fair share of the war’s burdens over time – if nothing else, because it cannot be sure of its ability to sustain a popular commitment to the war effort among its own citizens. In turn, such uncertainty does not have a clear analogue in the case of internal conscription.Footnote ²² So – a proponent of the hybrid view might argue – this uncertainty explains the normative discrepancy between the two kinds of conscription.

This uncertainty, however, is at least two steps removed from the hybrid view’s desired conclusion. First, the lack of enforcement authority given an agent’s unequivocal refusal to do its current part in an emergency does not entail a similar lack of enforcement authority where the agent is uncertain about its willingness to do its (likely) part later on. To see this, consider the following variant on train ride:

Your reasons not to miss your bank meeting do not carry any more normative weight than the physician’s identical reasons in the case we considered above. If he has a duty to miss this kind of meeting because of a medical emergency on board the train, the same duty applies to you in the same circumstances. But even though you are unsure that you will have the strength of will to live up to this duty at T2 (when the next stop arrives), why must it follow that you lack, at T1, the moral authority to enforce his (potentially onerous) duty to care for the emergency patient?

Second, it is not enough for a proponent of the hybrid view to establish that an agent lacks enforcement authority where the agent is uncertain about their own willingness to play their proper part later on. This is because the uncertainty highlighted by the hybrid view has to do with others’ stances. The thought (recall) is supposed to be that, at T1, a foreign government lacks the moral authority to enforce a moral duty borne by citizens of an invaded country if that government is unsure about its citizenry’s willingness to support that country at T2. But why should we accept this thought? Consider, by way of illustration, one more variant on our medical train tale:

If Next Stop does not bode well for the hybrid view, this is surely even more true for Family Ride. There is no obvious reason to think that we lack the moral authority to enforce emergency duties incumbent upon others at a given point in time, just because of our uncertainty about the future attitudes of third parties. This is the case even where the emergency duties in question are onerous, and even where we have a special moral responsibility to try to convince the relevant third parties to adopt the right attitudes.

Collective choice

I now want to consider a final account of the relationship between external and internal conscription. This account repurposes an idea flagged in section 2. There I noted that we cannot distinguish external from internal conscription by appealing to the wrongness of trapping innocent individuals in a war, because such trapping occurs in both kinds of conscription. The present account therefore invokes an adjacent yet distinct idea, which has to do with the trapping of a collective in a war. Internal conscription might lead individuals to take to the battlefield against their wishes. But, at least in a democratic setting, there is never a danger that internal conscription will lead a society into battle against its wishes. After all, a democratic society that does not wish to fight will refrain from enacting internal conscription ab initio. External conscription, then, is set apart because it pushes an invaded society to fight back regardless of its collective will.

To see why this account, too, goes astray, we should first note that it has both a bold and a modest version. The bold version objects to any foreign policies that meaningfully constrain another society’s freedom of collective choice, portraying any such policies as infringing upon collective self-determination. But this bold version is implausible. It counterintuitively rules out a great many policy measures that are routinely adopted by numerous governments worldwide to affect other governments’ conduct.Footnote ²³

Therefore, we should focus on the modest version of the account. This version only targets those foreign policies that not only affect matters that are fundamental to another society’s collective affairs, but which really force that society to make certain choices against its own wishes. So the question that we need to ask is whether external conscription should really be understood as forcing a society to fight against its collective wishes.

The answer, I believe, is negative, even when we consider the most ambitious forms of external conscription. More precisely, the answer is negative even though external conscription indeed aims to enforce moral duties to fight in a just defensive war.

The analytical key here lies in the precise relationship between ‘forcing’ and ‘enforcement’ of others’ duties. Up to this stage, our discussion reflected the common assumption that to ‘enforce’ someone’s duty simply is to ‘force’ them to do their duty. But, strictly speaking, this assumption is false. Even an enforceable duty can be, and in some cases ought to be, enforced in a way that intentionally leaves the duty-bearer with a real choice as to whether to live up to the duty. Moreover, I now want to argue that any morally defensible form of conscription – be it internal or external – represents this kind of duty-enforcement. If this argument is correct, then no morally viable form of conscription, whether internal or external, actually ‘forces’ any agent – be it individual or collective – to fight.

The argument’s starting point lies with a general distinction between two kinds of duty enforcement. First, there is what we might call outcome enforcement. Outcome enforcement is morally permissible when the enforcer, simply by using coercion, can achieve all relevant moral goals, entirely independently of the duty-bearer’s attitudes or behaviour. Suppose, for example, that, having been found liable in a civil suit, Jones is now under a duty (both legal and moral) to pay hefty damages to the victims of the extremely harmful lies he has spread. Suppose further that state authorities enforce this duty by unilaterally tracing and confiscating the relevant sums from Jones’ bank accounts. The authorities’ enforcement effort fully achieves its goals insofar as the money ends up with Jones’ victims, and this is true regardless of Jones’ wishes or actions.

By contrast, incentive-based enforcement of a duty is morally permissible when the enforcer’s success depends, at least in part, on the duty-bearer making a meaningfully autonomous choice to comply with the duty. Here the enforcer uses coercion to increase the likelihood of compliance by altering the duty-bearer’s incentive structure. To illustrate, suppose that you want to enforce your son’s duty to apologize to a friend he has badly mistreated. Suppose further that your moral goal is (at least partly) to get your son to better comprehend and respond to the value of friendship. This kind of goal might be advanced by incentive-based enforcement. For example, you might deny your son access to various extra-curricular activities he cherishes in the hope that he might use the resulting free time to contemplate the wrong for which you are sanctioning him. However, getting your son to better recognize the value of friendship is not the kind of goal that you can fully achieve simply through outcome-enforcement, because this goal clearly requires your son to exercise real agency of his own. This goal will not be advanced in the slightest if, for instance, you got the offended friend on the phone and somehow compelled your son to mechanically say ‘I am sorry’ on the other side of the line.

Now, my claim is that, even in an emergency, any morally defensible form of conscription – whether internal or external – involves only incentive-based enforcement rather than outcome enforcement. To make moral sense, conscription, although altering (prospective) conscripts’ incentive structure, must still leave them with meaningful space for autonomous choice such that it remains plausible to say that the duty-bearers chose (even if under real pressure) to comply with their moral duty. If conscription, as we stipulated in Rawls at the outset, can only be justified as a means of securing liberty, it cannot be coherently pursued in a manner that entirely erases any such space for free choice.

What concrete conscription practices does this constraint rule out? Consider an internal conscription policy that not only orders numerous ordinary citizens to join the military, but threatens them with truly draconian punishments in case they refuse – such as permanent separation from their children, life in prison, or, even, at the limit, being placed in front of a firing squad. No liberal democracy can pursue such measures without thereby undercutting the moral values in whose name it purports to be fighting. Of course, conscription laws, just like many other laws, can plausibly be enforced through serious state sanctions, including, at the limit, some prison time. But, whatever else it might include, any morally viable conscription policy will necessarily feature built-in constraints on the severity of these sanctions, to preserve at least a modicum of choice. Thus, for example, telling a would-be-conscript that he will head to prison for the duration of the war if he refuses to enlist alters his incentive structure but still allows him a meaningful degree of freedom to refuse to put his life at risk. By contrast, to tell a would-be-conscript that he would be shot if he did not put on the uniform is to take the matter entirely out of his hands. The latter enforcement measure can never be warranted, even in emergency circumstances, in which the former, tragically, can.

An analogous observation, in turn, applies regarding external conscription. Morally speaking, external conscription also comes with built-in constraints on enforcement methods, and these constraints mean that, just as internal conscription may not actually force (individual) agents to fight, neither should external conscription. Most importantly, no morally viable form of external conscription will seek to bar a foreign society from surrendering in the face of an invasion or from seeking an armistice on terms that are favourable to the aggressor. Similarly, no morally viable form of external conscription will involve draconian sanctions against those who refuse to fight. It is one thing, for example, for a neighbouring country to deny admission to many citizens who ought to stay in their own country to fight in a just war. But it is another thing altogether for this neighbouring country to announce that if these citizens do not fight, it will use its economic might to starve them or to divert the natural resources on which their livelihood depends. Just as with draconian internal conscription measures, these are, again, totalitarian tools that flatly contradict the moral values in whose name the war ought to be fought at all. So, it seems that, once more, we have failed to uncover a fundamental normative difference between internal and external conscription. Instead, we simply discovered that both types of conscription equally ought to be constrained in their enforcement methods.

In lieu of a conclusion: On fighting abroad

I have spent the previous pages arguing against a fundamental moral discrepancy between internal and external conscription. More specifically, I have examined arguments that appeal in different ways to the lack of foreign authority, to fairness, and to collective choices. I have suggested that none of these arguments ground a fundamental normative distinction between the two types of conscription.

This conclusion leaves open the possibility that contingent reasons will generate a difference between the two forms of conscription in particular cases.Footnote ²⁴ But, rather than delving into these particular circumstances, I want to close with more general comments regarding the seemingly radical implications of my argument. Suppose that, per the last section, an unjustly attacked democracy chooses to resist the attacker, and that (per section 4) its citizens play a distinct, crucial role in sustaining the fight in the early stages of the war. At least if the conflict lasts long enough, could there arise a point in which foreign countries may permissibly support the war effort by conscripting their citizens and sending them to battle abroad?

My answer is a qualified ‘yes’. With the aforementioned Rawls, I believe not only that conscription can be permissible ‘if it is demanded for the defense of liberty itself’ but also that the relevant liberties include ‘those of persons in other societies as well’ (Rawls Reference Rawls1999, 334). So, for instance, Rawls himself (alongside millions of other Americans) was conscripted into service in the Second World War, following the Japanese attack on Pearl Harbor. But even if this attack had never taken place, and the United States had faced no Axis threat, it would have still been morally permissible for the US government to draft its citizens into the fight against the Axis powers.

There are three main qualifications to this answer. First, conscripting locals into a just war abroad is impermissible wherever there is serious reason to think that sending arms and/or funds will be at least as conducive to the just war effort. Second, in many instances, the escalatory implications of committing troops abroad can be serious, and these implications are surely morally relevant (NATO’s enduring reluctance to commit troops to Ukraine being a prime case in point). Third, and perhaps most important, there is the priority of democratic majority will. As many democratic theorists have long noted, a democratic majority has the moral authority to make (morally and prudentially) mistaken decisions on many issues. Conscription is one of these issues. If a clear legislative and popular majority opposes conscription into a foreign defensive war, then the enactment of conscription is impermissible even if the war’s cause is eminently just. This constraint effectively means that conscription into a just defensive war abroad (as opposed to the encouragement of volunteers) may only happen if a democratic majority sufficiently identifies with the plight of that war’s foreign victims. But when such identification is present, the idea that it is morally permissible to conscript locals for the sake of a foreign fight starts to look less radical.

Admittedly, even when taken together, these qualifications may not fully assuage the concerns of those who fear that my account of external conscription is overly radical in the scope of conscription it licenses. However, the preceding arguments suggest that the fearful’s best response may not be to question the permissibility of external conscription. If external and internal conscription do indeed stand and fall together, the fearful might do best to reject both forms of conscription.