The Boundary Problem in Democratic Theory: A Methodological Approach

Abstract

How should political power and influence be allocated in democratic systems? That is, roughly, the core of the boundary problem in democratic theory. As of late, some authors have begun paying increased attention to the methodological aspects of this dispute. This paper attempts to make a twofold contribution to this ‘methodological turn’. On the one hand, it identifies and analyzes five desiderata of a successful principle of democratic inclusion. Any such principle, I argue, must be grounded in a clearly identifiable set of values, must provide a clear standard of inclusion, must specify the relevant modalities of inclusion, must provide sound bridging principles, and must be extensionally adequate. In the first half of the paper, these desiderata are developed and critically scrutinized. On the other hand, I also extract three lessons from the previous discussion. First, that a unique answer to the boundary problem (i.e. one grounded in one single principle) is unlikely. Second, that a satisfactory answer to the boundary problem need not necessarily appeal, as some have argued, to distinctively democratic values. And third, that intuitions about political exclusion (i.e. those requiring or permitting the exclusion of some individuals) should be treated with special caution.

Introduction

How should political power and influence (especially but not solely the right to vote) be allocated in democratic societies? That is the core of the so-called boundary problem in democratic theory. In the 1970s, Dahl (1970, p. 60) complained that this was ‘a problem almost totally neglected by all the greatest political philosophers who write about democracy’. Several decades later, this description is no longer accurate. In the preceding decades, political philosophers and theorists have become increasingly interested in this question, and a substantial body of literature has emerged around it. This question is of considerable normative significance: after all, if we are wrongly excluding some subjects from the demos, the resulting decisions will be less legitimate than they might have otherwise been—or, if the exclusion is sufficiently severe, not legitimate at all. And since all democratic societies make some exclusions (e.g. resident aliens, minors), we should better be able to tell which of them can be justified and which cannot.

Different answers have been offered. Some authors believe, for example, that those whose interests are significantly affected by political decisions are entitled to have a say in the corresponding decision-making processes (the All Affected Principle. Cf. Arrhenius 2018; Goodin 2007; Valentini 2014). Others, in turn, argue that simply having one’s interests (even normatively important ones) affected is not enough; what really matters—they claim—is subjection, to the law, or to the coercive institutions of the state (the All Subjected Principle. López-Guerra 2005; Abizadeh 2008; Beckman 2008, 2009, 2014; Miller 2009; Erman 2014). Still, others have held that we should focus on the kind of relationship that holds among the members of a political community: if they are members of the same society, or nation, or if they are to relate to each other as equals, then they should have a say—or so these arguments suggest—in the making of the decisions that will shape the direction and the nature of their community (cf. Carens 2013; Bengtson 2022).

It is not precisely clear, however, how this issue should be adjudicated. Many authors have proceeded by assessing the intuitive plausibility of the implications yielded by the different views, and an important part of the debate on the boundary problem has revolved around the alleged counterintuitive results of the many views on offer. Perhaps the most discussed example involves the potentially overinclusive implications of the All Affected Principle. On some formulations (namely, those that target all possibly affected interests), this principle appears to demand, as Goodin (2007, p. 68) famously put it, ‘giving virtually everyone everywhere a vote on virtually everything decided everywhere’. Many authors, however, regard this implication as highly implausible (Beckman 2009, pp. 351–352; Miller 2009,  p. 215; Song 2012, p. 49; Andrić 2017, pp. 267–269).

But judgments of intuitive plausibility, although probably unavoidable, have their own limitations. First, without a background theory about which kinds of intuitions are more or less reliable we cannot know how much weight we should give to our intuitions about political inclusion. Second, intuitive adequacy does not, by itself, tells us how to choose between theories when all of them yield at least some counterintuitive results. And third, and quite generally, intuitive adequacy is unlikely to be the only factor guiding theory-choice.

Some recent contributions have begun to pay increased attention to the methodological side of the dispute (Erman 2014, 2021; Donahue and Ochoa Espejo 2016; Maltais et al. 2019; Miller 2020; Lippert-Rasmussen and Bengtson 2021; Bengtson 2022).¹ Building on these, this paper sketches a methodological approach to the boundary problem. More specifically, this paper systematically formulates and evaluates the desiderata of a successful principle of political inclusion. Any successful principle of inclusion, I shall argue, must meet at least five requirements. First, it must be grounded in an identifiable (set of) value(s). Second, it must provide a clear standard of inclusion (i.e. that which must be met in order for an individual to be politically included). Third, it must specify the relevant modalities of inclusion (i.e. those through which political inclusion can be realized). Fourth, it must provide sound justificatory principles bridging its underlying values with its standards and modalities of inclusion. And finally, an adequate principle of inclusion must be extensionally adequate (i.e. in the absence of undermining defeaters, principles that conform with our considered judgments are preferable).²

Although I will not attempt to defend any particular answer to the boundary problem in this paper, I will nonetheless extract three lessons that I think can be motivated by the above analysis. First, I will suggest that because of its many facets and internal complexity, a unique answer to the boundary problem is unlikely—and, conversely, some form of pluralism seems more attractive. Second, I shall argue that a satisfactory answer to the boundary problem need not make reference exclusively to distinctively democratic values, as some authors have argued, and can instead draw from a broader palette of values and principles. Third, I will argue that intuitions about political exclusion (i.e. those requiring or permitting the exclusion of some individuals) should be treated with special caution, given the way in which such intuitions have been hugely mistaken in the past. In developing these three claims, this paper attempts to illustrate how by adopting a methodological approach we can become better equiped to critically scrutinize the tools theorists typically employ in dealing with the boundary problem. And this, even if it cannot replace first-order normative and conceptual theorizing, helps us make some progress.

The Desiderata of a Successful Principle of Inclusion

In this section, I formulate and articulate five crucial desiderata of a sound, or valid, principle of inclusion. By principles of inclusion, I refer to normative principles that require the political inclusion (in a sense to be specified) of individuals who possess a certain property, or stand in a particular relationship. A valid principle of inclusion, in the context of the boundary problem, is one that meets, at the very least, the following requirements.

Values, Standards, and Modalities

Principles of inclusion tell us that those who possess certain properties, or stand in a particular relationship, ought to be politically included. In order to fill in the gaps of this general scheme, a specific principle of inclusion must clearly specify (1) its standards of inclusion (i.e. which are the properties, or the relationships, in virtue of which someone should be included politically), and (2) its modalities of inclusion (i.e. which are the ways in which political inclusion can be realized). In addition, a principle of inclusion must also specify (3) its grounding values (i.e. those on which the principle ultimately rests; cf. Magaña 2022a).

The need to specify a principle’s standards seems straightforward enough (otherwise, the principle would be empty and untestable), but someone might wonder why we should also specify its modalities and grounding values.

On the one hand, specifying the modalities of inclusion matters because political inclusion can come in different ways. Although the right to vote constitutes the paradigmatic modality of political inclusion, it is not the only one: non-electoral representative devices—e.g. those that might be employed to represent the interests of children, future generations, or nonhuman animals (e.g. an ombudsman)—provide one such example (cf. Eckersley 2000; Ekeli 2009; Vink 2020; Magaña 2022b). Importantly, the literature on the boundary problem reflects the diversity of available modalities of inclusion. The All Affected Principle, for instance, has been used to allocate both rights to participation and representation (Goodin 2003, Ch. 11; Ekeli 2009, p. 445; Brighouse and Fleurbaey 2010,  p. 150). The All Subjected Principle, in turn, has been employed solely (to the best of my knowledge, at least) to distribute participatory entitlements.

If we do not specify which modalities of inclusion are relevant to a specific principle (1) it becomes harder to assess the content of the principle (as its requirements will remain compatible with many different possible interpretations), and (2) it also becomes harder to assess whether arguments that appeal to principles of inclusion are successful or not. Suppose, for instance, that one appeals to the All Affected Principle to argue that workers ought to be included in corporate boards. Suppose, further, that the All Affected Principle vindicates giving workers representation, or even some degree of voice, but not voting rights. Has the argument succeeded or not? Well, that will depend on which exact modalities of inclusion we wanted the All Affected Principle to vindicate in the first place. Thus, in order to assess the success of arguments by appealing to principles of inclusion, the assumed modalities of inclusion must be made explicit.

On the other hand, specifying a principle’s grounding values is pertinent because none of the statements expressed by principles of inclusion (e.g. that the affected, or the subjected, should be politically included) describe brute facts. There must be some reason why being affected, or subjected, or establishing certain kinds of relationships, should have any normative purchase to begin with.

This means, then, that principles of inclusion must be grounded in more fundamental values. The All Affected Principle, for instance, has been said to be grounded in the value of well-being (Goodin 2007, p. 50), fairness (Dahl 1989, p. 108), and individual and/or collective autonomy (Nässtrom 2011, p. 123; Lindley Wilson 2022). Attempts to justify the All Subjected Principle, in turn, often appeal to the worth of individual or collective autonomy (Abizadeh 2008, pp. 39–42; Miller 2009). And enfranchising those who stand in certain social relationships so that they can relate as social equals hinges on the value of social/relational equality (Bengtson 2022, pp. 14–18).

Any valid principle of inclusion must be anchored in some plausible value(s). Some authors have argued, though, that not any value will do: normative theorizing about the boundary problem must appeal to, or be constrained by, distinctively democratic values and ideals, such as political equality (Song 2012; Erman 2014; Miller 2020), solidarity (Song 2012; Miller 2020), or some kind of robust political agency (Saunders 2012; Erman 2014). In “We needn’t Necessarily Appeal to Distinctively Democratic Values”, I shall argue against this constraint.

A further reason why an adequate principle of inclusion should clearly state its grounding values is that different values will make a difference to our understanding of the principles at hand. To begin with, different principles will support different conceptions of which are the interests, states of subjection, or relationships relevant to democratic inclusion. For example, the interests that are relevant to the All Affected Principle will be different if we think that the principle is grounded in the value of autonomy, or in the value of well-being more generally—in which case the set of relevant interests will arguably be broader, for the principle would then cover both autonomy and non-autonomy-based interests. As a result, different values will ultimately provide different answers to the boundary problem, even for the same principle.

Justificatory Principles

Justificatory principles are intermediate-level normative principles³ linking a principle’s fundamental values with its standards and modalities.⁴ They are necessary because a principle’s standards, modalities, and values, are not, by themselves, enough to secure the validity of a principle. Some principle, for instance, might be grounded in values that few would doubt are normatively important (such as well-being, fairness, or autonomy), and still fail to succeed if there is no sound way to move from these values to the desired standards or modalities of the principle. One common criticism against the All Affected Principle, for instance, holds that the principle fails to show why a moral principle requiring that everyone affected by a decision should receive adequate moral consideration—a moral principle which few would deny—requires giving everyone affected by a decision participatory rights—a stronger thesis that some authors reject (Beckman 2009,  p. 47; Abizadeh 2012, p. 878; López-Guerra 2014, p. 46; Bengtson 2020, p. 97). Regardless of whether the objection succeeds or not, it illustrates how it is an open question whether a principle’s alleged grounding values can account for why those values should entitle someone to political inclusion, through some specific modality.

Specifying its standards, modalities, and values, then, is necessary to articulate a well-formulated principle of inclusion, but it does not suffice to have a valid or successful principle of inclusion. This requires, in addition, sound justificatory principles.

Justificatory principles can be instrumental and noninstrumental. Instrumental principles, on the one hand, claim that inclusion through some specific modality is required to produce, in the long run, outcomes that are substantively correct from a procedure-independent standpoint, and which we have procedure-independent normative reasons to pursue. One influential defense of the All Affected Principle, for instance, rests on the empirical claim that enfranchising those affected by political decisions provides the best way to protect their interests in the long run—which assumes, in turn, that we have compelling reasons to pursue such an outcome (cf. Goodin 2007, p. 50). Noninstrumental principles, on the other hand, claim that inclusion is essential for the realization of some value that we have normative reasons to realize. Thus, for instance, some have argued that enfranchising those affected by political decisions is partly constitutive of the ideal of collective self-rule (cf. Nässtrom 2011, p. 123), or that enfranchising those who stand in certain relevant social relationships is partly constitutive of the ideal of relational equality (Bengtson 2022).

As we saw earlier, some authors have defended the view that the grounding principles (upon which justificatory principles ultimately rest) must be distinctively democratic—e.g.  include values such as solidarity, political equality, or the worth of political agency. In addition, others have argued that the justificatory principles themselves should reflect certain features of the democratic ideal. Bengtson (2022, p. 2), for instance, has recently argued that justificatory principles must satisfy a Value Requirement, according to which justificatory principles (1) must be able to vindicate democratic inclusion (and not, for instance, mere consideration of interests), and (2) must also acknowledge that democracy is valuable and called for (i.e. it cannot rest on a value which might be equally, or even better, satisfied by excluding those that meet the principle’s standards).⁵ To be precise, Bengtson does not claim that principles of inclusion must necessarily satisfy the Value Requirement. More weakly, he contends that this is a requirement a principle of inclusion should satisfy in order to count as a fundamental principle of democratic inclusion—rather than, say, a derivative principle of democratic inclusion (what Bengtson, echoing G. A. Cohen, calls ‘a rule of regulation’).⁶ This leaves open the possibility, to be defended in “We needn’t Necessarily Appeal to Distinctively Democratic Values”, that not all principles of inclusion must be fundamental principles of democratic inclusion. Thus, rather than providing a general desideratum for a successful principle of inclusion, as we are trying to do here, the Value Requirement provides, more narrowly, a desideratum for a principle of inclusion to be appropriately classifiable as a fundamental principle of democratic inclusion.

Summing up, a successful principle of inclusion must involve sound justificatory principles bridging its underlying values and modalities of inclusion. I am, however, deliberately phrasing this desideratum broadly enough to allow, on the one hand, for the possibility of both instrumental and noninstrumental justificatory principles, and, on the other hand, to explicitly acknowledge as an open question whether principles of inclusion should appeal to democratic values (necessarily, or more weakly, as in Bengtson’s view, in order to qualify as a fundamental principle of democratic inclusion).

Extensional Adequacy

Finally, an adequate principle of inclusion should not depart too much from our considered judgments about political inclusion. Considered judgments are those whose reliability is not undermined by local debunking explanations.⁷ In John Rawls’s ([1971] 1999, p. 42) terms, they are acquired ‘in circumstances where the more common excuses and explanations for making a mistake do not obtain’. When someone expresses a considered judgment, we must assume that she engaged in a respectable amount of reasoning before endorsing it, or that she made a bona fide effort to discharge her epistemic duties, or at least that she acquired the relevant judgment through (generally) epistemically reliable means.

Conformity with considered judgments can be relevant for at least three dimensions of a principle of inclusion. First, for the plausibility of its grounding values. We do not need, for instance, to know anything about its standards, modalities, or justificatory principles to know that a principle of inclusion that relies on the value of, say, redheadedness is mistaken. Second, for the plausibility of the justificatory principles linking the grounding values, standards, and modalities of a principle of inclusion. Noninstrumental accounts (e.g. those that argue, for instance, that political inclusion is partly constitutive of autonomy or relational equality) can be more or less intuitively appealing in a quite straightforward way. Instrumental accounts, though, are a bit more complex. On the one hand, they rely upon empirical assumptions which our intuitive normative judgments are ill-equipped to adjudicate, and which must be assessed by appealing to the best available empirical evidence. On the other hand, though, even instrumental accounts contain normative and/or evaluative assumptions that might conform better or worse with our considered judgments. Someone might possibly contend, for instance, that, even if granting some group of individuals participatory rights over the decisions that affect them would better protect their interests in the long run, a purely outcome-based account is in itself counterintuitive. Finally, considered judgments can be used to assess the plausibility of the implications of the principles themselves. As we have seen, some believe that if the All Affected Principle entails the inclusion of ‘virtually everyone everywhere’ (as, on some formulations, it seems to require), then it is very implausible, or perhaps even plainly absurd. Likewise, on some readings, both the All Affected Principle and the All Subjected Principle appear to require the inclusion of transients and tourists—who are subjected to the laws of the jurisdictions within which they move, and who have important interests at stake by the political decisions made in those places. Many authors, however, find these implications unpalatable.⁸ Plausibly, conformity with intuitions in all these three senses matters. For the purposes of this paper, though, I shall focus specifically on the third class of intuitions: that is, intuitions requiring or permitting the inclusion/exclusion, of some group of individuals who share certain property or stand in some particular relationship. I shall call these, for the sake of brevity, intuitions about political inclusion (or exclusion).

Although there is an ongoing debate on the evidential credentials of intuitive judgments, I shall here assume that some degree of conformity with considered judgments is indeed important for theory choice in the context of the boundary problem: ceteribus paribus, if a principle of political inclusion conflicts with our considered judgments about political inclusion, this counts against the validity of the principle. Such a requirement has been explicitly postulated by others: Andrić (2021), for instance, speaks of a requirement of ‘Extensional Adequacy’, whereas Bengtson (2022) introduces a ‘Considered Judgment Requirement’.

Little has been said, however, about which intuitions should be counted as considered judgments. This, however, is regrettable, for, as I noted in the Introduction, without a background theory about which kinds of intuitions are more or less reliable it is difficult to know how much weight we should give to our intuitions—or, in other words, which intuitions we can treat as considered judgments.

In a recent discussion, Handby (2022) has usefully identified three types of biases to which intuitions in political philosophy might be especially vulnerable: (1) parochialism (when our intuitions track normatively irrelevant, but pervasive, facts about our own political units), (2) endogeneity (when our judgments have been shaped by the very institutions—or in general, by the political phenomena—that they are about), and (3) idiosyncrasy (when our judgments have been shaped by contingent facts about our own life history or individual psychologies). As I shall suggest later, some intuitions about political inclusion/exclusion might be vulnerable to these (and other) biases. If that is correct, theorizing on the boundary problem should pay greater attention to whether normative judgments have been formed under epistemically propitious circumstances or not. Although I cannot offer any general account in response, my main goal here (both in this section and in “Intuitions About Political Exclusion are Vulnerable to Biasing Factors”) is to highlight the importance of discussing which factors render normative judgments (un)considerate—as it is both a highly consequential, but somewhat neglected, aspect of the discussion.

Three Lessons

In this section I will articulate three theses that I think can be motivated by the above discussion.

Strict Monism About Political Inclusion is Unattractive

When discussing political inclusion, as in many other debates, one can adopt a monist or a pluralist approach.

Many principles of inclusion in the context of the boundary problem appear to assume what we may call an ‘if and only’ clause, according to which all those and only those who satisfy the principle’s standards ought to be politically included (cf. Beckman 2014, p. 253; López-Guerra 2005; Beckman and Hultin Rosenberg 2022, p. 191). On this reading, the elements of a successful principle of inclusion must not only be able to explain why those who satisfy the principle’s standards should be included (under some specific modality of inclusion), but they must also be able to account for, and support, the exclusion of those who do not meet them.

The ‘if and only if’ clause can be read in at least two ways. On a strong reading, one might hold that there is only one valid principle of inclusion for all modalities of political inclusion. This is a purely monistic view, in which questions of political inclusion are adjudicated by one single principle of inclusion. On a weaker reading, one might claim that there is only one valid principle of inclusion for each particular modality of inclusion. This is, in contrast, a moderately monistic (or, conversely, moderately pluralistic) view.

Some authors working on the boundary problem have defended moderately monistic/pluralistic views. Bauböck (2018), for instance, has argued that the All Affected Principle determines who ought to have their interests represented in the decision-making process, the All Subjected Principle identifies who was a right to contest governmental decisions, and to receive equal protection of their rights and freedom, and the Stakeholder Principle⁹ identifies who ought to be considered a citizen of a polity. Similarly, Erman (2021) has recently argued that both the All Affected Principle and the All Subjected Principle capture values that are partly constitutive of the democratic ideal (such as procedural fairness and political autonomy).

On the other hand, one could also defend a strongly pluralistic approach, according to which political inclusion under one specific modality might be supported by different valid principles of inclusion. On this account, then, different principles can be adequate, but they are not individually tied to one particular modality of inclusion.¹⁰

One reason to doubt purely monistic approaches is that explaining why it would be objectionable for some group (i.e. those who meet a principle’s standards) to be excluded from decision-making does not, by itself, show that there is anything objectionable at all with the inclusion of other groups meeting different standards. Saying, for instance, that individual autonomy requires that individuals be able to have a say in the decisions that subject them entails, if the argument is sound, that excluding those individuals from political decisions is objectionable. But it does not imply (not at least without any additional arguments) that individuals are wronged when others are included by other reasons.

The above methodological discussion suggests a further reason for skepticism. As we have seen, the ‘boundary problem’ can be disaggregated into at least two questions, one about the allocation of participatory rights, and another about the adjudication of rights to political representation. Other authors have introduced further distinctions. Lippert-Rasmussen and Bengtson (2021), for example, distinguish between at least three instances of the boundary problem: (1) the procedural problem (should democratic norms be themselves decided democratically?), (2) the conceptual problem (should the concept of democracy shape the constitution of the demos?), and (3) the value-focused problem (should the values that make democracy desirable bear on how the demos is constituted?). In a similar spirit, Erman (2021) has argued that the boundary problem must be sensitive to the different functions we might expect democracy to perform, which are in turn set by the value(s) that make democracy valuable, legitimate, or authoritative as a decision-making procedure.

All of these contributions to the discussion suggest that the ‘boundary problem’ constitutes an umbrella category which encompasses distinct, albeit related, questions, all unified by a shared bearing on the general question of how political power, influence, and relevance should be distributed in democratic societies. Now, the more multifaceted and decomposable a problem gets, the less likely it is that it can be adequately answered by one single principle. Suppose that you believe that considerations of autonomy vindicate that all and only all those subjected to state coercion be given a say in political decision-making. That might satisfactorily answer one of the questions that constitute the boundary problem—and might even help us answer the others, on the plausible assumption that these questions are related. But why should we believe that this account also applies to broader questions of political representation, to the procedural, conceptual, and value-focused formulations of the boundary problem, and to all the different functions that we may expect democracy to realize? Although, of course, there is nothing that decisively prevents a single principle from having such a large explanatory potential, it might seem more cautious to adopt, by default, a pluralistic approach, which leaves open the possibility that different principles might be doing some work vis-à-vis some specific aspects of the problem.

We Need not Necessarily Appeal to Distinctively Democratic Values

As we have seen, some authors believe that an adequate answer to the boundary problem must appeal to distinctively democratic values (such as solidarity, political equality, or the worth of political agency). It is, of course, quite difficult to precisely determine what counts as a distinctively democratic value. For the purposes of this paper, I shall take the three above values as prototypical distinctively democratic values. Conversely, I will assume that values that can contribute to explain why democracy is valuable, or legitimate, or authoritative, but which can be realized, and remain valuable, even when democracy is absent (such as well-being or individual autonomy) are not distinctively democratic.¹¹ Otherwise, it is not clear what the restriction under discussion would be able to rule out.¹²

One thought that might motivate the view that an adequate answer to the boundary problem must appeal to such values is that, after all, the boundary problem is essentially concerned with the basic architecture of democratic systems. An adequate answer to the boundary problem (or to the specific problems that compose it) must be, in short, an answer to a question about the structure of political decision-making in democratic societies. So stated, this claim is difficult to deny: an adequate response to a specific problem must be sensitive to the nature of the problem it purports to be an answer to.

It is not clear, though, that any of this entails that there should be a general constraint on how we theorize about the boundary problem that requires principles of inclusion to be grounded in distinctively democratic values. This is because it is an open normative question whether all of the reasons that we have to include individuals in the decision-making processes of democratic systems are based on such values. Consider, for instance, future generations and animals. Some authors have argued that we have reasons, derivable from at least some of the principles offered as an answer to the boundary problem—typically, the All Affected Principle—to represent the interests of those groups in democratic decision-making (Ekeli 2009; Vink 2020;  Magaña 2022a). These are intelligible, open questions. It might be possible, of course, that we have no compelling reasons to politically represent the interests of any of these groups. Now, if thinking about the boundary problem necessarily requires appealing to distinctively democratic values, then these questions would not even be open, since it does not seem that distinctive democratic values, as defined above, can adjudicate them—i.e. it is controversial whether solidarity and political equality can exist between us and animals or future generations, and in any case members of these groups are not political agents in a robust sense.

One could thus distinguish then between the question of whether someone should be included within the institutional machinery of democratic systems (for different sorts of reasons), and the question of which individuals should be included on the basis of distinctively democratic values. On a narrow reading, the boundary problem is thought to be essentially concerned with this second question (this is the sort of reasoning that I take to be motivating the claim that an adequate answer to the boundary problem must make reference to democratic values). On a broad reading, in turn, the boundary problem is thought to be concerned with the first question, and need not be restricted in the above way.

There are at least four reasons to favor the broad reading. First, the broad question already encompasses the narrow question. If we ask who should be included in democratic systems, for whatever reasons, this covers inclusion for strictly democratic reasons. Thus, adopting the broad reading does not require that we abandon more specific questions about inclusion for distinctively democratic reasons.

Second, the broad conception reflects more accurately how the boundary problem has been dealt with in the literature. For, as I pointed out above, some authors have argued that the principles developed in the context of the boundary problem have implications for collectives whose inclusion cannot be clearly grounded in democratic values, narrowly conceived. And many more have discussed those questions without seeming to worry about whether the relevant values—such as well-being or autonomy—could be construed as ‘distinctively democratic’.

Third, note that the broad conception does not contradict the thought that an adequate answer to the boundary problem must be an answer to a question about the basic institutional architecture of democratic societies, for this is exactly what the broad question asks. It is just that in asking about the allocation of political power, influence, and relevance in democratic societies, it does not restrict itself to the contribution of distinctively democratic values. Consider, for instance, Bengtson’s Value Requirement, which demands that principles of inclusion be able to support stronger forms of inclusion than a mere moral demand to consider all affected interests, and also that they be able to support democratic decision-making to begin with (as opposed to no democracy at all). There is no obvious a priori reason why principles of inclusion that do not appeal to distinctively democratic values cannot satisfy these two demands. Think, for example, about the classical consequentialist defense of both democracy and the All Affected Principle, which states that democracy (and, more specifically, enfranchising all those who are affected by political decisions) provides the best way to protect everyone’s interests in the long run (cf. Goodin 2007, p. 50). Regardless of whether the argument is ultimately sound or not, it is clearly one that attempts to respect, in its most basic formulation, the two demands of the Value Requirement.¹³ And yet it does not seem to appeal to any distinctively democratic values.

Fourth, it is not exactly clear what would be gained by constraining in such a way theorizing about the boundary problem. To begin with, it does not dissolve the substantive questions. Even if we stipulated that the boundary problem is only about democratic inclusion on specifically democratic grounds, one might then distinguish between the boundary problem and a broader problem of political inclusion in democratic systems, and focus on the latter. In response, one might argue that something that we gain is that by restricting the scope of the boundary problem we can focus on the specific contribution of distinctively democratic values to political inclusion. But, as I have argued above, this is something that adopting the broad conception does not preclude at all.

These four answers would also work for a version of the objection that claims, more specifically, that the boundary problem is only concerned with those modalities of inclusion that involve robust political agency (Saunders 2012). Again, a broader conception of inclusion does not preclude narrower questions about robust political agency and the distribution of participatory rights, reflects better the literature on the boundary problem (since, as we saw above, some authors have focused also on less demanding modalities of inclusion, for instance representation), remains a question about the basic architecture of democratic systems, and continues to raise normatively pertinent questions, which are not dissolved simply by adopting a narrower conception of the relevant modalities of political inclusion.

In a useful discussion, Erman (2014, p. 535) distinguishes between an exogenous view, according to which democratic theory (which focuses on, and works with, distinctively democratic values) cannot answer the boundary problem, and an endogenous view, according to which the boundary problem is internal to and part of the basic architecture of democratic systems. Similarly, Miller (2020, p. 6) contrasts an approach that imports elements from outside of democratic theory, with one that brings democratic values to bear on the boundary problem. If what I have argued in this subsection is correct, these two approaches are not rival: it is possible (and, I think, preferable) to accept both that theorizing about the boundary problem need not necessarily appeal to democratic values, and that those values can be relevant for (at least some) aspects of the boundary problem.

One interesting result of this discussion is that even if it turns out that the All Affected and the All Subjected Principles are unable to provide a fundamental principle of democratic inclusion, as Bengtson (2022) forcefully argues, it would not follow that they must be rejected. At best, it would follow that these principles are incapable of answering one important aspect of the boundary problem—but, if the arguments defended in the previous section are correct, that would not also be fatal.

Before concluding, it is worth noting that what I have said in this section is compatible with three weaker claims.¹⁴ First, one might claim that, albeit not strictly necessarily, appealing to distinctively democratic values is, ceteribus paribus, preferable. Although in this paper I remain agnostic about it, this possibility is fully compatible with the arguments defended in this section—which purported to show that appealing to other kinds of values is not necessarily worse, all things considered.

Second, one might also argue that, to the extent that one appeals to distinctively democratic values (or, more specifically, to a conception of democracy), then one’s principle of inclusion should cohere with the values underpinning one’s favored conception of democracy—as long, of course, as this conception is sound (i.e. captures something of what makes democracy valuable, or authoritative). If we endorse, for instance, a protectionist conception of democracy (i.e. of democracy as being, at least partly, about the protection of interests),¹⁵ then interests-protection must guide the choice among principles of inclusion (cf. Goodin 2007, p. 50; Miller 2009, pp. 204–207). Importantly, this would not mean that we must select a principle that explicitly appeals to interests. Suppose, for example, that the best way to maximize interests-protection is by enfranchising those subjected to political decisions. In that case, a defender of the protectionist conception of democracy might perfectly support the All Subjected Principle—but only because that is the rule that, in the long run, would maximize the protection of interests. Even though the values that underpin this conception of democracy do not explicitly figure in the corresponding principle of inclusion, the kind of coherence at issue is preserved at the justificatory level—i.e. the principle is chosen because it is the best way to realize the values underpinning the protectionist conception of democracy.

And finally, one might contend (in the spirit of Lippert-Rasmussen and Bengtson’s value conception of the boundary problem) that any answer to the boundary problem must appeal to the values that make democracy valuable. Since the value of democracy can be grounded in different kinds of values (only some of which need to be distinctively democratic in the above sense), this weaker requirement would be compatible with the arguments defended in this section.

Intuitions About Political Exclusion are Vulnerable to Biasing Factors

In the absence of undermining defeaters, conformity with our considered judgments renders a principle of political inclusion preferable. As I pointed out above, though, whether an intuition counts as a considered judgment or not depends on how vulnerable it is to the different factors that inevitably bias, to some extent, our normative judgments about political matters. In this subsection, I shall argue that we have some reasons to take our intuitions about who should be excluded with particular caution. This is not to say that these judgments can never be considered. The point is more modest: given the disastrous ways in which those judgments have proven mistaken in the past, we should avoid treating them as providing conclusive considerations against some principle of inclusion.

Intuitions about political inclusion and exclusion are arguably vulnerable to parochialism and endogeneity, as defined above. To begin with, some psychologists have argued that those of us who live in democratic, rich, and industrialized countries are psychologically peculiar (Henrich et al. 2010). Similarly, there seems to be some evidence that political preferences in democratic systems are endogenous (Fuchs-Schündeln and Schündeln 2015).

Although those facts recommend treating our intuitions about democracy (and, specifically, political inclusion) with some caution, I shall focus on one more specific—and potentially more troubling—feature of judgments of political exclusion. Most of us, at least among those who live in democratic systems, believe that it is inadmissible to exclude women, black people, or those lacking property from participating in democratic policy-making. But that was exactly the case for many centuries, and women, black people, and the propertyless were formally disenfranchised. Since they were already ‘virtually represented’, or so the argument went, they had no need for voting rights. In fact, as of 1900, only one country (New Zealand) had fully universal suffrage, while 17 additional countries only granted the right to vote to adult males (Przeworski 2009, p. 291). Even now, for instance, most democratic countries do not politically include alien residents—at least not for national elections. This, however, seems unwarranted on most (if not all) of the main accounts on offer, since they are deeply affected by political decisions, are subjected to the law, and to the state’s coercive structures, are part (or many of them, at the very least) of the societies within whose borders they reside, and there is no clear reason why they should not be treated as equals, as the ideal of relational equality demands.¹⁶

Now, if we got things so wrong in the past—and might still, in fact, be getting them wrong in the present—should not we adopt, as a rule of thumb, a skeptical stance towards intuitions about exclusion—i.e. those that require or permit it? Or, at the very least, give them lesser weight than we would otherwise do?

When a class of intuitions has such a track record, it seems sensible to treat them with caution. This is, admittedly, too coarse grained.¹⁷ After all, such an argument is unable to distinguish among intuitions about the exclusion of stones, children, and propertyless adults—as they are all intuitions about exclusion. But, clearly, there are important differences among these cases, and an adequate account should be able to capture them. Here, thus, we must introduce the following caveat: intuitions about the political exclusion of a set of entities must be assessed in terms of a body of other normative and evaluative judgments about moral exclusion. There is a well-established, large and coherent body of judgments—across many different domains of morality—featuring children and adults (male and female, property-owning or propertyless…) as fully morally considerable. Stones, however, do not typically elicit analogous responses. This, arguably, should be factored in when assessing the specific degree of (un)reliability of judgments about political exclusion: when there exists a large coherent body of judgments condemning the exclusion of some creatures and entities in other domains of morality, the case for caution holds sway (and it does so in proportion with the range and depth of the relevant body of judgments). Conversely, when such a body of judgments is absent, the corresponding intuitions about political exclusion might be less vulnerable to the debunking explanation suggested in this section.

Does this caveat render the precautionary principle innocuous? Not necessarily. Consider, for instance, the claim that the All Affected Principle entails that everyone, everywhere, should be enfranchised, or at least represented. As we noted earlier, many find this a reductio ad absurdum of the principle. But, if the above reasoning is correct, this response may not be warranted (not, at least, so straightforwardly), as it relies on a type of judgments with a very poor record track, and which also targets entities whose exclusion from other domains of morality seems precluded by a large and coherent body of normative and evaluative judgments. Briefly put, foreigners (even those in distant countries) are closer to women and the propertyless than they are, say, to stones. Of course, none of this means that this implication cannot be treated, at the end of the day, as a reductio ad absurdum of the All Affected Principle. My contention is, more weakly, that the underlying judgments belong to the category of intuitions which we must handle with special epistemic care.¹⁸

Conclusion

Answering the boundary problem requires having a good grasp of which desiderata an adequate principle of inclusion must satisfy. In this paper, I have identified five such requirements. An adequate principle of inclusion must, first of all, clearly specify its grounding value(s), and its standards and modalities of inclusion. Secondly, it must also develop a sound justificatory principle linking those three elements. Thirdly, it should also exhibit some degree of conformity with our considered judgments about political inclusion. In addition, I have defended three theses: (1) that given the complexity and multifaceted nature of the problem, a monistic answer is unlikely, (2) that an adequate answer need not make reference only to distinctively democratic values, and (3) that intuitions about exclusion should be handled particularly carefully, given their poor track record. This, of course, does not answer the boundary problem. But if the arguments here succeed, it might nonetheless help us think straighter about it.