A(nother) democratic case for federalism

Abstract

This work offers a new democratic case for federalism, understood as a form of governance in which multiple entities in a country possess final decision-making authority (viz., can make decisions free from others substituting their decisions, issuing fines, etc.) over at least one subject (e.g., immigration, defense). It argues that leading solutions to the democratic boundary problem provide overlapping arguments for federalism. The underlying logic and many details of the most commonly cited solutions focused on those relevantly affected by and subject to decisions each support three distinct arguments (focused on voteshares, other forms of democratic influence, and persistent minorities) for multiple demoi possessing authority in a polity. Federalism is the best available method for recognizing the distinct demoi. This not only supports federal governance as opposed to unitary governance and subsidiarity. It also suggests that democracy and federalism are importantly related and have several implications for institutional design.

This work offers a democratic argument for federalism stemming from prominent scholarship on the democratic boundary problem concerning who should be “eligible to take part in which decision-making processes” (Arrhenius, 2005: 14). ¹ The boundary problem initially arose partly as a problem of federalism (Dahl, 1983). Yet contemporary scholarship thereon rarely intersects with work on federalism. The following argues that otherwise-contrasting work on the boundary problem actually supports federal governance. The underlying logic and many details of two leading candidate solutions (viz., those focused on those relevantly affected by and subject to decisions) support arguments for multiple demoi possessing authority in a polity. Federalism, understood as governance in which multiple entities in a state possess final decision-making authority (viz., can make decisions free from others substituting their decisions, fines, etc.) over at least one subject (immigration, defense, etc.), is the best method for recognizing the demoi.

Leading solutions to the boundary problem more specifically support three distinct arguments for federalism as opposed to unitary governance and non-federal alternatives to centralized rule. Concerns with vote-shares, other forms of democratic influence, and persistent minorities each support dual jurisdictions in which all parties make some decisions and only some make others. Federalism affords stable institutional mechanisms for ensuring distinct jurisdictions that appropriately track justifiable allocations of decision-making authority over subjects. It also maximizes the chances that decisions on particular issues are made by the most appropriate entity. Further work must specify the form(s) of federal governance most likely to do this. But this article indicates several features most likely to fulfill democratic aims. ² It thereby further demonstrates the value of treating boundary and federalism-specific problems together.

The first two sections provide background information/motivation. The next section argues that proponents of plausible affect- and subjection-based solutions to the boundary problem should desire multiple voting fora within states that are best realized in federal rule. The following section argues that federalism offers distinct forms of influence over decisions beyond direct voting all democratic theorists should desire. A final substantive section argues that federalism can address the persistent minority problem(s). The substantive sections jointly provide a convergence argument for federalism: proponents of leading accounts of democracy have multiple reasons to support federalism regardless of how they characterize affect, subjection, democratic influence/participation, or other core features of their views. ³

Democratic boundaries and federalism

Philosophers of democracy and federalism face similar problems and burdens. Boundary problem scholars seek a democratic principle that can identify “whom should have influence [on a decision] and how much influence they should have” (Arrhenius, 2018: 102n25). This is not a merely a practical issue arising where a political community's basic features are contested, as in secession or immigration disputes, but implicates core theoretical questions. ⁴ If, e.g., legitimate authority rests on members of a democratic community having a particular kind or amount of input into decision-making processes or authoritative decisions require such input, a complete account of authority must identify which members qualify and the scope/nature of their claims. ⁵ Accounts can then be valuable even where perfect influence/authority allocations cannot be realized: theoretical solutions to the boundary problem offer normative ideals for allocating influence over decisions one can use to judge real allocations, including in practical debates. ⁶

The boundary problem was initially posed as one concerning relevant units for democratic decision-making. Dahl (1983) and Whelan (1983) questioned why many democratic theorists assumed the unit of democratic decision-making was pre-constituted: Democratic decision-making requires a previously-identified decision-making body, threatening regress. ⁷ Each found candidate principles for identifying the groups lacking but called for greater scrutiny. The goal was to identify who should be part of a given demos. Yet Dahl (1983: 3–4, 11) recognized solutions to the boundary problem are incomplete if they address “the problem of inclusion” (“What persons have a rightful claim to be included in the demos?”) without addressing the “scope of its authority” (“What rightful limits are there on the control of a demos?”). While Dahl (1970, 1983) severed analyses, ⁸ contemporary scholars (Arrhenius, 2018; Miller, 2020; Wilson, 2022) recognize one cannot identify the proper unit of democratic decision-making without addressing the decision space's scope and domain. Miller (2020) thus distinguishes democratic boundaries’ “domain (where and to whom do decisions apply), constituency (who is entitled to be included in the deciding body) and scope (which issues should be on the decision agenda).”

The need to identify who can possess authority over what and the scope thereof tracks elements of the assignment problem (Rodden, 2006) in federalism studies, which focuses on whether and how powers should be assigned between a country's different levels of governments. Democratic and federal theorists must each identify who can legitimately claim final decision-making powers (or ‘authority’ whereby no other entity can substitute its own decision or directly compel a change) over particular subjects (e.g., healthcare, immigration) or issues (e.g., where to place vaccination centres or grant citizenship claims) and when that entity should be able to and should actually make such a final decision. Each must explain whether, when, and why, e.g., a province should decide where to place a hospital. Federalism scholars focus on subjects as federal governance is constituted by a division of powers over subjects. However, problems of and burdens for properly allocating authority are consistent across domains. ⁹

Despite these parallels, contemporary work on boundary and assignment problems rarely intersects. ¹⁰ This is understandable. The boundary problem may be analytically prior to a federal ‘division of powers’ allocating authority to different governance units (central governments, provinces, cities, etc.). ¹¹ Moreover, many early works on the problem understandably debated whether democratic principles required extending citizenship, voting, or, less often, other forms of influence to non-citizens. ¹² Others (also understandably) focused on secession referenda. ¹³ This created path dependencies away from federalism. Many solutions are also cosmopolitan in ways that challenge federal theories’ state-based commitments (Wilson, 2022: 171). ¹⁴ While scholars working on the assignment problem sometimes appeal to ‘democratic’ solutions (e.g., Weinstock, 2001), non-democratic federations (e.g., the USSR) and non-federal democracies (e.g., Israel) are often considered legitimate. ¹⁵ Many question whether democracy and federalism are even compatible. ¹⁶ If the concepts conflict, they are unlikely to solve each other's problems.

Exploring whether and how democratic solutions address federal concerns nonetheless appears wise. Even if one worries that federalism presupposes answers to questions the boundary problem should solve, principles from one setting may apply in another. If they fail to do even this, they could guide analyses of what democracy should look like within pre-determined federal boundaries. ¹⁷ Democratic theorists working on the boundary problem (e.g., Bengtson and Lippert-Rasmussen, 2021a, b) are, e.g., beginning to address broader authority within states, like how to identify a group who should make decisions about immigration and even local traffic laws (e.g., Arrhenius, 2018), or identify challenges concerning democratic governance beyond states, like those posed by diaspora groups seeking self-governance (e.g., Spiro, 2018). Examining how proposals would assign powers in federal states helps scholars test whether conceptions of democracy plausibly answer questions about the domain and scope of such groups’ authority.

Whether democratic principles plausibly guide federal authority allocation can, more broadly, provide insight into debates about any relationship(s) between democracy and federalism. ¹⁸ Federalism is one of the most common forms of governance. ¹⁹ This demands justification. ²⁰ Self-avowedly ‘democratic’ normative principles compete or combine with appeals to pluralism, self-determination, epistemic goods, and other values as primary normative justifications for federalism and means of identifying when power can be decentralized (Føllesdal, 2003/2018; Da Silva, 2022). ²¹ Yet many such ‘justifications’ consist in remedies in perceived deficiencies with ‘pure’ democracy, as when decentralized federalism is said to provide “laboratories of democracy” in which decisions/policies can be tested and better ones can be identified and implemented. ²² Democratic considerations are, in turn, often said to render federalism illegitimate. Concerns federalism is anti-democratic appear across time and jurisdictions. ²³ Kelemen (2006: 221) succinctly states the concern: insofar as “[f]ederalism constrains majorities … it is clearly undemocratic.” Federal governance divides decision-making authority and so constrains a majority's ability to implement its preferred decisions: it “either effects an unfortunate compromise of democratic norms or constitutes a salutary corrective to democratic excess, a counter-democratic measure to disperse and restrain popular power” (Abizadeh, 2021b: 742). ²⁴ Institutional mechanisms for implementing federal governance may further limit democratic rule. Concerns that federalism eliminates democratic communities’ ability to set their own political agenda by placing certain powers beyond their purview or that federalism requires judicial review to enforce the division of powers are just two examples. ²⁵ The relationship between federalism and democracy thus remains vexed: “federalism and democracy may be viewed as mutually reinforcing, though the two can just as well contradict one another” (Sonnicksen, 2018: 31). The boundary problem literature can help resolve these disputes.

Scrutinizing the implications of conceptions of democracy is also important for understanding the nature and preferred structure of even theoretically acceptable forms of federal and non-federal governance. Democracy could require equal vote-shares for all polity members, making decentralization of authority at best non-ideal (Kelemen, 2006). Or it could require greater vote-shares for particular bodies to, e.g., protect them against undue central state power (e.g., Levy, 2007) or reflect their stronger claim over a matter due to, e.g., ways it will impact them more (e.g., Weinstock, 2014). The latter would at least support decentralization. Democratic concerns provide normative ideals one can use to judge authority allocations (Arrhenius, 2018: 188–189).

Definitions and scope

The following provides further reasons and proof of concept for pairing questions about the boundary problem and federalism. Prominent works on the boundary problem point towards democratic arguments for federalism. ‘Democracy’ and ‘federalism’ accordingly relate in philosophically interesting ways demanding further analysis. Detailing how solutions to the boundary problem specifically address aspects of the assignment problem focused on identifying particular powers’ scope and domain requires other articles. The (overlapping set of) democratic argument(s) for federalism detailed below demonstrate/s why such analysis remains desirable.

For these purposes, I understand federalism as a system of governance in which (at least legislative) power is allocated “between at least two levels so that units at each level have final authority and can be self-governing in some issue area” (Føllesdal, 2003/2018). I assume this ‘division of powers’ has some constitutional protection such that no entity that is provided with a power can direct face interference within its domain of authority (viz., be subject to a fine for or veto or substitution of a decision) absent a new constitutional understanding. ²⁶ Any definition of federalism invites controversy. ²⁷ This one at least reflects common ground among otherwise competing approaches, including the standard view on which federalism is defined by a combination of shared-rule and self-rule (e.g., Elazar, 1987; Watts, 2008), and “institutional” views defined by institutional commitment to a constitutional division of powers (e.g., Da Silva, 2024). ²⁸ It also reflects usage in the boundary problem literature. ²⁹ Where this working definition describes paradigmatically federal states, like the USA and Germany, and distinguishes them from unitary states in which one actor has all final decision-making authority, like France and Israel, and those using other forms of non-unitary rule below, it is also apt for comparisons here.

I follow Levy (2007) in recognizing that a defense of federalism must demonstrate that it is preferable to both unitary governance and other forms of governance in which decision-making authority is ‘decentralized.’ Federalism here is non-identical with a broad understanding of decentralized governance. ³⁰ The division of powers characterizing federal rule could provide a central authority with many powers so long as non-central bodies maintain some spheres of exclusive authority. Whether more or less centralized allocations of authority within federal states is desirable is, however, distinct from whether federalism is preferable to forms of governance in which power is allocated in other ways. The kind of ‘devolved’ authority recognized in the UK, in which the Westminster Parliament provides executive and legislative powers to Scotland, Northern Ireland, and Wales via statute but retains the right to substitute its own decisions or even revoke decision-making powers, is a core contrast case regardless of whether one characterizes it as a form of unitary governance or a form of decentralized rule.

I also follow Levy (and others) in viewing subsidiarity arrangements as a contrast case. ³¹ Subsidiarity is the principle whereby powers presumptively belong to the most local body possible (Føllesdal, 1998; Levy, 2007; etc.). It is sometimes offered as a principle for allocating decisions within federal states (Føllesdal, 1998). In such cases, subsidiarity speaks to the question of whether centralized or decentralized federalism is preferable. However, subsidiarity is also frequently invoked as an alternative to federalism whereby (at least some) allocative decisions should be beyond the scope of constitutional decision-making because each allocative decision requires application of the subsidiarity principle or because certain groups have distinct domains of authority with which states cannot interfere (e.g., Cahill, 2021; Muñiz-Fraticelli, 2014). ³² Subsidiarity is then said to demand, e.g., that decisions about where to place a vaccination centre be preceded by an analysis of which ‘local’ group is capable of addressing it or that states refrain from interfering with religious education. Either result contrasts with federalism as defined here.

Understanding federalism in terms of a constitutive constitutional division of powers provides a useful demarcation point for comparing the democratic bona fides of federal and non-federal rule. I take no further stands on disputes about the nature of federalism and whether, e.g., consociations are plausibly ‘federal.’ ³³ Federalism so-defined nonetheless comes in many forms. One distinction is between “cooperative” or “coordinate” federalism in which “in most areas, decision-making and implementation require action by both levels of government and thus their integration” and “dual federalism” whereby each “at least one exclusive (traditionally non-overlapping) domain of authority in which the entity can act as it sees fit” (Da Silva, 2024). A related distinction contrasts “vertical federalism” with distinct entrenched divisions of powers and “horizontal federalism” whereby different actors are represented in each other's institutions (Abizadeh, 2021b). Another contrasts “coming-together federalism” whereby pre-existing entities combine into a new federal union with some shared powers, as in the USA or Canada, and holding-together federalism, whereby a unitary body divides powers among sub-units to maintain a now-federal body, as in Belgium or Nepal (Stepan, 1999). Still others contrast republican and monarchical, parliamentary and presidential, and uni/mono-national and pluri-national systems (Watts, 2008, Hueglin and Fenna, 2015, Palermo and Kössler, 2017, Da Silva, 2022). Pluri-national states with multiple recognized ‘people’ are more likely to adopt asymmetrical federalism whereby some sub-state units have more powers, rather than symmetrical federalism whereby they have identical powers. ³⁴ I discuss such distinctions as relevant below. However, the arguments for federalism generalize even if some support particular forms more than others. ³⁵

My democratic analysis of and argument for federalism focus on the two most prominent families of proposed solutions to the boundary problem concerning those affected by and those subjected to a decision. I do not thereby take substantive views on alternatives, which merit attention elsewhere. ³⁶ Demonstrating federalism's import for democracy is simply best-established by first attending to the most commonly cited views. ³⁷

Affect-related proposals come in at least two forms. ³⁸ The more prominent ‘all-affected’ principle holds that “all people who are affected by a decision have a right to participate in making that decision” (Wilson, 2022: 170). This too admits multiple interpretations. Key distinctions concern how to understand ‘affect’ and ‘participation.’ Proponents debate whether only those whose ‘interests’ are affected have a right to participate (cf. Goodin, 2007, Arrhenius, 2005, Wilson, 2022) and, if so, whether all actually and all possibly affected interests trigger participatory rights (cf. Goodin, 2007, Owen, 2012). ³⁹ Accounts require diverse levels of participation from voting rights to other forms of representation in democratic processes, like deliberative fora. ⁴⁰ The principle is thus sometimes discussed in terms of “influence” over decisions (Arrhenius, 2005: 20; Miklosi, 2012: 483). ⁴¹

The second affect-related proposal focuses on those most affected by a decision. Bengtson and Lippert-Rasmussen (2021a: 573) note that an affect-based principle underlies many “ways in which we delimit the relevant demoi. For instance … subsidiarity reflects the fact that decisions ought to be taken by those (most) affected by them.” This is one of few boundary problem-related arguments prominent in federalism studies where the idea that those ‘most affected’ by a decision should make it is a possible justification for federalism (Weinstock, 2001). While that field rarely attends to diverse senses of ‘affect’ in boundary problem disputes and ‘most affected’ principles may be unable to support ‘subsidiarity’ and ‘federalism,’ ⁴² the parallel is notable.

Subjection-based views, by contrast, suggest a decision's legal, or law-like, force helps identify the relevant demos. The all-subjected principle holds that anyone who will face non-de minimus consequences from a decision should have a role in that decision. As Abizadeh (2021a) puts it, “for political rule to be democratically legitimate, those subject to it must have a right of democratic say.” ⁴³ The principle is often framed in terms of being subject to a law (e.g., Lopez-Guerra, 2005; Arrhenius, 2018: 105). However, many accept that subjection to a decision takes numerous forms, and some need not implicate formal law. Debates about whether people are relevantly ‘subject’ to non-territorial or regional laws (cf. e.g., Miller, 2020; Abizadeh, 2021a, b) notwithstanding, for example, many believe states subject persons to coercion that can take three forms: “coercive acts” (viz., “noncommunicative direct coercion”), “authorizing coercive acts” (“laws” subjecting persons to coercive acts), and “communicative legal coercion (in threatening punitive harms)” (Song, 2012: 17 (building on Abizadeh, 2010: 123)). ⁴⁴ I address additional relevant distinctions among subjection-based views when detailing arguments below.

Vote-Shares

The first democratic argument for federalism holds that those concerned with affect or subjection should both support split demoi in which some decisions are made by all relevantly impacted parties and others by those relevantly impacted to a greater degree. ⁴⁵ The logic of both families of solutions to the boundary problem produce this result. Even supporters of the seemingly most expansive conceptions of a proper demos also express commitments to some domains where those most impacted by decisions should make them. The bifurcated demoi are best produced and protected through federal governance. Where solutions to the boundary problem reflect democratic interests, overlapping results here present democratic support for federalism.

The case for (some) decentralization

To begin, subjection- and affect-related views provide arguments for splitting vote-shares to provide more influence for those most impacted by at least some decisions. Proponents of each can and should recognize that those relevantly impacted by a decision should have a greater say over it. Vote-shares are (at least) a useful proxy for what matters when making these decisions (Bengtson and Lippert-Rasmussen, 2021a: 574), particularly where they also identify representatives and provide mathematical guidance for tailored ‘influence’ supporting decentralization on different models. ⁴⁶ Votes at more ‘local’ levels have greater impact than those at higher ones.

Concerns with properly allocating vote-share-based influence support decentralization on subjection- and affect-based views. Federal governance balances these concerns with competing (oft-centralizing) demands better than alternative forms of decentralization, including devolution.

Subjection-based views offer a direct argument for forms of decentralization within states consistent with federalism. They are connected by commitments to the idea that those subject to rules should play a role in its creation. While being ‘most subject’ to a decision may sound bizarre linguistically, persons are not always relevantly subject to decisions in the same way. This basic idea is familiar to any lawyer. It supports some decentralization. ⁴⁷ Differential subjection to decisions should impact influence thereon. For instance, many rules primarily target particular populations, like city-dwellers. Consider a simple decision about the height of a bridge over a road connecting a city centre to a more peripheral beach. Low bridges limit beach access for those who would rely on public transit to reach it, including many urban residents who do not otherwise need/desire cars (e.g., Caro, 1974). A provincial government led primarily by rural voters who frequently use cars may be insufficiently aware of this complication. This presents a(n at least defeasible) case for more local rule. While one may quibble with particular beach access cases, ⁴⁸ they help motivate why city-dwellers should determine similar decisions.

Given the lack of precise correlation between the scope of laws (or other decisions) and those subject to them, practical application of this principle requires heuristics. It may seem odd if, e.g., cities make decisions about local parks where other provincial residents might use them. But rule at the provincial level is still less problematic in an immense country where residents of other provinces rarely use them due to geography. The normative ideal may remain that all subject to the law should have a role in its creation. However, those actually facing possible sanction have stronger claims to relevant subjection, favouring some decentralization. If applying this ideal requires concessions to practical reality and heuristics, it is no worse than alternatives. Proponents of the all-affected principle, e.g., grant the need for such concessions. ⁴⁹

Affect-based views likewise support some decentralization. This is most obvious for explicit proponents of ‘most affected’ principles who seemingly accept it by definition. However, anyone interested in affect as a guide to democratic input should recognize a most affected principle has some merit. Statements by prominent all-affected principle proponents suggest they agree. The all-affected principle is often taken to have an expansive account of the proper democratic polity and, by extension, the site of democratic decision-making. Goodin (2007) suggests it theoretically entails a near-universal franchise or other form of world governance given today's highly-interconnected transnational interests. Arrhenius (2005, 2018) and Wilson (2022) reach similar conclusions. Yet even these otherwise-globally-oriented all-affected scholars grant that theoretical arguments for global governance need not entail such institutions be realized. ⁵⁰ I contend that reasons purporting to support the global demos/oi they envision on theoretical grounds may not favour centralized rule in state-like bodies. ⁵¹ Those interested in matching affect with influence instead have reasons to sometimes permit levels of each to track.

Theorists on competing sides of boundary problem debates (Saunders, 2012, Bengtson and Lippert-Rasmussen, 2021a, etc.) suggest the all-affected principle also supports subsidiarity but refer to a most affected principle when doing so. ⁵² Goodin (2007), Arrhenius (2018), and Wilson (2022) actually suggest their principle fits well with a principle of subsidiarity whereby authority should presumptively be more local. Wilson (2022) and Goodin (2007) frame this in terms of relations between global bodies and states, but the idea that concerns with those ‘most affected’ should presumptively support allocating powers to local units also appears in domestic analyses. Shifting to concerns with those ‘most affected’ highlights that any plausible view admits distinctions. Vote-share-based arguments at issue note that one's influence over a decision is greater in a 1000-member group than in a 100000-member group. If only the 1000-member group is relevantly impacted, members should at least have greater influence on the outcome. Situations where the 1000-member group or its representatives (since members of the 1000-member group also has more say over the representative decision-maker chosen than those in the 100000-member group) make the decision ensure those most affected by the decision most influence it. This likely supports decentralizing decisions concerning many subjects and issues.

The case for federalism particularly

While Goodin, Arrhenius, Wilson, et al. accept subsidiarity, federalism better addresses affect- and subjection-based concerns. All-subjected and -affected principles are characterized by recognition that many decisions impact many persons in the relevant sense and this should provide them with influence over decisions. Basic arguments above suggest different vote-shares can accommodate different levels of impact decisions have on particular groups. But they are too basic on their own if they ignore each scholar's further commitment to all persons having influence on matters that affect them or subject them to certain treatment. Forms of dual jurisdiction, like federalism, better fulfill the needs to provide all persons with some influence over certain decisions than subsidiarity arrangements and other forms of decentralization and yet provide others with a larger sphere of influence due to their unique position. On such views, all persons have an equal vote on matters clearly impacting them all. Some have greater vote-shares over matters relevantly impacting them more. Those promoting ‘all-affected’ views, for example, are right to note that the impact of many ‘local’ decisions cannot be contained and providing no vote-share to those facing non-trivial impact on their rights demands justification. Federalism better instantiates these competing pulls than unitary governance, devolution, or subsidiarity.

Federalism provides an institutional framework for realizing commitments to equal vote-shares for some matters and weighted vote-shares for others. The uniquely impacted have more influence but all have some. Interests above are best protected through federal governance in which each decision-making body has final decision-making authority. First consider groups affected to a much greater degree by a decision. Those concerned with subjection should seek to ensure a ‘local’ decision-making body is not subject to others’ undue influence. The possibility of a central government substituting its own decisions always lingers in unitary states. Even those with devolved authority thus face a threat of coercion. If Scotland's authority remains revocable by the government in London, a consistent threat that a larger body could replace its decisions at any time remains. ⁵³ If a child can spend their allowance as they please but their spending power is revocable, the latter possibility will frame and limit their spending habits. Scotland is no child. Devolution can treat it like one. Federal governance in which provinces or even cities are sovereign in their jurisdiction avoids this possibility: federal governments cannot substitute their decisions, fine the provinces, etc. when they dislike provincial choices. Not all subjection-based views focus on coercion, but the idea that only entities with full authority can fulfill members’ rights to influence decisions has merit. Those concerned with affect should likewise seek a means of ensuring that most affected parties can uniquely influence decisions in some cases. Next consider the larger group of all relevantly impacted parties. Knowing one will always have some path to influence decisions that will impact one's vital interests is desirable. Knowing particular contexts in which this will occur long-term helps one strategize to protect those interests. A person with Scottish roots and cultural affinities in London should feel confident that they can influence rules affecting their rights even if they lack a direct vote for Scottish Parliament seats. In both cases, federalism protects relevant interests at distinct levels.

Institutional considerations also favour federalism. A constitutional division of powers provides stability that can obviate needs for continued discussions of boundary and assignment problems. ⁵⁴ A well-designed federal division of powers will foster state stability by settling some contentious political issues. In so doing, the settled allocation of powers provides all decision-making bodies with a means of planning decisions over time. This should increase opportunities to govern well. Opportunities to judge decisions remain necessary, but boundary problems scholars have long recognized that one cannot constantly relitigate the demos. ⁵⁵ Even a presumption of local rule maintains too many opportunities for relitigating allocations. ⁵⁶ A well-designed federal division of powers leaves some room for debate– e.g., contours of powers remain contestable –but limits the decision space in ways that foster necessary stability.

Getting a federal division of powers right should also increase the chances of well-placed allocations of authority over particular issues and particular decisions. One may suggest a demos should be (re-)constituted for all weighty decisions and allocating decision-making authority among subjects, rather than issues, forecloses this possibility. However, one cannot enter into complex deliberations over every weighty issue as theorists from Dahl to Goodin recognize. Even common referenda are implausible today, helping explain why Goodin et al. abandon them. A well-designed federal division of powers over subjects should make it more likely that decisions over issues are made by more local bodies where appropriate, providing the best approximation of even their regulative ideal. While we can judge decisions about when parties devolve powers on democratic grounds, a practical mechanism for ensuring powers are generally well-allocated remains necessary. Federal governance dividing powers over subjects in ways most likely to ensure proper levels will make decisions over issues provides an institutional means of best matching affect or relevant subjection with influence across time and space.

Where additional institutional mechanisms remain necessary to ensure consistently compelling impact-influence matches, divisions of powers provide a framework for debate. Difficulties amending constitutions in many states admittedly limit their capacity to correct severe errors. ⁵⁷ This need could simply support forms of flexibility within federal states. Concurrent jurisdiction, in which persons have influence in different degrees at different levels of governance indexed to impact on populations, reflects competing desires and permits some flexibility. ⁵⁸ For instance, a federal government could control environmental regulation, provinces state-specific commerce, and cities boating and fishing licencing within and rules for access to local waters. That arrangement could ensure waterway-specific decisions are made locally subject to broader norms. One can then adjust powers by applying rules to cases. If, e.g., historical understandings of a legislative power over ‘the environment’ consistently enables federal governments to dominate provincial minorities and underlying issues also plausibly fit under a provincial ‘trade and commerce’ power, the domination should trigger judicial re-articulation of relevant powers.

What qualifies as ‘good design’ for the purposes of fulfilling this aim can depend on factual circumstances in particular communities or on which other justifications for federalism apply and make additional demands. However, democratic considerations provide compelling guidance on these issues. If, for instance, one is in a plurinational state with a history of conflict and it is established that self-determination-based concerns provide additional arguments for and constraints on federal design, this may favour forms of asymmetrical federalism in which some provinces encompassing sub-state ‘nations’ have powers that other provinces do not. ⁵⁹ Democratic considerations could also favour this outcome if only such groups were uniquely impacted in ways that established valid claims to jurisdiction. The first point here is that some decisions could be made by pointing to other factors. Peacekeeping or self-determination-based concerns could provide additional or alternative support for asymmetry. However, the second point is that the kind of democratic concerns at issue are not silent on design questions. The forgoing supports mechanisms for fostering debate about the contours of powers; concurrent jurisdiction; and judicial review. It also offers reason to question subsidiarity within federal states. The argument for federalism does not rest on the desirability of these features. But democratic considerations pointing towards certain features suggests my view is generative.

Proponents of subjection- and affect-related views ultimately express commitments that are best instantiated via the forms of influence at ‘higher’ and ‘lower’ levels reflected in and best institutionally realized through federal governance. This presents a democratic argument for federalism. Persons are distinctly subject to or affected by decisions in relevant ways. This provides them with a strong claim to increased influence that can be well-realized through an increased vote-share, which itself is well-realized through more ‘local’ rule. Federal governance realizes this in a manner consistent with the general impetus towards an expanded franchise implicit in affect- and subjection-based views. Federalism provides opportunities for everyone to influence decisions that impact them in votes at higher levels and for those more affected to have more influence at lower ones. It may work better where concurrent jurisdiction is possible: concurrency offers multiple opportunities for influence over the same subject and for contestation where an allocation of authority over subjects leads to implausible results over many issues. But arguments here stand even if one rejects concurrent jurisdiction on other grounds.

Objections/replies

Understandable objections do not defeat this argument. One objection suggests that the preceding leaves important questions unresolved. For example, Miller (2020) suggests affect-related principles provide little guidance for the boundary or assignment problems: saying fishers should “have a voice when fishing quotas are being allocated because of their obvious interest in the outcome does not yet settle which level of decision-making” is right for that kind of decision.

Yet subjection- and affect-based views both supporting federalism of some kind is notable even if further work on its implications is necessary. Subjection- and affect-related views support some decisions being made at different levels. The above provides guidance on which groups can have powers and when: one should divide powers in ways that maximize the vote-shares of those most affected, however best-defined. Identifying relevant levels of impact (be it subjection or affect of some other kind) and distinguishing cases where ‘all’ persons are relevantly impacted from those where only a subset is impacted is difficult. But federalism also provides institutional contexts in which heuristics can be adopted and modified to better reflect distinctions. Increased vote-shares are non-determinative on who should make decisions about many subjects and issues. Even affected-based theorists should, e.g., grant those most affected by a decision may not always be institutionally capable of addressing underlying issues. If, e.g., a city lacks capacity to address healthcare-related concerns, this is a good reason not to provide it with such powers even if the city faces unique issues. However, this does not mean cities should never have increased power or defeat the democratic arguments above/below. All arguments in this area present prima facie claims to authority subject to other practical and theoretical realities.

Another objection suggests the affect-based arguments above distort the all-affected principle by, e.g., adopting a ‘weighted’ articulation of affect-based views (Brighouse and Fleurbaey, 2010; Bengtson and Lippert-Rasmussen, 2021b:1027; etc.) without sufficient argument. This is less plausible where those committed to “fundamentally egalitarian” affect-related views whereby all interests must be weighted equally for any decision, like Goodin (2007: 49), concede ‘most affected’ views’ plausibility. However, the idea that all interest holders should have an equal vote is intuitive and common. Luckily, one need not uniformly accept weighted views to accept that affect-based views support decentralization and federalism. The above suggests even egalitarians should recognize equal consideration of interests permits splitting vote-shares where necessary to protect some interests. Providing for different demoi, each of which reflects competing demands, helps do so. If, e.g., one wants to weigh all interest-holders or interests equally, circumstances where the character thereof or differential impact decisions will have thereon should permit different levels of interest-holder influence. This is likely why Goodin et al. accept the most affected making some decisions despite rejecting interest or affect weighting. Those seeking egalitarian demoi and opportunities for distinct communities should again favour federalism.

If different readings of ‘affect’ present diverse results on which decisions should be made by which bodies, that would not defeat the case for decentralization and federalism. Choices about how to identify those affected and levels of affect have theoretical and practical implications. But attendant challenges leave the present argument standing. Consider a proposed fishing quota in Maine. It will directly impact Maine fishers and less directly impact fishers in Massachusetts and even Nova Scotia. Whether it impacts fishers in those other territories of the USA and Canada more than broader food supply and environmental conditions and subjects thereof is less clear. If, e.g., affect is ‘direct impact on one's vital interests in a way foreseeable prior to the decision,’ this could favour smaller groups, permitting and supporting forms of decentralization. If Maine fishers alone face direct impacts from quota, they may appropriately set policy. So too could the Maine village of Stonington. If, by contrast (following Wilson, 2022), ‘affect’ is read in terms of a(n unweighted) “non-trivial affect of the context of others’ choice,” it implicates a much larger group: a fishing ban could have a non-trivial affect on the food choice of many, likely including residents of California and perhaps those in other countries relying on U.S. imports. Even this broad reading does not entail that decisions about fishing must always be made at a national, let alone global, level. If the best understanding of ‘affect’ led to the quota decision being best-made at the national level, for example, provincial or municipal governments may maintain significant influence on implementation. Even this most centralizing view can favour some decentralization. This example highlights how levels of influence should differ regardless of how we taxonomize and quantify affect or impact; most proposals support some decentralization. The preceding provided further reasons to think federalism best balances this impetus with competing concerns.

Other forms of influence

While Miller (2020)'s objection also raises questions about the form of influence an entity should possess over decisions in ways that may challenge the preceding arguments’ appeals to vote-shares, those interested in diverse forms of influence over decisions should favour federalism even more. This distinct argument faces similar challenges as the prior one, limiting the need for detail discussion of objections. The argument can also be more succinctly stated and evaluated.

Federalism, in short, presents greater opportunities for differential influence for differently impacted groups while maintaining a distinct sphere of full democratic equality. Alternative forms of influence facially challenge the case for federalism. While many boundary problem scholars focus on the franchise, others suggest that democratic solutions only require other forms of influence over decision-making, like representative governance (Arrhenius, 2018; Boudou, 2023; Miller, 2009; etc.). Owen suggests different forms of impact trigger claims to different forms of influence: some groups are “entitled to voice or representation of their interests in the decision-making process,” others are “entitled to contest the government's decisions,” and still others to “authorize the government's decisions” (2018: 184). ⁶⁰ If this is correct and the ‘right of participation’ triggered by relevant forms of affect- or subjection- only requires those impacted have some influence over a decision, final decision-making power need not rest with decentralized bodies (Owen, 2012, 2018; Saunders, 2012; Lagerspetz, 2015; Miller, 2020; etc.).

Affected interests, for example, may require X to have means of political participation short of decision-making authority. Many formal (e.g., public fora, public feedback on draft legislation, participation in administrative bodies) and informal (e.g., newspaper editorials, online interactions) feedback mechanisms offer opportunities for participation in and influence on policy development absent direct voting rights. If an increased vote-share is not necessary or sufficient for increased influence, decentralization meant to increase that influence, federal or otherwise, could prove unjustified. Some further believe solutions to the boundary problem require “institutional flexibility” (Wilson, 2022: 191). Federalism must not unduly limit possibilities for groups to exercise their affect- or subjection-based claims to influence decisions.

Yet if a proper solution to the boundary problem should involve different groups having different kinds of influence, this should actually favour federalism over alternatives. Federalism offers mechanisms for realizing appropriately differing levels of formal and informal influence that democrats should desire. Vote-shares were, recall, only a useful formal proxy for what mattered above, namely the ability to have influence over all decisions that impact you in the relevant way but more influence on decisions that relevantly impact you more. Federal governance appears particularly valuable on this score. It presents distinct opportunities to influence outcomes through political participation even absent a vote, etc. in the composition of a body with final decision-making authority over a subject or issue. Where it does so at multiple levels– e.g., one can lobby or participate in debates at federal, provincial, or municipal levels –it offers increased opportunities for entities impacted in different ways to exercise correspondingly different claims.

Another example further clarifies this argument. If one lives in Edmonton, Alberta, Canada, one has no direct say in the Canadian province of Ontario's decisions about which healthcare goods to publicly fund in an Ontarian hospital despite federal transfers partially funding every provincial healthcare system. One also lacks direct influence in the composition or decisions of the city council in Alberta's other major city, Calgary, which often has a more laissez-faire approach to governance. Calgary could make decisions about, for instance, city street maintenance that impact those frequently travelling between the cities, to say nothing about decisions about housing for local homeless, which is necessary in the very cold province, producing incentives to move from a city that does not provide it to one that does. Yet the citizen has a vote in federal elections that will impact the amount of federal funding available for healthcare services in Ontario and in provincial elections whose winners will set policies to which municipalities must comply (and who possess constitutional control over municipalities in Canada). Edmonton has multiple representatives in Alberta and Canada's respective legislatures. ⁶¹ So, ‘influence’ on at least subjects is possible on even the most basic metric. This too demonstrates that all persons having a right to participate in decisions that affect them need not entail that final decision-making authority even presumptively rest at particular levels. It need not undermine a case for federal governance. One could read the division of direct and less direct forms of influence over areas as reflecting proper levels of affect or subjection above. This argument too is stronger where some concurrent jurisdiction exists– if a decision space includes multiple decision-makers, it can more easily present opportunities to influence relevant outcomes –but does not require concurrency given that apt possibilities to informally influence decisions remain absent concurrent jurisdiction. Federalism could even permit those affected by decisions to have certain forms of influence while those subjected have another, as Owen et al. desire.

While other forms of decentralized governance or even subsidiarity arrangements may offer different forms of influence, arguments for federalism particularly in the last section also apply here. Some other forms of decentralization may also offer means for matching impact with diverse forms of influence, but devolution and subsidiarity arrangements face additional challenges. Devolution agreements, for example, offer multiple sites of influence. If claims for decisive forms of influence only ever applied to the entire community, they could offer useful means of allocating diverse forms of influence. However, these types of claims would be equally realized in unitary governance and it is unlikely that decisive claims are so-allocated in any case. Owen and Miller, for example, disagree on many points; yet neither suggests that decisive claims must always rest with a centralized body. While subsidiarity arrangements may then appear well-suited to the kind of “flexibility” desired by some advocates (e.g., Wilson, 2022) to maintain distinct forms of influence, such arrangements prove too flexible to secure meaningful influence long-term. One cannot know who one needs to lobby to have meaningful influence over a decision if domains of jurisdiction are in flux. One may, in fact, consistently need to influence the body that identifies authorities in given cases, creating a situation functionally similar to that of members of a unitary state. At worst, then, this argument makes federalism preferable to many other forms of government and makes a strong case therefore when combined with the other two.

Persistent minorities

A third democratic argument for federalism stems from the problem of persistent minorities. Proponents of affect- and subjection-based solutions to the boundary problem should seek to solve the problem of persistent minorities. Federalism offers one of the strongest solutions.

Some groups consistently fail to have their preferences recognized in a polity given the groups’ size and polity's decision-making processes. This raises questions about whether they have sufficient influence over given decisions as to be legitimately bound by them. This is the most basic form of the problem of persistent minorities. Abizadeh (2021b) defines persistent minorities as “groups whose members’ preferences are … negatively correlated with those of a persistent majority … over a broad range of issues over time.” Lee (2001: 262) translates the problem for the voting procedures central to many boundary problem arguments, defining a persistent minority as “a group of persons who are consistently members of the minority over a series of votes.” ⁶² Many discussions of persistent minorities focus on ‘national’ cultural groups. ⁶³ However, as Christiano (1994: 173–174) demonstrates, the underlying concern arises wherever there are large socially salient differences between groups that have significant impact across a range of issues and the relative size of the groups produce one longstanding majority who consistently ‘wins’ and (at least) one consequent minority who consistently loses. The minority then “doesn’t get its way enough” (169) on plausible metrics, which is intuitively problematic.

The problem of persistent minorities, like the boundary problem, takes several forms. ⁶⁴ Many discuss a “tyranny of the majority” in which a majority group attempts to and often succeeds in dominating another (Requejo and Sanjaume-Calvet, 2023). ⁶⁵ However, the distinct problem of persistent minority groups ‘losses’ can apply absent majority ill-intent. ⁶⁶ What, if anything, makes these losses problematic is contested. Many discuss the problem in terms of persistent minorities’ interests. Christiano, for example, details majoritarian decision-making processes creating conditions in which minority interests are not reflected in the “collective features of society” (1994: 175–177). ⁶⁷ This can amount to a de facto lack of opportunities to participate in the creation of a society (2010) and alienate minorities therefrom (1994), undermining stability. ⁶⁸ Others identify non-interest-based wrongs. Many are status-based. The problem is often framed as one in which members of minorities and majorities have equal power but the power is less valuable for members of the minority (Mráz, 2023). For instance, each person has one vote but members of majorities alone consistently see their votes affecting policy. “More-than-equal power” for minorities is then necessary to create equal status in terms of being able to effectively influence the shape of society (id.). Abizadeh (2021b) further argues that this form of unequal political standing is problematic where and because persistent minorities consistently lack power to enact their preferred policies and so are consistently subject to majority power. The problem is not simply that such minorities do not “get their way” but that decision-making processes render persistent minorities unable to influence decisions and so subjects them to unjust domination. ⁶⁹

Persistent minorities pose a problem for affect- and subjection-based views of democratic legitimacy regardless of whether the problem is framed in terms of interests or status. I accordingly take no strong view here on how to characterize the wrong or the precise threshold qualifying minorities as ‘persistent.’ ⁷⁰ While some deny that persistent minorities are possible, ⁷¹ the “consensus” in mainstream political theory is that they are a valid posit for ideal and non-ideal approaches (Mráz, 2023). The existence of persistent minorities challenges any view that takes a link between impact and influence as central to democratic legitimacy. Some groups will always ‘lose’ absent some domain of final decision-making authority. Such groups accordingly lack proper influence over rules that affect them or to which they are subjected. Absent creation of some new state for each persistent minority, federalism offers one of the best methods for providing these groups with a domain of decision-making in which they can have proper influence on at least some core decisions. It may prove necessary for that purpose.

Abizadeh offers an intuitive version of the argument, which I build on here. ⁷² Abizadeh understands the all-subjected in republican terms, seeking to minimize domination (2008) or other unjust power relations (2021b). ⁷³ Persistent minorities, at least, must sometimes determine decisions that impact them lest they be subject to domination. Subjection thereto is accordingly illegitimate. He provides an example: “If French Canadians would be perpetually outvoted in majoritarian institutions, then they would not be treated as political equals. Departures from majoritarianism—such as those enshrined in federalist arrangements—are sometimes indispensable for democratic equality” (743). Variations from country-wide majority rule across are, in other words, required to fulfill democratic concerns with ensuring those impacted by decisions can actually influence them. Absent such variations, Abizadeh argues, purportedly democratic procedures will violate what he takes to be the wrong of persistent minorities, namely circumstances in which majoritarians unduly wield powers over minorities.

Abizadeh's basic logic has broader application. Subjection- and affect-related views should address persistent minorities. One need not be a republican to find it problematic when one group always ‘loses.’ Those concerned with autonomy, for example, can and should be concerned when persons are consistently unable to bring about policies reflecting their will. Saunders (2010) highlighted the importance of ensuring persistent minorities maintain some influence in earlier work distinguishing majority rule and democracy. Saunders (id.: 153)'s justice-based approach was explicitly contrasted with a republican one. The interest-based version of the persistent minority problem then offers an even more straightforward account of why accounts of democratic boundaries should worry about persistent minorities. Persons being consistently unable to further their interests through democratic processes should worry anyone seeking to justify the composition of a polis on interest-based grounds. The fact of persistent minorities, then, challenges multiple subjection-based views and at least the interest-focused, if not all, affect-based ones. If a group with distinct interests that would otherwise be capable of protecting those interests systematically cannot exercise that capacity due to another group's control of democratic processes, this undermines the very goods purporting to justify democracy.

This should favour decentralization generally and federalism particularly. Control over policy-making is plausibly required to avoid systematic constraints on adequate minority influence in some circumstances. The full decision-making authority provided to sub-state groups in federal states can avoid concerns about subjugation found in other forms of decentralized rule. Threats of more central authorities retrieving their final decision-making powers can themselves qualify as subjugation, as our Scottish case made clear. It is, further, more likely that a group (e.g., Scots) will be able to protect its interests where it need not worry about interference by another entity (e.g., Westminster). These concerns support and can help structure well-designed federal governance even if they fail to fully justify federalism. For instance, persistent minorities’ concerns are especially acute when decisions speak to the minority group's constitutive ends, like language/culture, helping to explain Abizadeh (2021b)'s focus on cultural minorities. This results also helps identify which powers they can possess and their scope: those groups likely require powers over culturally-salient subjects, like language, to fulfill their constitutive ends.

Yet the persistent minorities problem is not merely cultural and requires greater institutional imagination about authority allocation, minimally demanding scrutiny of whether and how federal design could further democratic ends. Note, e.g., recent work demonstrating demographic sorting of more liberal individuals in metropolitan areas and more conservative ones in rural areas: liberals then persistently “lose” in federal and state politics (Rodden, 2019). If cities now not only face unique challenges due to, e.g., density but constitute populations with distinct values that cannot be realized at higher levels of governance, this supports stronger municipal powers. ⁷⁴ But municipal powers are parasitic on central or provincial ones in major federations (e.g., the USA, Canada), limiting city-dwellers’ ability to realize their values and leaving powers they possess subject to the will of another government and its voters. ⁷⁵ Stonington, Toronto, Edmonton, etc. are in similar positions vis-à-vis provincial bodies as Scotland is to Parliament in London and, worse, have less political power to avoid revocation.

North American immigration debates underline this reality. Most immigrants to the USA and Canada settle in cities. This is rational given aforementioned sorting and liberal tendencies to favour fewer immigration restrictions. However, municipal attempts to control immigration via ‘sanctuary city’ proposals have been severely limited by cities’ lack of formal constitutional status in those countries. ⁷⁶ Regardless of what one thinks about the merits of that example– one could plausibly hold that immigration ‘affects’ everyone and speaks to existential concerns beyond municipal competence –it is emblematic of a general mismatch between municipal values and municipal powers. City-dwellers accordingly plausibly constitute a persistent minority in many liberal-democratic states. At minimum, there is now reason to seriously consider further devolution of power to cities to avoid creating persistent minorities without proper democratic voice. However, any democratic deficit is then most likely to be solved via a revised form of federalism that recognized municipal powers over at least some subjects. Otherwise, cities will always face the possibility of undermining third-party interference. ⁷⁷

Federalism alone may not address all persistent minorities problems but is no worse for not being a panacea. A cultural minority with linguistic policies could, e.g., be otherwise dominated. Federalism may not be the only way to address the problem of persistent minorities even where federalism would address aspects of the problem. Abizadeh (2021b: 753) suggests federalism is most likely to address the problem where “the formal-procedural power of all subjects is equal and (a) no one has informal power over others’ political preferences; (b) there is only one social structure that induces a persistent majority/minority cleavage, and this is the only such cleavage; and (c) the persistent minority is territorially concentrated.” The absence of any conditions can call for different institutional responses. That of (c) particularly challenges federal solutions. Abizadeh thus understandably promotes other means of addressing the problem if/where it does not obtain. One should take care not to overstate the case for federalism. However, federalism appears particularly well-suited to address aspects of the persistent minority problem long-term while gaining further support from arguments above. If federalism only addresses the problem of persistent minorities under certain conditions, there is at least now a prima facie case for examining whether those conditions can obtain in any country at issue without causing further problems. ⁷⁸ And if and where federalism alone cannot resolve the problem, nothing here bars alternative approaches. Many existing approaches to the persistent minorities problem could be supplementary. For example, representational body reforms discussed by Abizadeh could work in tandem with federal governance. ⁷⁹ Likewise, federalism not only offers a method of instantiating ‘minority rights’ but can also work alongside specific minority rights provisions. ⁸⁰

Federal solutions to the persistent minority problem have the additional benefit of less radically departing from existing institutional practices than many proposed alternatives. For instance, Abizadeh (2021c) and Saunders (2010) also promote lottery voting/sortition as means of securing democratic equality for otherwise-persistent minorities. That more radically departs from the status quo than federalism, whatever its merits. Saunders notably considered his case for lottocracy secondary to his primary argument against majority rule. This work builds on Saunders’ primary argument (and Abizadeh) rather than treating the secondary one too briefly. It suffices here to note that even lottocracy/sortition that protects democracy against majority rule may not provide the distinct demoi supported by the present range of democratic arguments. Adopting federalism is also less radical than secession or other forms of state dissolution. Where decisions of an encompassing state will continue to impact persons in a breakaway state, this will remove a method for their properly influencing such decisions. While secession could be justified for many reasons– some suggest self-government alone can fulfill self-determination rights necessary to address a ‘tyranny of the majority’ (e.g., Requejo and Sanjaume-Calvet, 2023) –it is a radical response that may come with democratic costs. Indeed, even those who champion self-determination recognize that federalism has democratic benefits (id.: 197). Subsidiarity arrangements likewise present threats of instability outlined in preceding sections.

Remaining objections to this line of argument merely point to the need for further scholarship on democracy and federalism. One suggests concerns about persistent minorities do not specify which minorities should have which powers. Miller (2020), for one, also charges that the all-subjected principle may explain why one should have a right of participation in a decision but “assumes that the domain and scope issues have already been settled.” This concern is acute when one must also identify which groups can plausibly have final decision-making powers within states. Which groups are dominated when control over cultural and linguistic policy suffices to avoid domination is unclear. Perhaps some minorities cannot avoid domination within states. What, for instance, should one make of a territorially disperse group that can never be a majority and secure power even in sub-units of a territorially-defined state? ⁸¹ Pointing to the problem of persistent minorities cannot, critics suggest, help one validly choose between competing minority claims, let alone identify which powers ‘majorities’ can or must still possess.

Useful initial indications on how to address this concern above admittedly leave questions open. Yet such questions merely highlight that federalism is, again, not a panacea and spotlights a need for further analysis. Persistent minorities present genuine challenges on leading responses to a central problem of democratic theory that is likely best solved via some form of decentralization. Federalism appears especially well-placed to resolve underlying issues. If no stable federal arrangements can address attendant problems, the seeming connection may not hold. However, the preceding provided empirically-informed theoretical reason to hypothesize that federalism can resolve the problem and analyze whether and when that is so. Concerns that persistent minorities may remain in even a reasonably-just federation, especially where a minority is territorially dispersed, ⁸² likewise call for further analysis and suggest Abizadeh and others are right to also examine whether and how representative bodies can address any lingering concerns. However, no single approach is likely to solve every problem. A case for federalism of some kind remains strong where it offers useful means of addressing several distinct problems.

Another objection holds that addressing the persistent minority problem in the context of the boundary problem obscures operative moral interests. Saunders (2012: 281), for instance, acknowledges the problems above but suggests focusing on ‘affect’ or ‘subjection’ distorts analyses. Basic concerns with autonomy require that persistent minorities sometimes “win” in ways favouring some decentralization but that no one is “entitled to impose costs on those who are not part of the cooperative scheme … [so p]rohibition of or compensation for negative externalities imposed on outsiders is the first-best ideal.” Decentralized authority is minimally justified where it is necessary for realizing member autonomy but only where any negative externalities thereof are severely limited. But affect and subjection are distorting red herrings. ‘Persistent minority’ status need not track level of affect or ‘subjection’ status. A critic might add that neither status necessarily tracks ‘unjustifiably restricted in pursuit of autonomy.’ However, such a contention presents a similar outcome to the final objection to vote-share arguments. Indeed, it strengthens the present argument by identifying another set of democratically-oriented persistent minority concerns that favour decentralization generally and federalism particularly.

Conclusion

Dominant approaches to the boundary problem provide overlapping arguments for federalism. Even those promoting an ‘all-affected’ principle implicitly recognize the value of some forms of decentralization when appealing to how affect can promote subsidiarity. The logic and leading proponents of affect- and subjection-based views also point towards the value of having larger sites of democratic decision-making, even within the same jurisdiction. Federalism's commitment to multiple authorities with distinct jurisdictions is well-suited to institutionalizing these overlapping ends. Well-designed federal arrangements even present opportunities for properly weighing competing pulls towards local and broader rule and ensuring the appropriate entity is most likely to make decisions about discrete issues. Some present opportunities to revise arrangements, though the circumstances in which this may occur require scrutiny. Other factors can render federalism inapt for a given country (e.g., where it would undermine stability). However, leading theories of democracy provide strong theoretical bases for federal rule in their absence.

This result presents a strong prima facie case for a relationship between democracy and federalism and demands further research on its possible forms. The forgoing alone does not (yet) entail that federalism requires democracy and the multi-pronged nature of the convergence argument for federalism presents options for how federal states are ‘democratic.’ However, the ways in which distinct democratic concerns could justify federalism suggest just federal institutions may need to be democratic in some way. Even this basic link vindicates my inquiry.

The preceding also points towards other promising research projects. One would more closely examine the arguments above in light of the details of particular affect- and subjection-based views. Examining their implications for the scope and domain of authority can help determine whether and how subjection- and affect-related views need to be revised to plausibly guide authority allocation. It will also illuminate whether the best solution(s) to the boundary problem track the best solutions to the assignment problem, democratic or otherwise. One can then examine whether solutions to either problem plausibly describe or distort central concepts.

A related project would examine federal design and its implications for just federalism(s). Analyses above provide some guidance by e.g., strengthening cases for municipal powers as an element of just federalisms and for overlapping jurisdiction to enable necessary contestation and flexibility. However, much more must be said about how differences in affect- and subjection-related views favour particular forms of decentralization or federalism. For example, further exploration of the importance and limits of concurrent jurisdictions and areas where persistent minorities must maintain non-overlapping authority is now warranted. Examining democratic theories’ implications for particular subjects/issues should also prove valuable. Concrete applications of the above to queries concerning how similar countries (e.g., the USA, Australia, or even non-federal states exercising devolution, like the UK) should structure authority allocation would highlight the stakes and precisify the contours of debates about democratic governance and just authority allocation. I hope the above motivates this necessary work.