This chapter explores what I see as some problems with the All-Affected Principle (AAP), and with the proposals for redrawing political boundaries that have been made on its basis. I define the AAP as holding that everyone affected by a decision should have a right to participate (e.g. through voting, or some other form of direct influence) in a procedure governing that decision. My construal is neutral as to how votes should be weighted, whether equally or proportionally to the degree to which the decision affects individuals’ interests. My definition of the AAP is narrower than some, yet I believe this narrowness is necessary to capture the putative connection between the AAP and democracy. While a broader reading of the AAP might suggest that all those affected should have their interests considered, be represented by an advocate, or have a chance to plead their case, those broader principles are not obviously democratic. A benevolent monarch might well consider his subjects’ interests, allow them advocates, or grant them opportunities to make a case. Still, the monarch’s subjects would be deprived of all power in the monarch’s political decision making. I therefore opt for the narrower construal.

The first part of the chapter asks whether there is a convincing philosophical justification for the AAP. The second part asks whether it is best understood as a substantive or procedural principle. The third investigates whether the AAP provides a useful way to approach boundary questions. Unlike the AAP’s defenders, I argue that the principle does little to realize individual or collective self-determination. Whatever one thinks of the AAP, it is important to acknowledge that it has significant costs for self-determination as a moral ideal.

I should say at the outset that I agree with the AAP’s defenders that salient issues facing the world today cry out for global regulation: climate change and refugee crises are two obvious examples. But I worry that the AAP provides an overly blunt and simple approach to these problems. To see why, we need a better grasp of the principle.

Justifying the AAP

Let me start by asking: What does it mean to be “affected?” I propose that one is affected by a decision if it has a causal impact (or perhaps a sufficiently important causal impact) on one’s interests (or perhaps sufficiently important interests). Yet it is often counterintuitive to include everyone whose interests are affected in making a decision.¹ Consider Robert Nozick’s famous counterexample:

Proponents of the AAP have tried various strategies to deal with this “overinclusiveness” worry. Yet limiting relevant “affected interests” to self-determination and self-development (as does Warren, this volume) does not seem to avoid Nozick’s critique: surely one has a self-development interest in being able to marry one’s beloved. Nor does it help to adopt the proportionality interpretation of the AAP (favored by Gray and Warren, this volume).³ Since my marriage has a greater impact on my interests than those of others, perhaps we could give me the greatest say over this decision, and a lesser say to other interested parties, such as my suitors, parents, and friends. But is it really plausible that these others should have any say at all?⁴ To be sure, even if the decision is mine to make, I have a moral obligation to consider my suitors’ interests, and not to gratuitously harm them – e.g. I should express my decision in a way that is sensitive to their feelings. But I need not grant them any say in my decision making.

The lesson of Nozick’s example is that sometimes “ways of importantly affecting the lives of others are within the rights of the affecter.”⁵ Most of us believe that an individual’s personal autonomy rights grant them an important domain of choice regarding their own life, including – besides their right to decide whom to marry – rights of free expression, freedom of religion, and free choice of occupation. People are owed some range of options within which to make choices that realize their own personal self-determination, even when these choices affect others. Self-determination is not infinite, of course; it is limited by duties of justice. But within the limits of my self-determination rights, once those are properly specified, it is I, and not anyone else, who has a claim to decide.

In response to this concern, one might exclude personal autonomy rights from the domain of the AAP. When some individual has a personal autonomy right over a choice, on this view, the fact that outsiders’ interests might be affected does not entitle them to any say. Notice, however, that Nozick’s worry seems to extend beyond individuals to associations and organizations. For example, in 2016, Harvard University rejected 35,315 applicants. This decision certainly affected their important interests. Should these applicants have a say in Harvard’s admissions policies? It seems to me – by analogy to the personal autonomy case – that this is a choice Harvard alone has a right to make, in spite of its effects on outsiders’ interests. Since the number of rejected applicants is larger than the number of current Harvard University students and faculty, were they to be included in Harvard’s decisional processes, then, given their strong preferences to attend the university, they would likely have to be admitted. And were that to happen, Harvard’s educational purposes would become very difficult to attain.

Of course, like personal autonomy rights, the shape and scope of organizations’ decisional rights are limited by duties of justice. That is the reason why Harvard cannot deny admission to (among others) African American, Jewish, or female applicants on grounds of race, religion, or sex. All citizens have an important claim of justice to a fair opportunity for higher education, and excluding them on discriminatory grounds is not essential to the university’s attainment of its educational purposes. But though its decisional rights are limited by constraints of justice, within its rights, Harvard has a claim to determine which applicants to take, without granting them a say, even when the decision importantly affects their interests.

One might respond here by further limiting the AAP, holding that when either an individual or an organization has an autonomy right to make a decision, affected nonmembers lack any claim to participate in making that decision. It is worth noting that some adherents of the AAP reject this response. They argue that firms, universities, and churches ought to be democratically organized and to include all those whose interests they affect (see Gould, this volume).⁶ On this interpretation, the AAP becomes a radical principle. Most people do not think there is any nonnegotiable demand for democratic inclusion and decision making in all associational contexts.

I believe we should reject this radical interpretation. A just society should leave appropriate space for associations to have “a free and flourishing internal life.”⁷ Consider a limit case: should the nuclear family be democratically organized and include all those whose interests it affects? State pensioners and recipients of welfare benefits have an interest in my child’s future earnings. Must I grant them a say on whether they should take after-school piano lessons or learn computer programming? To require all affected outsiders to be included in the family’s decisions would undermine the goods the family makes possible for its members. This is not to say that outsiders’ interests are irrelevant. Yet the right way to respect these interests is to place justice-based limits on the family’s decisional power, drawing the boundaries of its autonomy rights in the appropriate way. Thus, the choice not to educate one’s child, or to bequeath them one’s large estate free from tax, is not one any family should have the right to make.

Suppose, then, that one accepts this further suggested limit to the AAP: both individuals and associations/organizations can have (limited) autonomy rights to make decisions without granting outsiders a say. The question then becomes: How much is left of the principle? Some authors postulate a category of “decisions in principle open to democratic adjudication,” exogenously defined, and hold that when it comes to those decisions, one has a claim to inclusion insofar as one’s interests are affected.⁸ But in light of our discussion, this response seems ad hoc. If a nonpolitical organization can have rights to make autonomous decisions when those decisions affect others’ interests, then why can’t a political organization have such rights as well? Suppose the Navajo Nation is considering whether to opt for English-only or Navajo language instruction in its elementary schools. Various nonmembers’ interests are at stake: the profits of textbook salesmen, the employment opportunities of English teachers. But I believe this decision is one for the Navajo Nation alone to make. If this is correct, then the work of determining who should be included in the demos is done not by the AAP, but rather by an independent account of individual and collective autonomy rights and the constraints that justice imposes on the shape and scope of those rights. A just institution should “build in” decisional permissions to facilitate individual and associational pursuits, “while defining their boundaries by general standards.”⁹ This is as true of a well-ordered global framework as it is of a domestic one.

One might object here that my argument merely shows that certain decisional permissions ought to be granted as a matter of substantive justice, if we are to secure important individual and associational interests at tolerable cost to other values. But who should have the authority to decide which permissions our basic structure should recognize? Perhaps everyone affected – not a bureau of technocrats, or some unelected judge – ought to decide. Someone could hold that the all-affected possess legitimate authority to determine the shape of the global basic structure, while simultaneously holding that they ought to recognize personal and collective autonomy rights. Some theorists have argued, for example, that a global democratic institution is required in order to legitimately demarcate the boundaries of the world’s constituent political units.¹⁰ On this view, as Abizadeh puts it, “the self-determination of differentiated democratic polities” should be viewed as “derivative of the self-determination of the ‘global demos’ as a whole”¹¹ (italics in original).

It is true that the AAP might plausibly be construed as offering a theory of legitimate authority rather than a theory of substantive justice. Still, a convincing theory of legitimate authority needs to be constrained by some “core” elements of substantive justice, including basic autonomy rights. Legitimate authority cannot take a purely procedural form: decisions at odds with “core” justice requirements can undermine the authority of a democratic procedure altogether.¹² These core requirements, in my view, include not only rights integral to the proper functioning of a democratic procedure itself – like freedom of political speech and association – but also other rights, like freedom of conscience, personal privacy, and the freedom to choose one’s occupation. Were a democracy to fail to recognize these autonomy rights, its citizens would have no reason to consider its verdicts binding.

Further, I believe the preconditions of democratic legitimacy extend to collective autonomy rights as well as personal ones. Were a higher-level demos (i.e. the US federal government, or a possible future global democracy of the all-affected) to decide that the Navajo Nation was not entitled to self-governance, forcibly merging its members into a wider polity against their will, I believe this decision would undercut the legitimacy of that higher-level demos. Members of the Navajo Nation would have no reason to see that verdict as binding on them. So even if the AAP is construed as a theory of legitimate authority, personal and collective autonomy rights may still constrain it.

Four Distinctions

The AAP is often invoked to support the view that we should aim for global democracy. The idea is that in the contemporary era the domain of affectedness has expanded, through global trade, investment, communication, and cross-border environmental impacts. These developments have displaced a prior equivalence between territorially based populations and the effects of political decisions that is often (in my view dubiously, given the history of global trade and colonialism) presumed to have obtained in earlier eras. To comply with the AAP, we must extend the reach of democracy over time and as circumstances allow, ultimately to the global level. Does global democratic enfranchisement indeed follow from the AAP?

It is helpful here to distinguish between two different construals of the AAP:

1. (1) Substantive Justification: when someone’s justice-related interests are affected by a decision, decision makers are required to take that interest into account (along with other affected interests), aiming for a fair balance of fulfillment of justice-related interests among all affected parties; and

2. (2) Procedural Justification: when someone’s justice-related interests are affected by a decision, that person should be provided with institutional influence (e.g. through a voting procedure) over that decision.

International interdependence clearly changes the scope of substantive justification, and this may give rise to justice constraints that should limit states’ decisions. I absolutely agree that when individuals and collectives exercise their autonomy rights, they are morally bound to consider the interests of those affected. Yet just about every moral view endorses (1), including those theories that limit themselves to negative duties not to violate others’ rights or to harm them in significant ways. Construed this way, the AAP does not say anything particularly novel. Most defenders of the AAP instead take it to imply (2): we have a duty to include the affected in an authoritative decision-making procedure. All affected interests should enjoy a say over the outcome.¹³

Why is procedural justification necessary? One thought is that a democracy of the all-affected is the institutional setup most likely to lead to a substantively justified outcome. (This is the traditional instrumentalist defense of democracy.) When individuals whose justice-related interests are affected by a decision are provided with mechanisms to represent and advocate for those interests, this is likely to promote a fairer balance of interest fulfillment. On this view, “making decisions democratically … is the best way to protect and promote people’s interests.”¹⁴

This is often a plausible view, but it is subject to important empirical preconditions. Democratic decision making only reliably leads to a substantively justified outcome when voters are able to gain good information about others’ interests, and when they are disposed to take those interests into account in voting. Suppose that factories in Peru emit pollutants blown downwind to Bolivia, affecting the well-being of people there.¹⁵ Many advocates of the AAP suggest that in situations like this, we ought to establish a democratic institution that can promulgate environmental policies binding both Peru and Bolivia, and that citizens of both countries should be given a vote in determining these policies. But is this the best way to promote a fair balance of interest fulfillment? It depends.

Ideally, the 30 million Peruvians would be informed about the impact of their downwind emissions on their 10 million Bolivian neighbors, and ideally they would care about these effects, voting for emissions limits even at some cost to their own prosperity. Suppose, however, that nearly all the Peruvians favor allowing their factories to pollute, at whatever cost to their neighbors, since the emitting industry is a “national” champion on which many jobs depend. In these circumstances, justification through a democratic procedure is unlikely to lead to a substantively justified outcome. A fairer balance of interest fulfillment might instead be secured by an institution (say, an international regulatory commission) that allowed for representation of affected interests and gave them a chance to contest or appeal its verdicts, but did not give them direct influence over the decision. Of course, it may well be possible to design a democratic mechanism that would do better than a simple majority vote. But the point is that whether democratic participation promotes a fair balance of substantive interest fulfillment in a given scenario is contingent. Where it does not, then from the point of view of substantive justification, we should prefer a procedure that does not grant the all-affected any say over the outcome.

Were we to institute a global democracy at present, I think the trade-off between substantive and procedural justification would be huge.¹⁶ In a domestic context, we rely on the media, personal, associational, and social ties, and shared educational institutions to gain some (imperfect) understanding of our compatriots’ interests. Of course, there is a global media and, increasingly, a global network of social and associational ties. But these networks are fragmented: they unify mainly elites, and they exclude large parts of the world entirely. Cultures, economic circumstances, and political values still vary much more widely across countries than within them. This means that it is hard for ordinary citizens of the Netherlands, say, to get a good grasp on the interests of people in the Central African Republic, and vice versa. Without denying that some subnetworks have established dense global ties, when we look at the world as a whole, diversity and lack of connectedness remain great, and our resources for bridging these gaps are not well developed. Similar worries could be raised about political motivation. Even where individuals have an understanding of distant others, they are likely to prioritize their own and their compatriots’ interests. So in the near term, we face contexts where, as Valentini puts it, “the preconditions for democracy to be successfully established at the global level are … missing.”¹⁷ Presently, a global democracy is not likely to lead to substantively justified outcomes.

Is that a reason not to aim for global democracy? One could argue that someday a reasonably just global democracy might become feasible, and it is worth putting into place the preconditions for that scenario by gradual steps.¹⁸ Perhaps global social movements should educate people around the world to develop solidarity with distant others and to establish better links across societies, which would unify masses as well as elites. I leave it open whether an instrumentally justified global democracy might become feasible in the future, and I do not rule out social movements that attempt to establish the preconditions for it. However, at any given moment, to make the instrumental case for global democracy, one must argue that it would produce more substantively justified outcomes than a nondemocratic institution would. In current circumstances, I doubt that this case can be made. I want to stress that I fully agree that global decision makers ought to be externally constrained to take into account the interests of those affected by transnational processes, e.g. not to impose significant harms on them. But I doubt that global democracy would be better at providing these constraints than would alternative, nondemocratic institutions, like international courts or regulatory bodies.

Are there other, noninstrumental reasons to work towards global democracy? Here we need some understanding of why democratic influence might be intrinsically important, apart from its effects on substantively justified outcomes. Of course, many people believe that even if a benevolent dictator or bureau of technocrats were to make substantively well-justified decisions in our behalf, there would still be an important objection to their rule. But what exactly is the nature of this objection, and does it apply to global decision making as strongly as it does in the domestic context?

Here there are two prominent arguments for democracy’s intrinsic value. The first – which I mention only to set aside – invokes public equality. It holds that to respect people as equals, it is not enough to merely to take their substantive interests into account. Instead, some institutional recognition of their equal status is required, and (at least in certain circumstances) this recognition should take the form of democratic enfranchisement. Disenfranchising people brands those excluded as inferior, fails to extend equal respect to their judgments amid disagreement, or fails to treat them as social equals.¹⁹ Whether the public equality argument would prescribe global democracy depends importantly on the nature of status inequalities at the global level, whether these inequalities are perceived as rendering some people publicly inferior to others, and finally whether granting all persons an equal say in a global democratic procedure is the best way of affirming their equal moral status. Though the public equality argument raises many interesting issues, I lack space to examine it here. For that reason, this chapter should not be seen as articulating an all-things-considered case against global democracy. I leave it open whether the demands of public equality might extend beyond the nation-state, in a way that demands the institution of global democratic procedures.

Instead, I take up a second argument for democracy’s intrinsic value: that granting people democratic influence over what affects them helps to realize a moral ideal of self-determination.²⁰ I scrutinize the self-determination argument here because the AAP’s defenders often invoke it (for examples, see Gray, Warren, and Gould, this volume), and also because I myself believe that self-determination has considerable value. But I believe democratic voting procedures are neither necessary nor sufficient to realize self-determination. So the self-determination argument cannot support a clear obligation to pursue, over time, equal political influence at the global level. Instead, I will suggest that once we better understand the ideal of self-determination, we will see that global democracy is in significant tension with it.

Self-Determination, Democracy, Boundaries

The self-determination argument for democracy begins from the idea that each person has a fundamental interest in being the author of their own life. This interest is commonly thought to ground individual liberties, like freedom of expression, freedom of religion, and the freedom to marry and form personal relationships. These liberties allow a person to express their evaluative judgments in their life-commitments. But government decisions also have an impact on the shape of a person’s life, profoundly affecting their natural and social environment in ways that pose a prima facie threat of domination. (Here I gloss over an important debate over what exactly it is about government decisions that threatens individuals’ autonomy. Is it the imposition of authoritative legal obligations? The state’s enforcement of directives, through threats or coercive sanctions? Or simply the causal impact that these decisions have on an individual’s ability to lead their life as they see fit? While this debate is important – and it is not clear to me that the AAP gets the best of it – for the purposes of this chapter, I shall assume that the AAP correctly describes the prima facie concern about domination.)

The self-determination argument then suggests that if a political subject is to be fully autonomous, the prima facie threat of domination from political decisions must be overcome. How is this to be accomplished? The answer is that, just like their life-commitments, the political decisions applied to an individual should somehow reflect their own judgments and choices. The final premise is that by granting people a right to participate in a democratic process, we enable them to be “authors” of the decisions under which they must live.²¹ In affording each citizen an opportunity to influence the state’s decisions, democracy neutralizes the threat that political rule poses to their autonomy, turning alien rule into self-rule. The argument can be extended to global democracy: since the decisions made by foreign states, corporations, and international organizations have consequential effects on people’s lives, to be fully autonomous, they must be given a right to participate in these decisions.

I agree with proponents of the self-determination argument that political decisions often pose a prima facie threat of domination. Yet while I agree that self-determination is an important value, I believe democratic voting procedures are neither necessary nor sufficient to secure it. Note here that I adopt a narrow definition of “democracy,” which refers to a procedure granting each person equal opportunity for influence: a procedure such as majority rule, or perhaps a lottery system.²² I argue that a shared commitment to collective political action is crucial to enabling self-determination: in the absence of shared commitment, democratic procedures have little significance; in its presence, they are not required.

Let me first explain why voting procedures are insufficient for self-determination. Recall the worry that individuals are not fully autonomous when they are substantially affected by decisions beyond their control. Yet it is not clear how global democratic procedures will solve this worry, since these procedures also would not afford individuals meaningful control over life-affecting decisions. Since 7 billion other citizens share voting power in a global democracy, whether or not decisions reflect my priorities and judgments will always be partly up to them. Indeed, they may impose their alien views on me, as happens whenever I am in the dissenting minority on some serious political question. Here, it seems, we are ruled by the global majority, “subject to the collective will of multiple others”; we do not rule ourselves.²³

Thus, it seems impossible for global democracy to extend an autonomous individual control over decisions that affect them. Global citizens have only a minuscule, utterly negligible political influence, and this influence will not afford them the power to ensure that political decisions reflect their own judgments and choices. Further, in any global democracy, large numbers of people will still be outvoted, as the worry about tyranny of the majority illustrates, and these individuals will certainly see central features of their lives shaped by decisions they reject, including the possible overriding of local interests they regard as important. Given all this, in what sense does global democracy enable autonomous self-rule?

In general, individual autonomy is furthered by personal control over decisions, not by an infinitesimal share of collective control. Suppose you are out to dinner with a large group of friends. Are you more autonomous if the group takes a majority vote to decide what you will order, or if you get to choose your meal for yourself? It seems that you are autonomous only in the second case.²⁴ Only then can you ensure that the meal you will eat will robustly reflect your own judgments about what meal would be best.

Given this line of thought, one might be tempted to conclude that self-determination is simply an illusory political ideal: political decisions cannot be authored by the entire group of people subject to them. I do not accept this conclusion. Instead, I believe self-determination is a realistic and valuable ideal. But I do think these reflections show that democratic voting plays little role in safeguarding self-determination.

In response, I propose to conceive self-determination somewhat differently. I hold that an individual is self-determining when they are governed on the basis of values and priorities that they in some way share. Of course, no individual’s personal priorities can be mirrored in each and every political decision, but there is a second-order sense in which an individual’s priorities are often reflected in group decisions – namely, when they share a commitment to a valued political enterprise and to certain shared policies by which they believe that enterprise should be structured. A commitment to participating in collective political action, on this view, is very important in enabling self-determination.

Consider a small-scale example, say, a partnership running a joint venture together. Often, in the context of such cooperative activities, a group develops shared commitments about how their enterprise should go. This does not mean that the participants converge in all their first-order judgments. Still, even when they do not converge, partners are often able to generate shared commitments – not reducible to their own judgments – about how their joint venture should be run. These commitments frequently emerge as compromises in the face of disagreement or bargaining about how the group should structure their enterprise. For example, a group might form a shared commitment to certain methods of making decisions, such as majority voting after public discussion, or to certain higher-order ideals or values.²⁵

Participants can freely accept their group’s commitments, and the outcomes that result, even when these outcomes diverge from their first-order judgments. Here, the participant is perfectly autonomous when they comply with the group’s decisions. To take an example: while I often disagree with my colleagues about whom to hire, I prefer that we make our own hiring decisions together, according to our accepted consultation procedures, even though that means accepting some decisions with which I disagree. Indeed, I would consider myself disrespected if the dean overruled our collective decision, even when the result was to impose my preferred candidate. Though my colleagues’ decisions do not always correspond to my first-order preferences, there is still an important, second-order sense in which my priorities are reflected in these decisions. I share a commitment to a valued cooperative enterprise together with these colleagues, and to certain shared policies by which I believe our enterprise should be governed. If this is correct, then my interest in self-determination can be fulfilled even when I am subjected to decisions that I do not personally control and with which I may disagree, so long as I value my participation in the collective venture and endorse the group’s higher-order values and procedures. Though I may not endorse every outcome, I am governed by institutions that I view as appropriate.

When a group of citizens share commitments of this kind, I will say that they share a political will. A shared political will is an interlocking structure of joint intentions among individuals to cooperate together in a political enterprise and to endorse higher-order policies as to how their enterprise should run. Collective self-determination, as I understand it, requires a (robust) correspondence between citizens’ shared political will and their institutions, under conditions that enable their free deliberative reasoning.

On my view, it is a shared political will – not the existence of democratic voting procedures – that is essential for self-determination. Note that voting procedures say nothing about the composition of the demos ruled by those procedures, including whether it contains subordinated minorities. But the ideal of collective self-determination is not indifferent to the demos’s composition. Consider the following case:

I suggest that this annexation would not be legitimate. Majority voting procedures are not naturally authoritative independent of considerations about the constitution of the group ruled by those procedures, including whether that group contains unwillingly subjected minorities.²⁶ Self-determination is not a purely procedural ideal, it is a normative theory of the conditions under which political power is consistent with the autonomy of those subject to it. These conditions are substantive, not just procedural; they include certain basic rights, as well as the institution of appropriate boundaries between political groups.

More controversially, I also hold that democratic voting procedures are not necessary for collective self-determination. Recall that I am adopting a narrow definition of “democracy” as a procedure that grants each person equal opportunity for influence. A broader definition of “democracy” refers to any system that ties legitimate political power to a process of shared deliberation among free and equal citizens.²⁷ There is significant overlap between my account of self-determination and this broader democratic ideal. Since correspondence between a group’s shared will and their institutions is valuable only where citizens’ judgments are freely formed, my view stresses the protection of basic liberties of conscience, free expression, and free association, and the importance of deliberative public opinion.

While my account might be said to rest on broadly democratic values, I doubt that democracy in the narrow, procedural sense is necessary for self-determination. A particular group’s policy for self-governance might involve equal votes and majority rule, but it also might not. Participants may agree, under authentic deliberative conditions, that a particular individual – say, their monarch – is especially wise, virtuous, and good at interpreting the group’s shared commitments, and defer to their judgments day-to-day. So long as participants share commitments about how their joint venture should go, and so long as the monarch’s decisions count as reasonable elaborations of those commitments, then participants will be as self-determining under their monarch as they would be in a democracy. If this seems far-fetched, consider that many nonpolitical associations do exactly this. Many churches, schools, and businesses are governed by nondemocratic norms that their participants genuinely accept, and which protect their fundamental interests.

One might object here that to be self-determining, it is not enough for political institutions to reflect the authentic shared will of the population. Instead, that group must have the ability to control their institutions to serve their commitments, should they change their minds as to how they wish to be ruled. Voting procedures, it might be argued, are necessary to ensure this. I agree that there must be some way for the people to revoke authorization of their regime. But a range of different mechanisms might serve this purpose. Lockean rebellion seems antiquated and unreliable, but it might be sufficient in a society where the power of leaders depends heavily on the cooperation of the people and technologies of social coercion are undeveloped. Even under contemporary conditions, in states that possess armies and secret police, I believe it is possible for citizens to authorize a government that is not democratic – say, a constitutional monarchy – so long as there is some mechanism by which to initiate a process of constitutional reform, e.g. an amendment procedure.²⁸ If this is so, elections and voting are not necessary to self-determination.

One might also wonder whether my argument implies that democratic voting is not a requirement of justice. I think the public equality argument explains why justice often demands democratic procedures: formally equal votes reinforce citizens’ equal social status. While I hold that voting procedures are insufficient to ensure self-rule, I agree that they enable a valuable recognition of citizens’ equal status, and this is an important reason for preferring them (see also Gray, Gould, and Warren, this volume). Still, a nondemocratic state can be legitimate – with a right to rule its population free from interference – even when its institutions are not fully just (say, because they fail to fully enable equal moral recognition). On my view, a state is legitimate if it protects its citizens’ basic rights and provides for their self-determination, even if this occurs through nondemocratic institutions. Such a state has a right not to be forced to become a democracy.

Thus, the ideal of self-determination, as I understand it, holds that it is important that political subjects see their demos as a cooperative enterprise that they value, and that they generally endorse its institutions, even when they disagree with particular decisions. This facilitates their political freedom; it gives them the ability to appropriately see themselves as governed in a way that reflects their own values and priorities, rather than being subjected, against their will, to hostile or dominating powers. So unlike the AAP, the ideal of collective self-determination suggests that political boundaries should be drawn so as to enable people to affirm their inclusion in a particular demos and to endorse the institutions that structure it.

It is natural to object here that however desirable shared political commitment may be, it is unachievable among territorially defined populations; there are always dissenters among us. While there is much to be said about this issue, I do not believe that self-determination requires that collective commitment be unanimous. Sometimes it is permissible to coerce dissenters who fail to share the commitments underwriting political institutions. One such case is where dissenters are alienated only because they hold gravely unjust political values. I believe these dissenters lack a claim to self-determination, because self-determination is conditioned upon a commitment to the principle of equal autonomy from which its value is derived. A second case is where there is no feasible institutional configuration that could accommodate dissenters’ priorities and still carry out morally mandatory state functions at reasonable cost. Here I appeal to the importantly territorial nature of our duties of justice; we cannot establish property and contractual rights, enforce those rights, and punish violators unless people who live in proximity and interact regularly are subject to common institutions. Lone dissenters therefore have a duty to accommodate to some feasible political scheme, and to compromise with their neighbors in order to do so. To claim self-determination, then, a group must be territorially organized and possess broadly representative practices.

Still, I believe the ideal of self-determination will often support (re)drawing political boundaries to allow groups to be governed by institutions that reflect their shared commitments. Here I have in mind groups (a) who have political commitments that are consistent with basic justice for others, (b) who possess or can create a territorially organized structure of representation, and (c) whose dissent can be feasibly addressed, at reasonable cost, by granting them separate institutions. Thus, self-determination favors decolonization over a wider metropole in the case of former subject peoples. It also grounds a preference for independent political institutions for Indigenous groups incorporated into settler states, and greater autonomy for persistently alienated minorities in e.g. Scotland, Catalonia, or Quebec.

The ideal of self-determination need not be predicated on the existence of prepolitical “peoples” marked out by characteristics – like language, shared culture, or common interests – that specially suit them for self-rule. I doubt that there is any such Archimedean point of view from which to delineate political peoples. Instead, on my view, a people is born only when its members engage in institutionalized political cooperation, and come to value that cooperation. Some structure of institutional representation is necessary to create groups with sufficient corporate agency to act as peoples. But this “people-forging” process can succeed or fail. Sometimes political institutions generate “uptake” among their constituents, who come to value their joint enterprise and to willingly support the institutions that govern it. But in other cases such “uptake” fails to occur, and subgroups remain persistently alienated, either because of a historical legacy of conflict or oppression, or because they share many distinctive political priorities, which go unrecognized by the majority. In this case, the ideal of self-determination may call on us to reconfigure political boundaries, to enable persistently alienated groups to be governed by institutions that better reflect their shared commitments.

Conclusion

Thus, the ideal of self-determination has implications for boundary drawing that are rather different from those suggested by defenders of the AAP; it suggests the continuing possibility of fission, rather than pressure toward global fusion. Because individuals have an important interest in seeing themselves as willing coauthors of the institutions that govern their lives, we ought to draw political boundaries, to the extent possible, to enable people to participate in cooperative enterprises that they identify with. In defining the people, we look to the patterns of affirmation and alienation that emerge as artifacts of our currently existing institutions. We then ask: Are there feasible institutional alternatives – consistent with maintaining a stable, minimally just structure of political authority – that would better realize self-determination for those who currently lack it? We delineate a new “people” – when we do – not because we are recognizing something that already independently exists, but because we have some reason to hope that a new institutional configuration will lessen alienation at reasonable cost. On this approach, the process of constituting the people is never finished, once and for all. The “people” is a mutable entity, and negotiating and renegotiating political boundaries is a process that we can expect to be ongoing.

I should stress here that self-determination is not an absolute right, but rather a moral claim that must be applied with due regard for circumstances, and it can be outweighed by competing concerns. In concrete cases, we will need to weigh the grievances of the persistently alienated against the countervailing risks to just institutions, including the potential for civil unrest, instability, ethnic conflict, or rights violations. I also do not hold that self-determination necessarily implies a right to a sovereign state; federalism, devolution, or internal autonomy may be appropriate vehicles for self-determination in many circumstances.

Yet if it is to avoid engendering pernicious alienation, a just global framework needs to make space for collective self-determination. We should extend groups the permission to form separate institutions and to order their affairs in accordance with their shared priorities, because this facilitates a valuable form of political freedom. As with the autonomy rights of individuals and associations, there are justice-based limits to self-determination: a demos is not entitled to make decisions that significantly harm others or threaten their essential rights. Transnational decisions that have these effects must be regulated. Yet global democracy is not the only – or the best – way to address these problems; international courts or oversight bodies – perhaps requiring proportionate representation from non-Western states – are another option. Self-determining peoples can also be required to form global institutions where their cooperation is essential to sustaining basic justice, e.g. by combatting climate change and ensuring that refugees receive a new home and protection for their human rights. These and other justice-duties will place limits on the demos’s autonomous decision-making power.

Still, securing basic justice for others is the only reason why a self-determining people ought to be forced to submit to the rule of global institutions. Since it is especially valuable for people to be ruled in a way that reflects their own priorities, global justice should “build in” permissions for collective self-determination, allowing groups to establish separate jurisdictions and granting them autonomy rights to make their own decisions, even where those decisions affect others’ interests (so long as they thereby violate no duties of justice). If this is correct, then it is not clear that the moral ideal of self-determination supports an obligation to pursue global democracy.