Hostname: page-component-586b7cd67f-rdxmf Total loading time: 0 Render date: 2024-11-22T09:08:19.095Z Has data issue: false hasContentIssue false

Subsidiarity and the Allocation of Governmental Powers

Published online by Cambridge University Press:  08 November 2022

Michael Da Silva*
Affiliation:
University of Southampton, United Kingdom

Abstract

Every country must allocate final decision-making authority over different issues/subjects within its boundaries. Historically, many scholars working on this topic implicitly assumed that identifying the features providing entities with justified claims for authority and the entities possessing those features would also identify which groups should have which powers (or vice versa). However, many candidate allocative principles select multiple entities as candidates for some sub-state authority and yet fail to explain which powers each should possess. Further work must explain which groups should possess which powers when and what to do when two groups can make equally-valid authority claims using the same principle. Subsidiarity, the principle under which authority should presumptively belong to the entity representing those ‘most affected’ by its exercise and capable of addressing underlying problems, is one of the few principles focused on identifying which groups should have which powers. Unfortunately, subsidiarity alone does not provide guidance on many issues/subjects. Useful subsidiarity-related guidance relies on balancing underlying justificatory interests, which do the real allocative work. Another allocative principle remains necessary. A deflationary account of subsidiarity’s allocative potential nonetheless provides insights into how to articulate a new principle and accounts of subsidiarity that can fulfill other moral roles.

Type
Research Article
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright
© The Author(s), 2022. Published by Cambridge University Press on behalf of University of Western Ontario (Faculty of Law)

Every country must make decisions about how to allocate governmental powers. Each faces a host of questions ranging from the existential (e.g., ‘When should a person be able to immigrate?’) to the prosaic (‘Should a barbershop to be allowed to open on this street?’). Countries who claim legitimate rule should provide a principled basis for identifying who can answer these questions. The ‘authority allocation problem’ examines how to justifiably allocate decision-making authority, understood as the power to make decisions free from direct interference (substituting decisions, fines, etc.) from others.Footnote 1 While often framed as an issue for federations, all countries, including centralized ones, must decide which entities will have authority over which issues (viz., discrete questions at given times, like ‘Should we approve this licence?’) and subjects (viz., law/policymaking domains, like ‘healthcare’). The government in Paris cannot address every pressing or prosaic far-away issue in classically centralized France.

Many scholars working on these issues assume that identifying the features providing entities with justified claims for different kinds of authority and entities possessing those features would also identify who should have which powers when (or vice versa). Yet developments below highlight the need to distinguish the features necessary to make something a candidate authority and the factors that fully justify allocating particular powers to particular candidates. Many purportedly allocative principles select multiple entities as justifiably able to claim some ‘sub-state’ powers and yet fail to explain which powers they should possess or what to do when two groups have equally valid claims under that principle. Most candidate principles cannot, for example, explain when and why a province, rather than a city, should control healthcare policy.Footnote 2

This article demonstrates the need to distinguish ‘ontological’ features that can identify candidate authorities and the genuinely ‘allocative’ principles needed to specify who should have authority over which issues and subjects. It then explores the allocative potential of subsidiarity, the principle under which authority should presumptively belong to the entity representing those ‘most affected’ by its exercise and thus most capable of addressing underlying problems. Subsidiarity is one of the few candidate allocative principles in international law, European Union (EU) law, domestic constitutional laws, and legal/political philosophy.Footnote 3 It purports to appropriately allocate authority within countries and shows some promise for doing so. If, for example, a province and city are both candidates for a power under a principle whereby unique cultural entities can be authorities, subsidiarity could allocate it to the ‘closer’ municipality. ‘Most affected’ and ‘closeness’ are, of course, contested, possibly non-equivalent terms. Yet rough understandings guide real-world allocations. Subsidiarity is thus a good test case for analyzing allocative principles. Identifying allocative ‘work’ that subsidiarity can or cannot do also identifies its strengths and limitations, thereby contributing to subsidiarity studies. If, for example, subsidiarity must be fundamentally linked to communitarian interests to serve any allocative function, interpretations of the principle justifying it on communitarian grounds will be preferable.Footnote 4 If subsidiarity-based allocations are then consistently plausible, subsidiarity is valuable: it resolves a common, persistently challenging problem. Exploring competing conceptions’ allocative potential could help us choose between them or highlight the need for other ways to establish subsidiarity’s value.

My analysis ultimately demonstrates that subsidiarity as such is not a useful allocative principle. Subsidiarity does not provide guidance on many issues/subjects, especially regarding divisions of constitutional powers. Useful subsidiarity-related guidance balances underlying justificatory interests, which do the real allocative work. At least one other allocative principle is accordingly necessary, and proponents must justify subsidiarity’s value in other ways.

To establish this conclusion, I first outline the need for allocative principles and how conceptions of subsidiarity purport to allocate authority within countries. I then argue that no existing conceptions of subsidiarity fulfill the desiderata of a plausible allocative principle—namely, the need to provide concrete, intuitively acceptable guidance on which entities should possess authority over particular subjects (and when) that can at least apply in liberal democracies and to explain at least a de minimus number of real-world allocations of final decision-making powers—and subsidiarity’s characteristic feature, a presumptive allocation of authority to the ‘closest’ level of governance. I further argue that any account of subsidiarity that appears plausibly capable of providing concrete guidance collapses into a commitment to supporting principles that make the further appeal to subsidiarity redundant—and could lead to a misallocation of authority where supporting principles do not actually support ‘close’ rule. Despite this deflationary conclusion, analyzing subsidiarity’s allocative potential remains valuable. It explains why and when the principle seems compelling (and, consequently, why one may wish to identify another moral role it could fulfill) and it highlights the need for and desiderata of another allocative principle, the pitfalls candidate principles may face, and a method for identifying a better principle.

1. The Importance of Scrutinizing Potential Allocative Principles

Contrary to present scholarly trends,Footnote 5 questions about how to allocate authority within countries should not be confined to federal and provincial (including canton, länder, US states, etc.) governments, federations, or constitutional moments. A compelling allocative principle should guide decisions about which powers should rest with which government in federations, but this need not apply to federal and provincial governments alone. For instance, cities may also be relevant.Footnote 6 Moreover, countries need principles for when authority should rest with which groups for their everyday functioning. Central governments require principles for which powers they should delegate to others and when. A plausible authority allocation principle should provide guidance on the subjects over which entities should have authority in constitutional divisions of powers, where authority over issues should rest, and, by extension, when one should delegate it. This principle should allocate authority over subjects and issues in centralized and federal countries.

Recent developments challenge traditional allocations and allocative principles in ways that highlight the importance of—and require further work on—authority allocation. Rawlsian political philosophy initially assumed unitary countries and only gradually explored how minority rights and federalism impact the structure of just societies.Footnote 7 Theoretical discussions of authority allocation thus often occurred in law and political science; focus on constitutional divisions of power was common.Footnote 8 While some works addressed basic allocative concerns, common emphases on narrower issues (e.g., areas of ‘overlapping’ federal and provincial jurisdiction) was understandable given the low chances of constitutional change and generation of plausible allocative principles that explained real, prima facie justified decisions.Footnote 9

Yet recent events challenge the allocative status quo. For example, sanctuary city debates and demographic differences between municipal and rural residents led to calls for greater authority for cities.Footnote 10 For another, cities, Indigenous nations, and other sub-state ‘national’ groups demanded increased authority over public health management after purportedly unresponsive or otherwise problematic central/federal/provincial actions related to COVID-19.Footnote 11 At the same time, political philosophers increasingly recognize that principles ‘justifying’ provincial control over some topics equally ‘justify’ control by other entities. Per Daniel Weinstock, such principles (analyzed below) equally, if not better, justify municipal authority.Footnote 12 So, countries that “incorporate … [decentralizing principles] should on pain of arbitrariness apply them to cities.”Footnote 13 Avner de-Shalit, Ran Hirschl, and others make similar claims; per Hirschl, the principles also justify ‘rural’ authority.Footnote 14 Helder De Schutter et al. then argue that even sub-state ‘nations’ who lack their territory still possess unique cultures that could justify forms of non-territorial autonomy (and thus sub-state authority) on the self-determination-based grounds justifying provincial control in some countries.Footnote 15 Accounts of authority (and political justice at large) should explain if, why, and when these demands should be fulfilled.

These developments also highlight an often-overlooked distinction between features providing entities with justified claims for authority and factors specifying which groups should have which powers. For instance, self-determination is often discussed as if it answers the ontological question, ‘What entities can justifiably possess powers in a country?’, which addresses the features that make groups valuable in ways that could justify providing them with some authority, and the more specific allocative question, ‘Which entity should possess a specific decision-making power?’Footnote 16 Yet if, for example, a province and a city each have self-determination-based claims to control language policy, another principle must allocate authority between them. Several scholars view parallel arguments for provincial and municipal control as evidence that those who allocate powers to provinces should allocate them to cities, too.Footnote 17 Yet they might just make both of them the kinds of things that could have some authority. Other considerations should explain which powers each should possess and when. They may not require more constitutional powers for cities, but a complete account of authority allocation should address municipal claims.

We must accordingly separate the ontological and allocative questions and scrutinize the genuinely allocative potential of candidate principles for resolving the latter. Detailing issues with existing principles clarifies the problem. To begin, most existing principles fail to specify the precise powers entities should possess. For instance, the scope of moral self-determination rights is contested. Claims that they should entail powers over cultural policy for unique cultural groups are plausible.Footnote 18 Yet institutional forms other than sub-state autonomy may best protect cultures.Footnote 19 Moreover, even a ‘right’ to authority (in the sense at issue) could be defeasible. Self-determination rights can be constrained by, for example, the need for broader state stability.Footnote 20 Relationships between features grounding self-determination rights and particular issues/subjects are also unclear, leaving self-determination underspecified as allocative guidance. For instance, while some claim that self-determination rights-holders should control health policy,Footnote 21 culture and health are hard to link absent controversial posits about health’s cultural determinateness, and considerations above seemingly support central control.Footnote 22 International law’s minimalist commitments on the scope and institutional forms of cultural autonomy are thus unsurprising.Footnote 23

Other candidate allocative principles also raise problems: ‘decentralizing’ one’s appeal to entities’ unique interests/values or challenges; democratic benefits of providing authority to the entity ‘closest’ to the citizens capable of addressing it or best reflecting the interests of the ‘most affected’; or epistemic benefits of ‘local’ control.Footnote 24 Yet provinces, cities, and sub-state nations could receive some authority under each. No principle clearly specifies which entity should possess which powers where multiple entities possess equal claims under that principle. Attempts to break justificatory ‘ties’ then often beg questions. Groups may, for instance, disagree on how to draw the boundaries of ‘most affected’ groups or importance of territory for protecting cultures.Footnote 25 ‘Most affected’ status or geography cannot then be justificatory tiebreakers.

The prima facie justification of various authority allocations further suggests that the ontological features of candidate authorities cannot determine all allocative questions. There is, for instance, wide variance on where healthcare authority resides in federations: self-determination-based decentralization decisions and coordination-based centralization decisions each seem legitimate in some contexts.Footnote 26 At best, one must scrutinize the underlying principles to explain why prima facie justified choices are not fully justified.Footnote 27 The features of entities with prima facie justified authority claims alone cannot resolve these issues. A theory is likely better if it explains how the features that make an entity ontologically valuable could justify allocating powers to it, but different principles may apply at each stage. While self-determination and subsidiarity are discussed as competing principles,Footnote 28 subsidiarity could guide allocating powers amongst entities with equally compelling self-determination ‘rights’-based claims. A province and a city could both have extraordinarily strong self-determination rights with an epistemic understanding of subsidiarity best distinguishing their claims to particular powers.

A genuine allocative principle capable of addressing the preceding should provide guidance on which powers should belong to which entities in federations where powers are formally separated, what federations should do in areas of shared jurisdiction, when authorities in all countries should devolve to other entities, and where authority should rest where no entity possesses explicit powers. Ideally, it should also specify which powers should belong to which entities in a way that secures stability, but it should at least explain when decisions should rest with each entity. A fully action-guiding principle should explain why and when particular entities should possess decision-making powers. To play an oft-intended “democratic structuring” role,Footnote 29 it should explain which entities should have authority over which subjects, not merely when it should be able to decide an issue. This need not require that all parties to a debate share one understanding of a principle or its implications. It only requires that the principle have some basic features that present distinct moral reasons favouring particular allocations. For instance, while self-determination admits several interpretations and its application in real cases will often be contested, self-determination-based claims are usually grounded in moral reasons and particular groups should possess particular powers for common ends.Footnote 30 These reasons appear distinct from appeals to, for example, the epistemic value of particular allocations even absent perfect specifications of when and how the related arguments differ.Footnote 31

It is possible, even probable, that no one principle plausibly answers all allocative questions. Even then, one should explain the restricted scope of any principle one invokes. For instance, if different considerations apply to devolution and federal divisions of power, one should explain why this is so and why narrowly-tailored principles applying only to one are desirable. If, in other words, asking one principle to answer all allocative questions is unfair, we should still examine which candidates offer distinct, intuitively compelling results such that they should be operative principles stakeholders can validly raise to support particular claims in a context. I will now demonstrate that appeals to subsidiarity do not establish a general presumption that the most local entity should possess authority as claims, and the presumption’s seeming plausibility in some cases is best explained by other (e.g., democracy- or epistemic authority-based) principles. Moreover, real-world applications do not provide intuitively compelling results favouring local control. Subsidiarity, then, should not be even one of our standalone allocative principles. Any value that the concept may have will stem from other potential moral roles.

2. Subsidiarity’s Allocative Potential

Subsidiarity, again, holds that authority should presumptively belong to the entity representing those ‘most affected’ by its exercise and capable of addressing underlying problems. Subsidiarity, too, admits multiple interpretations.Footnote 32 Yet this provisional understanding refers to an influential concept. The EU’s power-sharing rules reflects its basic features.Footnote 33 Subsidiarity is also oft-discussed as a means of structuring divisions of governmental powers at international and domestic levels.Footnote 34 While it could fulfill other functionsFootnote 35 —and the best theory thereof may answer questions about, for example, how to allocate powers between state and non-state entities, rather than allocating constitutional powers within states or state-like entities like the EUFootnote 36 —subsidiarity is often discussed as the best principle for allocating authority within countries.Footnote 37 Its application to this task is often taken as self-evident: its “most obvious implications are for constitutional ordering within” countries.Footnote 38 Subsidiarity promises to allocate authority over issues and subjects.Footnote 39 While it is viewed as most compelling for dividing powers during constitutional transition,Footnote 40 its allocative potential within countries is supposed to extend to any allocative moment, constitutional or otherwise, such that it identifies when it is appropriate to delegate constitutional authority and create quasi-constitutional forms of revocable authority.Footnote 41

Subsidiarity is thus a plausible starting point for addressing authority allocation questions. Addressing subsidiarity’s allocative potential should also contribute to subsidiarity studies: if subsidiarity cannot fulfill this role, accounts should attend to its other possible roles (e.g., its ability to plausibly resolve state v. non-state conflicts).Footnote 42 Cataloguing subsidiarity’s forms and purported justifications suggests that subsidiarity could guide authority allocation in several ways. Evaluating each conception’s ability to fulfill the allocative role could help us choose between them. While some believe subsidiarity should identify candidate authorities and the issues/subjects over which they should possess authority,Footnote 43 subsidiarity would be valuable if it could only allocate authority over particular issues/subjects amongst groups previously identified as having good claims to some authority. Recall the case of a province and a city with equal self-determination-based claims. If subsidiarity-based allocations are consistently plausible, subsidiarity beneficially resolves a problem common to all countries. If it cannot plausibly allocate authority, we may seek an account that can fulfill another moral role.

Subsidiarity could guide authority allocation at several stages in a country’s development and in several ways, but all uses of subsidiarity for authority allocation purposes seek to establish a presumption favouring decentralization. Subsidiarity could, for instance, guide decisions about how to allocate powers at constitutional moments (e.g., an initial constituent assembly, moments of reform) or when entities decide when to devolve their powers to other entities. Realpolitik proponents may question whether abstract principles actually influence either type of decision. Yet decision-makers in both contexts normatively should—and often are politically required to—provide reasons for their decisions. Subsidiarity is at least proffered as a reason for decisions in both relevant contexts, and subsidiarity’s advocates claim that subsidiarity can create a default presumption for local control over many issues/subjects.Footnote 44 Even if, moreover, decision-makers did not appeal to principles to justify their decisions, a set of normative principles we can use to judge allocative decisions remains desirable. Subsidiarity is at least a good candidate for that role. It may also play a role in evaluating how entities understand their powers—for example, highlighting how the powers should be understood as necessarily enmeshed with decision-making at other levels—or create a presumption of common devolution of those powers to more local levels.Footnote 45

Subsidiarity could also, and purportedly does, guide judicial decision-making about constitutionally-entrenched powers.Footnote 46 Courts use the principle to resolve jurisdictional disputes to clarify explicit allocations of authority and to fill jurisdictional gaps. One can examine whether subsidiarity provides adequate guidance on how to do so and whether invoking it actually provides intuitively compelling allocations to local authorities. Even if one is primarily interested in how to allocate authority for constitutional divisions of powers or devolution, literature on judicial decision-making helps clarify how the concept can be understood in practice and offers another way in which subsidiarity could guide decisions. I accordingly discuss some judicial decisions even when focused on other allocative moments below.

I make no assumptions here on whether an inability to allocate authority within countries will challenge its ability to fulfill other tasks. The question ‘What makes entities candidate authorities?’ is largely taken as answered to explore subsidiarity’s allocative potential. The next section (surprisingly) demonstrates that no form of subsidiarity as such provides proper allocative guidance but attending to its forms/justifications clarifies its potential and limits. It demonstrates that arguments for subsidiarity either fail to provide specific, definite guidance on how to allocate powers; provide guidance that is unintuitive and at odds with prima facie acceptable existing authority allocations; or only present plausible guidance at the expense of subsidiarity’s most fundamental commitments or collapsing it into other principles. To be clear, my concern is not that subsidiarity is not a panacea. Rather, my concern is that no plausible specification of subsidiarity fulfills (at least one of) the concept’s primary task(s): providing normatively acceptable guidance on how to allocate authority based on a defeasible presumption of local control. This is so of each form of authority allocation in the preceding paragraphs.

Further detailing different conceptions of subsidiarity that aim to guide authority allocation within countries and claim consistency with liberal-democratic norms clarifies the issue. For instance, Andreas Føllesdal’s prominent EU-influenced view holds that “powers or tasks should rest with the lower-level sub-units … unless allocating them to a higher-level central unit would ensure higher comparative efficiency or effectiveness in achieving” state aims.Footnote 47 Others remove or add features. Hirschl does not appeal to comparative efficiency or effectiveness; his version “refers to the norm favoring the preservation or allocation of power to local authorities.”Footnote 48 N.W. Barber requires allocation to those ‘most affected’ by a policy, not the ‘lower’ unit alone: subsidiarity “does not just embody a preference for smaller units over large ones: it allocates powers to the states containing the people who will be affected.”Footnote 49 Still others connect EU-influenced and earlier ‘communitarian’ conceptions focused on protecting all ‘local’ groups, including trade unions and religious groups. John Finnis’s subsidiarity principle holds “that it is unjust for a higher authority to usurp the self-governing authority that lower authorities, acting in the service of their own members[,] … rightly have over those members.”Footnote 50

These definitions share features (permitting provisional definitions) but disagree on core issues. Føllesdal echoes Barber in invoking a most affected principle.Footnote 51 Yet authors disagree on subsidiarity’s scope of application. While the EU views it as applying in areas of “shared” authority,Footnote 52 it is also offered as a principle for allocating unique powers in the first place.Footnote 53 While Barber characterizes subsidiarity as a principle of ‘constitutionalism’ in the sense that it applies to all countries at all times and should be implemented in the institutional structure of all ‘constitutional’ countries,Footnote 54 others treat it as an allocative option.Footnote 55 Subsidiarity’s relationship to federalism and self-determination also differs across accounts, leading to different outcomes.Footnote 56

Subsidiarity is also alternatively discussed as creating a presumption in favour of ‘local’ control,Footnote 57 a burden that must be met before powers can be allocated to more central authorities,Footnote 58 a duty to “aid” others in exercising their powers,Footnote 59 and/or a duty to assist them in fulfilling their self-defined ends.Footnote 60 Views on what can defeat subsidiarity’s presumptive decentralizing tendency vary. Even Føllesdal notes that his “comparative efficiency” condition can be understood as a necessity condition or an effectiveness condition.Footnote 61 Whether federal governments are justified in acting when provinces will not act or only when they cannot do so also arises in the constitutional law of countries adopting subsidiarity or functionally equivalent principles.Footnote 62

Moreover, while extant definitions view subsidiarity as a decentralizing principle favouring allocating powers to the most ‘local’ unit possible, they may not always establish subsidiarity as decentralizing. The Treaty on European Union states that decisions should be made “as closely as possible to the citizen.”Footnote 63 This could make decentralization a necessary component of accounts of subsidiarity that seek to reflect EU law.Footnote 64 Yet subsidiarity may not decentralize if the burden for overcoming its presumption is low.Footnote 65 Many powers could be allocated to federal governments due to comparative efficiency. Indeed, subsidiarity centralizes power in existing countries’ legal practice (at least when applied to ‘residuary’ powers).Footnote 66

Subsidiarity’s purported status as a decentralizing principle likely stems from frequent ‘justifications’ based on purportedly decentralizing (sub-)principles. Arguments appealing to those most affected by an issue/subject, democratic concerns, and/or unique local interests or knowledge each purport to justify decentralization.Footnote 67 The ‘most affected’ principle is challenged for being too vague (viz., it is often unclear who is most affected by what issues, let alone subjects),Footnote 68 but its emphasis on stakeholder status appears important.Footnote 69 Stakeholder status also grounds democratic arguments for decentralization and subsidiarity:Footnote 70 local self-governance purportedly ensures the greatest representation possible in decisions affecting one (e.g., under subsidiarity, decisions are made by the group in which one’s vote has its greatest impact).Footnote 71

Protecting unique local interests is likewise proffered to justify decentralization and subsidiarity. Subsidiarity is sometimes discussed as an efficient way to sort interests.Footnote 72 It may best protect unique interests.Footnote 73 The EU version is purportedly justified by a need to protect local interests and the defeasible presumption that ‘local’ groups best protect them.Footnote 74 Arguments for decentralization speaking to the unique interests and challenges of local entities or epistemic value of local knowledge could also support subsidiarity if the ‘closest’ groups can be expected to protect the interests or are epistemically well-positioned to address given issues/subjects.Footnote 75

Justifications for non-EU-based conceptions of subsidiarity likewise appeal to decentralization. Two ‘liberty’-based views are representative.Footnote 76 The ‘confederalist’ view highlights the need for sub-state groups to whom citizens remain loyal as ‘bulwarks’ against tyranny.Footnote 77 Loyalty to a sub-state group of a sufficient size counter-balances central power, justifying local control to protect against unchecked central authority.Footnote 78 The existence of too many groups will leave no group strong enough to check power, requiring allocation only to some groups, but a subsidiarity principle limited to a small number could protect against tyranny.Footnote 79 The ‘Calvinist/communitarian’ view alternatively posits that local groups have unique values that must be protected from state interference.Footnote 80 Finnis suggests that these communitarian goods are so important as to justify subsidiarity even if it generates inefficiencies.Footnote 81

Subsidiarity, then, admits several forms grounded in compelling (sub-)principles. Nearly every form produces many prima facie plausible results, explaining why, for example, military and foreign policy powers often belong to central/federal governments while zoning powers often belong or are delegated to ‘local’ units.Footnote 82 If one form best allocates powers, it best fulfills subsidiarity’s primary task. This will resolve the authority allocation problem and identify subsidiarity’s best conception. Unfortunately, as we will now see, each modern conception of subsidiarity leaves important allocative questions unaddressed or relies on other principles to justify their conclusions, undermining the case for subsidiarity itself as an allocative principle.

3. The Case Against Subsidiarity as a Genuine Allocative Principle

Subsidiarity is prima facie compelling and has broad support, but its appeal rests on underlying (sub-)principles such that “apparent consensus on it has been gained only by obfuscation.”Footnote 83 Subsidiarity shorn of its underlying justifications is a mere stipulation.Footnote 84 Using it as an allocative principle requires attending to its justifications and accepting its institutional implications. Yet existing conceptions underdetermine authority allocations. Even plausible allocative decisions that subsidiarity could generate rest on underlying principles that do the real allocative work. They may not make local control a presumptive norm and could even lead to centralization. Subsidiarity does produce centralization in practice. Subsidiarity, then, is not merely an incomplete allocative principle.Footnote 85 It is not a standalone allocative principle at all.

The first issue with subsidiarity as a standalone authority allocation principle relates to familiar boundary-based considerations. Subsidiarity does not provide univocal, consistently compelling understanding of ‘closest’/‘most local’/‘most affected.’ ‘Close’ and ‘local’ can be geographically-defined. Yet the plausibility of municipal control suggests that subsidiarity as a geographically-defined principle should presumptively favour cities.Footnote 86 Even if the principle need not extend further to neighbourhoods, which some contest,Footnote 87 this account struggles to explain the lack of municipal constitutional powers in most federations or the inconsistent history of delegation of authority to cities, which suggests that even historical decisions to provide (revocable) powers to cities are not viewed as even presumptively compelling.Footnote 88 An often-overturned ‘presumption’ favouring the ‘most’ local is barely a presumption. While one may argue that this is just evidence that existing allocations do not meet basic theoretical standards, the large explanatory gap remains problematic absent some heretofore-absent explanation of why present realities seem intuitively problematic grounded in subsidiary-related concerns.

‘Closeness’ is, moreover, difficult to specify absent a potentially undermotivated geographical stipulation.Footnote 89 This is a clear issue for accounts of subsidiarity that require group identities to track boundaries.Footnote 90 It is likely an issue for any account where allocative principles should be able to address real claims with some plausibility. Claims by non-territorial groups on self-determination-based grounds cannot be rendered implausible by geographical fiat. Even geographical boundaries can be split further. Maria Cahill’s “primary” groups possessing presumptive decisional authority include non-governmental associations, like boxing clubs.Footnote 91 While clubs need not have authority if there are other reasons to bar them from possessing it (e.g., illiberalism), concerns about how to specify the ‘local’ raise questions about whether a presumption should favour greater authority for all ‘local’ groups. Subsidiarity at least struggles to explain existing authority allocations on any account that identifies anything ‘closer’ than provinces as candidate authorities. Border-drawing concerns will often make it difficult to apply.

Appeals to the ‘most affected’ entity require greater specification of the domain of an issue or subject and may not present a uniformly compelling answer favouring any entity, let alone presumptively favour decentralization. This concern is most acute with respect to division of powers-based concerns regarding authority over subjects. It is hard to establish the entity most affected by immigration or healthcare.Footnote 92 COVID-19 demonstrated public health policies’ differential impacts, but the entity ‘most affected’ by public health law is non-obvious.Footnote 93 A most affected principle could select two entities. It may not even provide clean answers in otherwise-easy cases favouring local control.Footnote 94 Consider municipal highways/roads. Should commuters have more say on a related policy because they use roads (supporting provincial control) or less because they do not pay property taxes?Footnote 95 ‘Most affected’ principles underdetermine this point. This problem extends to authority over issues. Consider a provincially-owned/-operated park within city limits but most often used by suburban residents. Who is ‘most affected’ by decisions to close that park to protect public health? Subsidiarity alone provides no answer.

These problems do not arise only because cities or sub-state nations have prima facie compelling claims. The claims’ prima facie plausibility establishes the need for a further allocative principle and difficulties addressing the claims are problematic, but the larger issue is that subsidiarity alone has difficulty explaining when it should apply. Perhaps cities, sub-state nations, et al. will never justifiably receive power, or one can justify only providing authority to central, federal, or provincial governments. Subsidiarity could be useful where the only entities recognized as even candidate authorities are federal and provincial governments (though this would tie subsidiarity’s plausibility to a substantive response to the ontological question and limit its application to a small number of cases, retaining the need for another allocative principle). Even then, difficulties identifying which powers must belong to more ‘local’ groups and when and how the presumption can be overcome would remain. Whether subsidiarity as such can plausibly set the relevant standards is at best debatable (as further demonstrated below).

Even stable meanings of ‘closest’/‘most local’/‘most affected’ do not uniformly favour subsidiary entities as authorities over discrete decisions, let alone constitutional powers. It is non-obvious that, for example, ‘local’ cities should presumptively possess many powers. Arguments for municipal control over immigration and healthcare powers remain controversial. It is difficult to accept that the burden is on others to disprove them. Similar problems occur even where we only analyze federal and provincial governments. Whether national or local cultures are most affected by immigration decisions is debatable.Footnote 96 Mere stipulations that, for example, provinces should control healthcare policy are difficult to justify, especially given distributional inequities across provinces who possess it.Footnote 97 While provincial control may be best in both cases, a stipulated presumption cannot justify provincial control. Treating subsidiarity as a principle for allocating authority over issues, rather than subjects, cannot avoid these concerns. For instance, even greater knowledge of where people are most likely to want to get tested for an illness during a pandemic may not make cities even epistemically-best-positioned to place them.

Even if one treats subsidiarity’s purported justificatory sub-principles as fundamentally decentralizing, they may not uniformly allocate powers to the most local entity. Sub-principles include a most affected principle, democratic principles, principles stressing the import of unique local interests, challenges, or knowledge, and/or principles appealing to the need for bulwarks against tyranny. Many cannot identify which ‘local’ entity should possess which powers. Parallel arguments for provinces and cities should lead to more municipal powers on a subsidiarity-based view favouring most local control.Footnote 98 Yet subsidiarity’s justificatory sub-principles may not produce this result. Even appeals to combinations of sub-principles could produce justificatory ties. For instance, both entities may secure equal ‘loyalty’ from members and be equally knowledgeable about the implications of a decision related to whether the parks above should be closed. Considerations that could break ties, like institutional capacity, are not always available. Available considerations likely favour provincial control, defeating presumptions favouring control by the most ‘local’ city, and again suggesting that subsidiarity is at best a strong candidate authority for allocating powers between federal and provincial governments alone. Even then, the decisions are likely only plausible regarding issues, not subjects. Strong cases for ‘local’ control over subjects, like land management or mining, will (at best) prove rare.

Issues like those undermining ‘most affected’ principles above also apply to ‘democratic’ and ‘local interest’ arguments.Footnote 99 Necessarily brief comments on each motivate general problems.

Some democratic concerns stem from boundary problems. It is difficult to determine the scope of subjects or most affected parties, making it hard to identify who should have what level of representation in decision-making procedures. Other concerns stem from democratic approaches’ lack of clear guidance on which powers should belong to which entities and/or how they provide intuitively problematic guidance. Proponents of municipal authority highlight how cities face unique challenges (re: diversity, homelessness, criminality, etc.) and are uniquely affected by existing federal policies (re: migration, etc.); their lack of decision-making authority in responses to those challenges and/or development of policies is worrisome.Footnote 100 Yet cities and sub-state nations are uniquely affected by many considerations. Cities face unique issues due to traditional features, like population density, and to recent developments, like greater immigration into cities (and, then, development of local cultural groups) and demographic ‘sorting’ whereby cities tend to be more ‘liberal’ than rural communities.Footnote 101 Their lack of authority over related matters arguably raises democratic issues, which can be acute where cities are more liberal and institutional considerations lead to more conservative federal and provincial governance.Footnote 102 If demographic sorting leads cities to consistently ‘lose’ on matters that uniquely impact them, cities likely lack adequate representation on the matters that most impact them.Footnote 103 Parallel arguments apply to sub-state national groups.Footnote 104 Yet democratic appeals to subsidiarity struggle to explain which powers must belong to cities or sub-state nations and likely cannot provide intuitively compelling results. Whether ‘local’ groups’ experiences should require municipal control over immigration policy is at best debatable.Footnote 105 If one accepts that result, problem cases remain. For instance, higher crime rates in cities suggest that criminal law ‘most affects’ city-dwellers.Footnote 106 A presumption favouring primary municipal authority over crime is uncompelling.

The same issues arise with appeals to local interests, challenges, or values divorced from democratic concerns. These principles provide some guidance on when cities or sub-state nations should possess particular powers. If, for example, one is best-positioned to protect a local culture, providing it with powers over cultural policy is intuitively compelling. Yet the scope of necessary cultural protections is difficult to parse. A presumption that it should, for example, control all education policy is hard to justify on local interests alone given education’s broad societal impact. Moreover, the range of unique interests, challenges, and values is wide enough to again generate justificatory ties between candidates. Entities could, for example, have equal self-determination-based claims for controlling language policy. A self-determination-based approach to subsidiarity would then be at best an incomplete allocative principle. Finally, even if one could identify which entities have unique interests, challenges, or values, it is hard to know whether and when control over which subjects is necessary to complete them. This could change over time.Footnote 107 These ‘justificatory’ (sub-)principles too at best allocate authority over issues.

Denying that powers should be provided to cities or sub-state nations cannot avoid these concerns. For instance, Loren King discusses parallel arguments for provincial and municipal authority but argues that subsidiarity should not apply to cities.Footnote 108 Yet an issue remains even if we grant the substantive ontological account generating that result: King jettisons the presumption favouring most local control and limits subsidiarity’s application more narrowly than one would desire of a complete allocative principle. Subsidiarity alone cannot explain why cities and sub-state nations’ prima facie compelling claims should not be realized if it is justified using features that establish those claims. Another allocative principle remains necessary.Footnote 109

Epistemic arguments better guide when ‘local’ entities should make decisions but still provide incomplete guidance in division of powers cases. What qualifies as knowledge is contentious and whether more ‘local’ entities consistently have better knowledge in a way that generates a presumption favouring local decision-making is questionable—local governments may often lack knowledge of a decision’s position in a larger decision-making context and impact elsewhere—but there is, at least, a decision procedure for identifying when local entities should make decisions on epistemic approaches. I cannot, however, accept that local entities generally have more knowledge about policy areas. I do not even know how to judge such claims about epistemic authority over subjects. Even epistemic approaches thus have limited scope. Moreover, epistemic positions may not sustain over time. Even if cities are best-positioned to decide where to place testing units for a particular virus on a particular date, they are not obviously best-positioned to handle public health policy generally and their knowledge may not sustain and warrant long-term delegation over narrow health facilities placement issues.Footnote 110

‘Bulwarks’-based arguments explain why most powers should belong to provinces but may not do so in a uniform manner and limit subsidiarity’s scope of application more than proponents desire. They hold that non-central/federal powers must be allocated to groups that sustain loyalty over time and do not eliminate concurrent loyalty to the country or splinter authority among so many groups so as to leave none with enough power to be bulwarks.Footnote 111 While some think these conditions only apply to provinces,Footnote 112 empirical evidence suggests people maintain loyalty to cities and sub-state nations over time and can do so while maintaining commitments to a ‘state’ identity.Footnote 113 Extending authority to many entities may limit any entity’s ability to do so, but three or four levels of governance, including municipal or ‘sub-state national’ levels, could act as bulwarks.Footnote 114 Questions about how to allocate powers between provinces and cities, for example, thus remain. Resolutions may not favour ‘most local’ control. Moreover, bulwarks-based arguments alone do not address which powers ‘lower’ authorities require to be bulwarks.Footnote 115 Even if we could resolve those problems, in turn, the idea that our allocative principle would only apply to central, federal, and provincial governments would still leave it with an undesirably narrow scope, retaining the need for another principle.Footnote 116

Examining each possible justification, then, challenges any case for subsidiarity as a genuine allocative principle. Any presumption favouring local control is likely only plausible regarding issues, not subjects. This is especially bad for the efficiency-based view, which is supposed on contingent realities and needs constant updating.Footnote 117 But the problem generalizes: No subsidiarity principle appears to provide uniformity on decisions over subjects, leaving questions about, for example, immigration and healthcare reform, or what to do in constitutional moments where subsidiarity is supposed to be most compelling, unresolved. This claim does not make the “mistake” of viewing subsidiarity as mere efficiency.Footnote 118 It just recognizes subsidiarity’s limits.

These considerations would be less forceful if subsidiarity consistently guided judicial decision-making in a compelling manner, but problems recur when applying the principle. One may contend that nothing in the forgoing undermines judicial use of the principle for authority allocation purposes.Footnote 119 Judicial application of the principle could vindicate its utility in concrete cases. My claim is that subsidiarity fails to create a plausible theoretical presumption favouring local control, but the point underlying this critique is fair: empirical data on the use of the principle is important on any plausible view. However, judicial use of subsidiarity does not consistently support its real-world plausibility (if it does so at all). To wit, European law heavily relies on national parliaments to enforce the subsidiarity principle through political pressure and European judicial treatments of subsidiarity amalgamate points from multiple distinct intellectual traditions, each of which raises the analytic problems undergirding critiques above.Footnote 120

While domestic courts invoke the principle elsewhere, in turn, their results are hardly inspiring. Canadian jurisprudence is instructive. The Constitution of Canada does not contain a subsidiarity principle.Footnote 121 This alone raises a challenge for many countries’ use of subsidiarity: it is often unclear whether and when judges can validly invoke it.Footnote 122 Yet scholars suggest that the federal government of Canada’s power to legislate with respect to issues of ‘national concern’ should be understood as reflecting a deeper constitutional concern with subsidiarity and that subsidiarity itself is recognized as a valid interpretive principle in Canadian constitutional law.Footnote 123 Extant jurisprudence admittedly provides some support for both claims. However, the same case law suggests that subsidiarity is not an action-guiding principle for resolving jurisdictional disputes that can provide intuitively compelling results favouring presumptive local control.

With respect to issues of national concern, Canada’s federal government possesses a power to legislate to ensure Canadian “Peace, Order, and good Government” (POGG); this is a ‘residuary’ power to legislate only where there is a gap in jurisdictions (e.g., neither federal nor provincial governments can explicitly legislate in the domain), in cases of national emergencies, and over issues of national concern.Footnote 124 Subsidiarity proponents suggest that this framework operates in a subsidiarity-like manner since the federal power only exists where provincial authority proves inadequate to the task at hand.Footnote 125 Yet even if we accept that contentious reading, we will now see that it does not appear to support practical uses of the subsidiarity principle.Footnote 126

Canadian judges have long struggled to develop a clear test for identifying matters of national concern and recent legal developments do not support the use of subsidiarity-like principles.Footnote 127 Historical tests were “difficult to apply.”Footnote 128 The most recent re-articulation of the test, in the References re Greenhouse Gas Pollution Pricing Act may not clarify matters, maintaining the basic problem.Footnote 129 Under that test, the national concern branch of POGG can be validly invoked where there is (i) a “matter is of sufficient concern to the country as a whole to warrant consideration as a possible matter of national concern” that has (ii) “singleness, distinctiveness, and indivisibility” (SDI) that distinguishes it from provincial concernsFootnote 130 and (iii) a scale of impact on provincial areas that is reconcilable with the fundamental division of powers.Footnote 131 Item (ii) requires that the legislation concern “a specific and identifiable matter that is qualitatively different from matters of provincial concern” and evidence that “establishes provincial inability to deal with the matter.”Footnote 132 Provincial inability refers to whether a “matter is of a nature that the provinces cannot address either jointly or severally, because the failure of one or more provinces to cooperate would prevent the other provinces from successfully addressing it, and if a province’s failure to deal with the matter within its own borders would have grave extraprovincial consequences.”Footnote 133 Yet this test may not be easier to apply than its predecessors. Indeed, it appears to maintain many problems with historical tests.

SDI issues, for example, remain difficult to identify under the new test. Historically, SDI was supposed to be identifiable through common law reasoning, but past judgments were unhelpful. Why “the control of marine pollution dumping of substances” counts, particularly (and perhaps exclusively) when limited to a particular kind of body of water like saltwater,Footnote 134 but laws addressing pollution in a river “not confined to a narrow range of toxic chemical substances … that have a severely harmful effect on human health and the environment whose pollutant effects are diffuse and persist in the environment”Footnote 135 does not remains unclear. ‘Is this like saltwater or a river?’ is more like a gnomic riddle than a legal test. New references to the “extraprovincial and international” nature of many SDI matters may not solve our riddle as well as the majority claims; the statement that international agreements “may in some cases indicate qualitative difference” is not terribly action-guiding.Footnote 136 Moreover, new statements on provincial inability may not fully resolve questions about which of “political incapacity, political unwillingness, or legal inability”Footnote 137 suffices to establish true provincial ‘inability,’ let alone which of these should suffice; the reference includes material supporting multiple interpretations of this criterion. The majority judgment also does not resolve historical questions about the meaning of ‘reconcilability.’Footnote 138 It only offers two paragraphs on how to understand that final stage.Footnote 139

There are, then, reasons to question whether Canadian national concern doctrine is properly action-guiding. The result in References re Greenhouse Gas further highlights how the branch itself does not reflect a meaningful presumption of local control. That case concerned a federal emission standard. All parties agreed that the provinces could set their own standards. Several chose not to do so. There was accordingly no sense in which the provinces were literally unable to set the relevant standards. However, the Supreme Court of Canada permitted strong incursions into provincial jurisdiction in the name of national concerns. While an Alberta Court of Appeal decision appealed to the subsidiarity principle to highlight the constitutional problems with the act in question,Footnote 140 the Supreme Court of Canada did not even directly engage with subsidiarity arguments.Footnote 141 It instead permitted strong incursions into admitted areas of provincial control, which combined with federal paramountcy rules to aid centralization.Footnote 142 Any purported presumption favouring local control is, then, apparently very weak. I doubt its existence.

Subsidiarity as a standalone interpretive principle does not fare much better in Canadian constitutional law.Footnote 143 Many references to the principle that scholars use to highlight its importance were mere obiter and did not explain how the principle should apply.Footnote 144 Even the most famous case referencing the principle, which upheld a municipal anti-pesticide law while highlighting the need for tri-level governance, was decided on the grounds that the municipal law fell under the terms of a valid provincial law, rather than the importance of local control as such.Footnote 145 The first detailed analysis of the subsidiarity principle in Canada was, in turn, part of a 4-4 split with minimal force; it again relies on provincial authority for minimal municipal powers and explicitly states that subsidiarity cannot alter the division of constitutional powers.Footnote 146

More recent attempts to invoke subsidiarity to save local laws largely falter.Footnote 147 Appealing to subsidiarity did not, for example, save local environmental standards that would impede a federal undertaking (viz., a pipeline), undermining the suggestion that a concern with environmental regulation best explains References re Greenhouse Gas: centralization seemingly trumps environmental concerns in practice.Footnote 148 While scholars also argue that democracy- and local interest-based subsidiarity arguments should require that provinces defer to municipalities on the drawing of electoral boundaries, the Supreme Court of Canada recently held that provinces can unilaterally change districts mid-election.Footnote 149 Subsidiarity can bolster some attempts to save bylaws in rare cases,Footnote 150 but the case law does not consistently favour local control. It also may not track intuitions on appropriate local control given its limits on provincial and municipal powers.

If one takes a more sanguine view on the theoretical and jurisdictional issues above, subsidiarity is still at best able to address delegation-based and divisions of powers concerns on a question-by-question basis, leaving large holes in subsidiarity’s ability to serve its intended allocative function. But even a uniform ‘presumption’ in discrete decision-making contexts is unlikely to hold. It is not clear that we should presumptively defer to cities on questions about where to place testing centres within municipal boundaries or when to close a park. Principles that could justify municipal control, including democratic and local control-based arguments, may equally favour provinces. Whether any tiebreaker principles could uniformly establish a presumption favouring subsidiarity as a basic allocative norm is at best unknown. Appeals to those ‘most affected,’ democratic concerns, or local interests or knowledge are unlikely to frequently allocate to the ‘closest’ level of governance if it is defined territorially—for example, most stakeholders will seek to move many powers away from cities—and such appeals to those ‘most affected’ or territoriality likely beg important questions anyway. Attempts to limit the scope of the analysis to central, federal, and provincial governments may not avoid this result and would limit subsidiarity’s scope of application in a way that still requires another allocative principle.

Where, in turn, ‘justifications’ for subsidiarity favour a ‘local control’ presumption, other (sub-)principles do the allocative work, making subsidiarity (at best) a mechanism for weighing other concerns. If, for example, communitarian ties justify presuming ‘local’ control over particular issues and identifies areas—for example, value-laden areas that are unnecessary for wider state stability, possibly including cultural policy, healthcare, or (more controversially) immigration powers—as best belonging to cities or sub-state nations, the community values principle guides allocation, not subsidiarity. Likewise, appeals to subsidiarity do no extra work where analytically severable democratic principles plausibly justify general allocative decisions.Footnote 151 If we are interested in democratic considerations, protecting local interests, leveraging special local knowledge, or bulwarks, we should appeal to those considerations to justify claims and weigh their value against other interests. Appealing to subsidiarity adds nothing to this weighing process.

The issue, then, is not merely that subsidiarity relies on other principles for its purported value. Rather, appeals to ‘subsidiarity’ either lead to implausible conclusions by the light of those very underlying principles or add nothing distinct to the moral decision-making process. Appeals to subsidiarity may even obscure relevant moral considerations in practice. The Supreme Court of Canada did not need to engage with subsidiarity directly in a municipal bylaw case since subsidiarity is only an interpretative principle.Footnote 152 Yet democracy is an even more fundamental principle of Canadian constitutional law.Footnote 153 One wonders if concerns about subsidiarity distracted from the important democratic concerns that may have been more important in the case. We cannot know for sure but even that possibility should be worrying.

Ultimately, then, subsidiarity is not a unique allocative principle and cannot serve its primary intended function. This does not mean that subsidiarity lacks any value but makes it less valuable and stresses the need to reflect on other allocative principles, such as the self-determination or epistemic value principles, to see if they play distinct allocative roles as claimed. Even if subsidiarity does not add meaningful content to discourse, it may, for example, have an important rhetorical role. For instance, some believe that subsidiarity’s inability to answer all allocative questions is unproblematic because invoking it occasions reflection on important moral concerns (e.g., the justificatory sub-principles above) and leads to more informed allocative decisions.Footnote 154 Appealing to subsidiarity might contribute to legal discourse on authority allocation, which could itself serve valuable ends, like building community in a polis and strengthening the legitimacy of decisions therein.Footnote 155 Simply adopting subsidiarity could also promote stability in the face of deep disagreement within countries.Footnote 156 Yet these cases for subsidiarity are practical (viz., non-theoretical), empirically contestable, and contingent. Moreover, ‘legal discourse’ must be ongoing to accrue the benefits. Ongoing constitutional debate is rare, again suggesting that subsidiarity will at best allocate authority over issues. Another allocative principle thus remains necessary on this best-case ‘discursive’ view. If one finds subsidiarity compelling, I recommend appealing to the underlying principles that make it seem compelling in a case instead. They provide more direct, normatively distinct guidance.

One still should not reject an otherwise valid claim in real-world cases simply because it is couched in subsidiarity language, but we should lament it if valid claims must be couched in that language to succeed. For instance, appeals to subsidiarity have strengthened Indigenous authority claims in Canada in the face of federal matrimonial property laws that would have otherwise applied, thereby serving a community-building role and helping local communities better protect individual rights than the federal government.Footnote 157 That case suggests that subsidiarity can be rhetorically valuable for fostering Indigenous self-governance and protecting universal rights in local settings. One of these underlying concerns may warrant appealing to subsidiarity despite its deficiencies as a standalone authority allocation principle capable of fulfilling its intended ends. I do not think we should remove a valuable tool from real-world battles for justice. Yet self-determination, anti-colonialism, and individual human rights protections could justify the Indigenous matrimonial property law absent an appeal to subsidiarity. It is a shame that the relevant communities needed to appeal to a problematic subsidiarity principle if their case for authority was already over-determined. I would prefer that direct appeals to the principles underlying subsidiarity’s seeming plausibility had more rhetorical force. This would permit more direct weighing of relevant moral interests. But public discourse need not mirror philosophy seminars.

4. Conclusion

Subsidiarity cannot serve one of its primary intended roles: allocating authority within countries. Its allocative potential is at best limited to cases where there is shared or no clear jurisdiction over issues, and relies on justificatory sub-principles. A ‘presumption’ favouring local control based on those principles likely cannot withstand scrutiny, and principles other than subsidiarity to do the allocative work when allocating authority to more ‘local’ entities is justified.

Testing subsidiarity’s potential as an allocative principle nonetheless furthers discussions in several literatures. Regarding subsidiarity, while the basic concept could be valuable even if it cannot identify all the entities that have prima facie compelling claims to some authority, the forgoing suggests that ‘subsidiarity’ must serve a function beyond guiding authority allocation in countries to be valuable. One may still argue that this merely creates a burden for future theories of subsidiarity to provide better allocative guidance. The way some conceptions of subsidiarity and underlying principles seemed more compelling than others suggests some paths are more likely to be fruitful and provides reason to prefer those conceptions. Yet even the best conceptions appear unable to serve as wide-scope allocative principles. For instance, while epistemic considerations may do a good job of allocating powers, questions remain about how to specify greater knowledge, and epistemic accounts appear to justify wider municipal powers than many are willing to accept. Limiting the principle to federal and provincial governments may better explain observed prima facie justified allocations but limits subsidiarity’s application more narrowly than proponents desire. Even then, epistemic value, not subsidiarity, is the guidepost.

Absent more compelling accounts of subsidiarity’s allocative value than those above, then, theorists should (surprisingly) no longer view an allocative function within countries as (one of) subsidiarity’s key moral role(s). If subsidiarity is to be more than a structuring framework for democratic decision-making, conceptions that fulfill other moral functions are preferable. Subsidiarity may, for instance, still be valuable as a means of allocating authority at the international level or explaining why associations, like guilds or boxing clubs, are ontologically valuable.Footnote 158 Conceptions capable of fulfilling these functions will be preferable.

Regarding authority allocation, the forgoing highlighted the need for a new allocative principle and desiderata thereof. The principles undergirding subsidiarity not only engender discussion,Footnote 159 but demonstrate that deferring to central/federal governments is also likely unwarranted as a general rule. Considerations explaining subsidiarity’s possible allocative value may not justify a presumption favouring control at the most ‘local’ level possible, but they are important moral concerns and likely equally defeat presumptions that uniformly favour central/federal governments. Practically, this finding provides reason to question central or federal governments who rarely delegate any authority and divisions of powers that recognize few provincial powers. It also undermines paramountcy, a principle under which federal laws always supersede provincial ones within predetermined areas of shared jurisdiction.Footnote 160 It is simply not obvious that federal decisions will always be ‘better’ in any or all relevant respects.

Finally, regarding method, the need for allocative principles to address real political claims—and ways subsidiarity (and now paramountcy) cannot uniformly and plausibly allocate authority within countries—stresses the need to examine whether other principles can supplement or replace subsidiarity and more plausibly allocate authority. An acceptable principle should guide each allocative decision above in a way that resolves the cases motivating this work, including conflict cases, and explain what justifies providing specific powers to specific entities (and when it does so). Perhaps no one principle can set general presumptions about how to justifiably, let alone ideally, allocate authority within countries. Decisions may need to be made through democratic processes.Footnote 161 Even then, we should seek a small set of principles for evaluating democratic decision-making. Allocations that appeal to features that made entities ontologically capable of making plausible authority claims benefit from simplicity. But separating ontological and allocative concerns permits necessary variation in the decisions.

The task moving forward is to identify whether any candidate allocative principles lack subsidiarity’s issues. If every principle faces issues like those above, the case against subsidiarity will admittedly have less purchase. One should not, however, assume that result ex ante. The preceding provided tools for assessing candidate principles and suggested that some which purportedly support subsidiarity are more plausible candidates than subsidiarity itself. The next step is to determine whether they and other principles can survive the kind of scrutiny above.

Acknowledgements

I would like to thank Geoffrey Sigalet and Daniel Weinstock for comments on previous versions of this paper; the Research Group on Constitutional Studies at McGill University for comments on a predecessor paper that inspired this one; and Hannah Da Silva for discussions on the issues. I would also like to thank my anonymous reviewers and Andrew Botterell for helpful guidance.

References

1. See Jonathan Rodden, Hamilton’s Paradox: The Promise and Peril of Fiscal Federalism (Cambridge University Press, 2006) for a federalism-specific ‘assignment problem.’ ‘Authority’ here does not perfectly mirror use in Raz, Dworkin, Wolff, Simmons, Green, et al. It does not focus on whether those allocated de facto ‘authority’ provide reasons or duties to obey them. Yet my approach fits traditional use in federalism studies and politics. Also, questions about the moral reasons justifying allocating decision-making ‘powers’ within states relate to those in Raz et al. Many principles that would make something a candidate for ‘powers’ here mirror those that would make it a classical ‘authority.’

2. Details appear in Section 3.

3. Any ‘modern’ conception of subsidiarity originates in EU law, including Treaty of Lisbon Amending the Treaty on European Union and the Treaty Establishing the European Community, European Council, 13 December 2007, C2007/C 306/01 [Treaty of Lisbon]. For subsidiarity’s use in Canada, see Andreas Føllesdal & Victor Muñiz-Fraticelli, “The Principle of Subsidiarity as a Constitutional Principle in the EU and Canada” (2015) 10:2 Ethics Forum 89; Hoi Kong, “Republicanism and the Division of Powers in Canada” (2014) 64:3 UTLJ 259; Hoi Kong “Subsidiarity, Republicanism, and the Division of Powers in Canada” (2015) 45:1-2 RDUS 13; Ran Hirschl, City, State: Constitutionalism and the Megacity (Oxford University Press, 2020) at 223. For use in the US, see Andreas Føllesdal, “The Principle of Subsidiarity as a Constitutional Principle in International Law” (2013) 2:1 Global Constitutionalism 37 at 37. For use in international law, see Paolo G Carozza, “Subsidiarity as a Structural Principle of International Human Rights Law” (2003) 97:1 AJIL 38 [Carozza, “IHRL”]; Samantha Besson, “Subsidiarity in International Human Rights Law—What is Subsidiary about Human Rights?” (2016) 61:1 Am J Juris 69 (noting critiques in Paolo G Carozza, “The Problematic Applicability of Subsidiarity to International Law and Institutions” (2016) 61 Am J Juris 51).

4. See Andreas Føllesdal, “Survey Article: Subsidiarity” (1998) 6:2 J of Political Philosophy 190 at 200-03 (on ‘Althusian’ accounts); Loren King “Cities, Subsidiarity, and Federalism” in James E Fleming & Jacob T Levy, eds, Federalism and Subsidiarity (New York University Press, 2014) 291; Maria Cahill, “Theorizing Subsidiarity: Towards an Ontology-Sensitive Approach” (2017) 15:1 Intl J of Constitutional Law 201; John Finnis, “Subsidiarity’s Roots and History: Some Observations” (2016) 61:1 Am J Juris 134. On Cahill’s account, Althusius presents a version connected to ‘sphere sovereignty,’ the view on which institutions (federal governments, churches, etc.) have natural domains of authority; subsidiarity is valuable insofar as it helps promote that natural order. Vischer views Althusius as liberty-focused. See Robert K Vischer, “Subsidiarity as a Principle of Governance: Beyond Devolution” (2001) 35 Ind L Rev 142.

5. Scholarly works increasingly recognize that other entities could possess powers and stress that some allocative principles, including subsidiarity, can be rivals to federalism. See generally Fleming & Levy, supra note 4; NW Barber, The Principles of Constitutionalism (Oxford University Press, 2018). The authority allocation problem also figures in debates about the separation of powers. The somewhat-simplified generalization nonetheless reflects (at worst) many relevant scholars’ primary focus. Even Fleming & Levy self-identifies as philosophy of federalism. Placement of many works in law journals is also notable. Many leading philosophical works specifically focus on federalism; see e.g. Dimitrios Karmis & Wayne Norman, eds, Theories of Federalism: A Reader (Palgrave, 2005); Wayne Norman, Negotiating Nationalism: Nation-building, Federalism and Secession in the Multinational State (Oxford University Press, 2006) [Norman, Negotiating Nationalism]. Allocative principles appear under the ‘federalism’ banner in Andreas Føllesdal, “Federalism” in Edward N Zalta, ed, The Stanford Encyclopedia of Philosophy, Summer 2018 ed, online: https://plato.stanford.edu/entries/federalism/

6. See theoretical arguments for providing powers to cities/sub-state nations in Sections 2 and 3. For comparative scholarship on different power-sharing forms, see e.g. Michael Burgess & John Pinder, eds, Multinational Federations (Routledge, 2011); Alan Fenna & Thomas O Hueglin, Comparative Federalism: A Systematic Inquiry, 2d ed (University of Toronto Press, 2015); Francesco Palermo & Karl Kössler, Comparative Federalism: Constitutional Arrangements and Case Law (Hart, 2017); Nicholas Aroney & John Kincaid, eds, Courts in Federal Countries: Federalists or Unitarists? (University of Toronto Press, 2017).

7. The dates of publication of John Rawls, A Theory of Justice (Harvard University Press, 1971), Will Kymlicka, Multicultural Citizenship: A Liberal Theory of Minority Rights (Oxford University Press, 1995), and Karmis & Norman, supra note 5 remain remarkable (though some works between them addressed some sub-state allocative concerns).

8. Recalling earlier caveats about simplification, note for example the locations of many publications (and homes of many authors) in sources cited herein. The major ‘federalism’ journal, Publius, publishes philosophy but more often features work in other fields. Texts cited throughout this work (or entries therein) support these generalizations.

9. See Hoi Kong, “Toward a Federal Legal Theory of the City” (2012) 57:3 McGill LJ 473 at 476. Kong thus views these issues against a backdrop of constitutional realities. Yet even works focused on such realities suggest expanding accounts of allocation to consider municipal and/or sub-state national claims. For example, Heather Gerken stresses the importance of addressing a broader array of claims to, and seeming exercises of, authority. See Heather Gerken, “Dissenting by Deciding” (2005) 57:6 Stan L Rev 1745; Heather Gerken, “Foreword: Federalism All the Way Down” (2010) 124:1 Harv L Rev 6; Heather Gerken, “Federalism as the New Nationalism: An Overview (2014) 123:6 Yale LJ 1889.

10. See Nilanjana Roy, “Cities Offer Sanctuary Against the Insularity of Nationalism”, The Financial Times (5 April 2017), online: https://www.ft.com/content/b54093f0-191f-11e7-9c35-0dd2cb31823a. While these claims often appear in the press, sanctuary cities appear as tools for protecting against ‘Trumpism.’ See Ilya Somin, “Making Federalism Great Again: How the Trump Administration’s Attack on Sanctuary Cities Unintentionally Strengthened Judicial Protection for State Autonomy” (2019) 97 Tex L Rev 1247. On demographic sorting, see Jonathan Rodden, Why Cities Lose: The Deep Roots of the Urban-Rural Political Divide (Basic Books, 2019). For sorting-related proposals, see Benjamin R Barber, If Mayors Ruled the World: Dysfunctional Nations, Rising Cities (Yale University Press, 2013); Benjamin R Barber, Cool Cities: Urban Sovereignty and the Fix for Global Warming (Yale University Press, 2014); David Miller, Solved: How the World’s Great Cities Are Fixing the Climate Crisis (University of Toronto Press, 2020).

11. See Michael Da Silva, “COVID-19 and Health-Related Authority Allocation Puzzles” (2021) 30:1 Cambridge Quarterly Journal of Healthcare Ethics 25. There I suggest COVID-19 even led to ‘decentralization’ demands in France.

12. See Daniel Weinstock, “Cities and Federalism” in Fleming & Levy, supra note 4, 259 [Weinstock, “Cities”]; Daniel Weinstock, “Pour une philosophie politique de la ville” (2009) 63:1 Rue Descartes 63; Daniel Weinstock, “Self-Determination for (Some) Cities?” in Axel Gosseries & Yannick Vanderborght, eds, Arguing About Justice: Essays for Philippe Van Parijs (Presses universitaires de Louvain, 2011) 377.

13. Weinstock, “Cities”, supra note 12 at 265. ‘Decentralized’ here means ‘allocated to entities other than global/central/federal governments.’ Its ‘proper’ meaning is beyond my scope of inquiry but see Paolo Dardanelli, “De-centralization” in John Kincaid, ed, A Research Agenda for Federalism Studies (Edward Elgar, 2019) 106.

14. See Hirschl, supra note 3 at 219-229; Daniel A Bell & Avner de-Shalit, The Spirit of Cities: Why the Identity of a City Matters in a Global Age (Princeton University Press, 2011); Avner de-Shalit, Cities and Immigration: Political and Moral Dilemmas in the New Era of Migration (Oxford University Press, 2019). See also Rainer Bauböck, “Reinventing Urban Citizenship” (2003) 7:2 Citizenship Studies 139. Jurisdiction-specific arguments for municipal ‘authority’ include Ron Levi & Mariana Valverde, “Freedom of the City: Canadian Cities and the Quest for Governmental Status” (2006) 44:3 Osgoode Hall LJ 409; Roger Keil & Douglas Young, “A Charter for the People? A Research Note on the Debate About Municipal Autonomy in Toronto” (2003) 39:1 Urban Affairs Review 87; Alexandra Flynn, “Operative Subsidiarity and Municipal Authority: The Case of Toronto’s Ward Boundary Review” (2019) 56:2 Osgoode Hall LJ 271; Kong, supra note 3; Kong, supra note 9. King and Blake discuss the status of cities in global authority allocation; see Loren King & Michael Blake, “Global Cities, Global Justice?” (2018) 14:3 J of Global Ethics 332. Others notice the parity of reasoning but believe further principles make provinces better authority candidates. See King, supra note 4; Richard Briffault, “‘What About the “Ism”?’ Normative and Formal Concerns in Contemporary Federalism” (1994) 47:5 Vand L Rev 1303.

15. See Helder De Schutter, “Territoriality and Personality: Concepts and Normative Considerations” in Jean-François Grégoire & Michael Jewkes, eds, Recognition and Redistribution in Multinational Federations (Leuven University Press, 2015); Ephraim Nimni, ed, National Cultural Autonomy and its Contemporary Critics (Routledge, 2005); Ghislain Otis, “Territorialite, personnalite et gouvernance autochtone” (2006) 47:4 Cahiers de Droit 781; Ephraim Nimni et al, eds, The Challenge of Non-Territorial Autonomy: Theory and Practice (Peter Lang, 2013). See also John Croakley, “Approaches to the Resolution of Ethnic Conflict: The Strategy of Non-territorial Autonomy” (1994) 15:3 Intl Political Science Rev 297. Cf Rainer Bauböck, “Multinational Federalism: Territorial or Cultural Autonomy?” (2001) Austrian Academy of Sciences Working Paper No 15.

16. See e.g. Allen Buchanan, Justice, Legitimacy, and Self-Determination: Moral Foundations for International Law (Oxford University Press, 2006) (on self-determination); Barber, supra note 5 at 187; Føllesdal, supra note 4 at 212 (on subsidiarity). As discussed below, subsidiarity need not answer the ontological question to be morally valuable. Like many major philosophical terms, ‘ontological’ and ‘allocative’ have multiple meanings in different literatures. In the subsidiarity literature, one ‘ontological’ approach holds that subsidiarity can help protect the value of groups; see e.g. Cahill, supra note 4. Our ontological question asks what features of entities make them candidate authorities. The value of groups may answer that question, but my use of the term ‘ontological’ relates to, but is non-identical with, use in Cahill et al.

17. See generally e.g. Weinstock, supra note 12; Hirschl, supra note 3.

18. Many thus allow exceptions to general allocative rules for cultural policy. This dates to at least Kymlicka, supra note 7. It continues. For example, Anna Stilz argues that territorial control is necessary to justifiably exercise most ‘powers’ but (deferring to De Schutter) permits possible exceptions for cultural policy, education, and natural resource management. See Anna Stilz, Territorial Sovereignty: A Philosophical Exploration (Oxford University Press, 2019) at 116, 232. Allen Buchanan’s account also distinguishes our questions. See Allen Buchanan, “Uncoupling Secession from Nationalism and Intrastate Autonomy from Secession” in Hurst Hannum & Eileen F Babbitt, eds, Negotiating Self-Determination (Lexington Books, 2006) 81.

19. Indeed, while the International Covenant on Economic, Social and Cultural Rights provides self-determination rights to all persons, General Comment 3 clarifies that all rights do not require or preclude any form of government. See International Covenant on Economic, Social and Cultural Rights, 16 December 1966, 993 UNTS 3, art 1 [International Covenant]; General Comment 3: The Nature of States Parties’ Obligations, United Nations, Committee on Economic, Social and Cultural Rights, 5th Sess, 14 December 1990, UN Doc E/1991/23 at para 8.

20. Further to arguments surveyed in Buchanan, supra note 16, note that the legal right in the International Covenant, supra note 19, is itself subject to limitations in arts 4-5, and self-determination rights must be consistent with realization of other rights.

21. For a recent example, see Aimée Craft et al, “COVID-19 and First Nations’ Responses” in Colleen M Flood et al, eds, Vulnerable: The Law, Policy and Ethics of COVID-19 (University of Ottawa Press, 2020) 51. Indigenous claims may raise unique concerns, but many national minorities have claimed general rights to control social policy. See Daniel Béland & André Lecours, Nationalism and Social Policy (Oxford University Press, 2008).

22. See also Michael Da Silva, “Nations as Justified Sub-State Authorities” (2022) 28:3 Nations & Nationalism 806.

23. See e.g. supra note 19, supra note 20.

24. See King, supra note 4; Weinstock, “Cities”, supra note 12.

25. See generally supra note 14, supra note 15, supra note 21.

26. See comparative federalism works in e.g. supra note 6. See also Patricia Popelier & Bea Cantillon, “Bipolar Federalism and the Social Welfare State” (2013) 43:4 Publius 626, on the broader range of social policy power choices.

27. Work on this topic is minimal, but for one philosophical discussion of some choices see Douglas MacKay & Marion Danis, “Federalism and Responsibility for Healthcare” (2016) 30:1 Public Affairs Quarterly 1.

28. See Barber, supra note 5 at 188.

29. Nicholas W Barber, “The Limited Modesty of Subsidiarity” (2005) 11:3 Eur LJ 308 at 308.

30. Compare e.g. works by Buchanan, supra note 16; Buchanan, supra note 18; Kymlicka, supra note 7; Norman, Negotiating Nationalism, supra note 5 (and related legal texts).

31. For a strong epistemic argument, see Yann Allard-Tremblay, “Divide and Rule Better: On Subsidiarity, Legitimacy and the Epistemic Aim of Political Decision-Making” (2017) 34:5 J of Applied Philosophy 696.

32. Cf Føllesdal, supra note 4; Cahill, supra note 4. If it is “essentially contested,” a conception should still reflect use in restricted domains. WB Gallie, “Essentially Contested Concepts” (1956) 56 Proceedings of the Aristotelian Society 167 at 167. See ibid for analysis of essentially contested concepts.

33. See Treaty of Lisbon, supra note 3, art 5.3.

34. The authors in supra note 32 agree on this. See also Føllesdal & Muñiz-Fraticelli, supra note 3; Hirschl, supra note 3 at 219-29; Fleming & Levy, supra note 4; Barber, supra note 5; Barber, supra note 29. Even those who say it originated in EU law (see e.g. Levi & Valverde, supra note 14) or also view it as a transnational principle (see e.g. Yishai Blank, “Federalism, Subsidiarity, and the Role of Local Governments in an Age of Global Multilevel Governance” (2010) 37 Fordham Urb LJ 510) recognize that its domestic structuring role is now central.

35. Some believe it could answer ontological questions, structure international law, establish boundaries of transnational authority, constitute state boundaries, keep peace, or justify transfers of funds between entities. See Føllesdal, supra note 4 at 191; Barber, supra note 5; David Miller, “Boundaries, Democracy, and Territory” (2016) 61:1 Am J Juris 33; Joseph Drew & Masato Miyazaki, “Subsidiarity and the Moral Justification of Intergovernmental Equalization Grants to Decentralized Governments” (2020) 50:4 Publius 698. Private law uses are less common; see Peer Zumbansen, “Happy Spells?: Constructing and Deconstructing a Private-Law Perspective on Subsidiarity” (2016) 79:2 Law & Contemp Probs 215; Matt Campbell “Subsidiarity in Private Law?” (2020) 24:1 Ed L Rev 1.

36. For a discussion of sphere sovereignty, see supra note 4 and accompanying text.

37. See Barber, supra note 29. The EU’s ‘modern’ conception was introduced to allocate powers in an association without set legislative powers. Yet subsidiarity is most often invoked outside that context as a means of allocating authority (including legislative authority) within countries. See Føllesdal, supra note 4 at 191.

38. NW Barber & Richard Ekins, “Situating Subsidiarity” (2016) 61:1 Am J Juris 5 at 5.

39. Barber views the issues as ‘primary.’ See Barber, supra note 5 at 205. Scholars generally disagree on the ‘primary’ issue subsidiarity should address, but most agree that issues and subjects are both relevant.

40. Ibid at 191; Andreas Føllesdal, “Subsidiarity, Democracy, and Human Rights in the Constitutional Treaty of Europe” (2006) 37:1 J of Social Philosophy 61 at 64.

41. See supra notes 32-38.

42. See supra note 35.

43. See supra note 16.

44. See supra note 3. See also overviews in supra note 4, and the comparative treatises in supra note 6.

45. See e.g. Flynn, supra note 14 (on how it should frame provincial decision-making); Éléonore Gauthier, “Spending Power, Social Policy, and the Principle of Subsidiarity” (2017) 22:2 Rev Const Stud 261 (on how it should frame federal decision-making). Both suggest that even a non-justiciable subsidiarity principle could fulfill this role.

46. I thank an anonymous reviewer for pushing me on this point. The sources mentioned in supra notes 3, 4, and 6 also establish the use of subsidiarity for this purpose. I discuss examples of legal application in the EU and Canada in Section 3.

47. Føllesdal, supra note 4 at 190. See also Føllesdal, “Federalism”, supra note 5; Carozza, “IHRL”, supra note 3; Allard-Tremblay, supra note 31.

48. Hirschl, supra note 3 at 221.

49. Barber, supra note 5 at 191 [emphasis added].

50. Finnis, supra note 4 at 134. See Dominic Burbidge, “The Inherently Political Nature of Subsidiarity” (2017) 62:2 Am J Juris 143 for another virtue-theoretic take that cites Finnis. Cahill adds that alternatives assume an unsustainable geographical hierarchy. She thus discusses ‘primary’ and ‘subsidiary’ authorities, not ‘local’ and ‘central’ ones. See Cahill, supra note 4.

51. See Føllesdal, supra note 40 at 64.

52. Føllesdal, supra note 3 at 38.

53. Further to points in Barber and others above, Føllesdal argues that it is necessary to foster the ‘dual loyalty’ characteristic of federalism. See Føllesdal, supra note 40 at 63-4. See also Jacob T Levy, “Federalism, Liberalism, and the Separation of Loyalties” (2007) 101:3 American Political Science Review 459.

54. See Barber, supra note 5 at 187ff.

55. See e.g. Weinstock, “Cities”, supra note 12.

56. Cf ibid ; Barber, supra note 5 at 215; Canadian caselaw in sources cited at supra note 5.

57. See Allard-Tremblay, supra note 31 (also stressing the burden needed to overcome it); Trevor Latimer, “Against Subsidiarity” (2018) 26:3 J of Political Philosophy 282 (arguing against the EU-related version).

58. See Føllesdal, supra note 4; Føllesdal & Muñiz-Fraticelli, supra note 3.

59. Barber & Ekins, supra note 38 at 5, 8.

60. See Allard-Tremblay, supra note 31 at 697.

61. Føllesdal, supra note 4 at 193 [emphasis in original] (also noting that it can proscribe or require central action).

62. Ibid . Compare Canadian caselaw discussed in Føllesdal & Muñiz-Fraticelli, supra note 3, and detailed further in Section 3.

63. Treaty of Lisbon, supra note 3, art 8 A.3.

64. See e.g. Hirschl, supra note 3 (though he suggests ‘decentralization’ is itself a principle of efficiency and democratic control that justifies subsidiary, ibid at 222).

65. Daniel Weinstock thus suggests its ‘decentralizing’ tendency is a matter of presentation (in conversation).

66. See Føllesdal & Muñiz-Fraticelli, supra note 3 at 100 (admittedly only discussing cases where most ‘powers’ are already defined but a new issue does not fall under the ambit of existing specified powers). I will return to this point.

67. See Barber, supra note 5; Hirschl, supra note 3 at 222, also discussing other (e.g., ‘Republican’) accounts.

68. See Robert Goodin, “Enfranchising All Affected Interests, and Its Alternatives” (2007) 35:1 Philosophy & Public Affairs 40 at 68.

69. See Hirschl, supra note 3 at 220.

70. See Weinstock, “Cities”, supra note 12; King, supra note 4 (to some extent). Cf Føllesdal, supra note 4 at 198. Subsidiarity could present tensions with democracy; see Cahill, supra note 4 at 205-06; Føllesdal, supra note 40 at 69-70, 72-74. This is likely true of any allocative principle other than the democracy principle itself.

71. See Weinstock, “Cities”, supra note 12; Hirschl, supra note 3 at 219-20. Cf Trevor Latimer, “The Principle of Subsidiarity: A Democratic Reinterpretation” (2018) 25:4 Constellations 586.

72. See Føllesdal, supra note 4.

73. Allard-Tremblay hints at this possibility. See Allard-Tremblay, supra note 31.

74. Ibid . Føllesdal and Barber also read it this way. See Føllesdal, supra note 4; Barber, supra note 5 at 192-93.

75. See Hirschl, supra note 3; Weinstock, “Cities”, supra note 12; King, supra note 4 (to some extent).

76. See Føllesdal, supra notes 4, 5.

77. Ibid . Cf Barber, supra note 5.

78. See Føllesdal, supra note 40 at 63-64.

79. See Føllesdal, supra notes 4, 5, 40.

80. See Føllesdal, supra notes 4, 5, 40. See also Cahill, supra note 4; Finnis, supra note 4 (with the latter also discussing a Catholic conception).

81. Per Føllesdal, other historical conceptions, like the Catholic one, likewise speak to local interest protection. See Føllesdal, supra note 4. This helps explain why Cahill, supra note 4, denies the existence of a new ‘modern’ conception. Yet the conception found in modern constitutional, EU, and international authority at least purports to break from tradition and address a different question. The Catholic conception often addresses questions related to sphere sovereignty, which may not be identical to the authority issues addressed here—though, as discussed above and below, it may offer a better understanding of a unique role subsidiarity can play in moral theory. On the Catholic conception, see e.g. Finnis, supra note 4; Russell Hittinger, “The Coherence of the Four Basic Principles of Catholic Social Doctrine: An Interpretation” in Margaret S Archer & Pierpaolo Donati, eds, Pursuing the Common Good: How Solidarity and Subsidiarity Can Work Together (Pontifical Academy of Social Sciences, 2008) 75; Michelle Evans & Augusto Zimmermann, eds, Global Perspectives on Subsidiarity (Springer, 2014) at ch 1, 2, 3.

82. See e.g. comparative federalism texts, including those in supra note 6.

83. Føllesdal, supra note 4 at 190.

84. See Cahill, supra note 4 at 206, 212. Føllesdal & Muñiz-Fraticelli, supra note 3 at 102 make a similar claim.

85. This would be consistent with Levy, supra note 53; Allard-Tremblay, supra note 31.

86. See municipal authority-focused sources, including those in supra notes 13, 14.

87. See e.g. Jerry Frug, “Decentering Decentralization” (1993) 60:2 U Chicago L Rev 253.

88. Weinstock, “Cities”, supra note 12 is inspired by many states’ failure to provide powers to cities.

89. This explains Cahill’s point in Cahill, supra note 4, outlined in supra note 50.

90. See Miller, supra note 10.

91. Cahill, supra note 4 at 223-24.

92. Central/federal governments usually control immigration. Yet, per de-Shalit, supra note 14, most immigration is to cities, whose residents are ‘most affected’ by it. Similarly, see Da Silva, supra note 22.

93. See Da Silva, supra note 11.

94. Similar worries motivate Burbidge’s claim that subsidiarity is inherently ‘political’ and debates about its scope of application must be resolved through democratic processes. See Burbidge, supra note 50.

95. Cf Hirschl, supra note 3 at 220-21; Lior Glick, “Commuters, Located Life interests, and the City’s Demos” (2021) 29:4 J of Political Philosophy 480.

96. de-Shalit’s conclusion in Cities and Immigration can also be generated by noting different patterns across provinces. See de-Shalit, supra note 14.

97. See e.g. Jamila Michener, Fragmented Democracy (Cambridge University Press, 2018) (on the US).

98. See supra notes 14, 15.

99. See Latimer, supra note 57, suggesting that major arguments also face empirical challenges.

100. See Hirschl, supra note 3 at 224, building on Weinstock, “Cities”, supra note 12.

101. See Hirschl, supra note 3; Weinstock, “Cities”, supra note 12; de-Shalit, supra note 14.

102. See Rodden, supra note 10.

103. Ibid . One may argue that that municipal bodies should set the terms of their powers, but claim that is too strong. Moreover, as discussed later in this section, subsidiarity does not protect municipal control over democratic process in practice.

104. See generally supra note 15.

105. See supra note 92.

106. See Hirschl, supra note 3 at 220-21.

107. On contingencies, see also Latimer, supra note 57.

108. See King, supra note 4.

109. Indeed, King may just offer a further, practical allocative principle. See ibid .

110. I build on Da Silva, supra note 11 here.

111. See e.g. Somin, supra note 10; Levy, supra note 53.

112. See Levy, supra note 53; King, supra note 4.

113. See supra notes 14, 15.

114. On the possibility of three or four levels of governance, see e.g. Barber, supra note 5 at 194 n 24.

115. Ibid at 191 notes this problem occurs even before bringing in cities. Moreover, other arrangements may be better bulwarks; see Latimer, supra note 57 at 289-90.

116. Sophisticated variants also face issues. For example, Allard-Tremblay’s epistemic account provides powers to entities best-positioned to address ‘common’ goals. Conditions requiring ‘common actions’ specify when a presumption that the most ‘local’ entity can be defeated. Yet such goals are often unidentifiable. Allard-Tremblay’s value pluralism bars his stipulating them or what groups are most ‘local.’ Similar issues explain why he views subsidiarity’s benefits are largely discursive and does not focus on direct allocative guidance. See Allard-Tremblay, supra note 31.

117. See Føllesdal, supra note 4 at 207.

118. Barber & Ekins, supra note 38 at 11.

119. I thank an anonymous reviewer also for this point.

120. Føllesdal & Muñiz-Fraticelli discuss the procedure under which national parliaments monitor EU proposals and “give out a ‘yellow card’—if they think the decision violates subsidiarity.” Føllesdal & Muñiz-Fraticelli, supra note 3 at 91. They also note how EU law on this topic combines features of five different traditions. Føllesdal further highlights questions and provides an overview of debates about who should apply the European principle. He also discusses how interpretation of the principle has not resolved conflicts or addressed worries about centralization. See Føllesdal, supra note 40 at 66-68. See Besson, supra note 3 for similar (albeit admittedly non-identical) points in the international human rights law context.

121. See Constitution Act, 1982, s 35, being Schedule B to the Canada Act 1982 (UK), 1982, c 11. No principle appears in the constitutional text or the unwritten constitution. Even supra note 3 sources grant this.

122. See e.g. Salt Spring Island Local Trust Committee v B & B Ganges Marina Ltd , 2007 BCSC 892. It also follows Canada Western Bank v Alberta, [2007] 2 SCR 3 [Canada Western] in deciding on interjurisdictional immunity grounds.

123. See e.g. Føllesdal & Muñiz-Fraticelli, supra note 3 at 98-101.

124. Constitution Act, 1867, 30 & 31 Vict, c3, s 91, online: https://laws.justice.gc.ca/eng/Const/index.html. On POGG as a general residuary, see R v Hauser, [1979] 1 SCR 984 [Hauser]. AS Abel, “What Peace, Order and Good Government?” (1968) 7 West Ont L Rev 1 argues that the federal and provincial governments each possess residuary powers. K Lysyk, “Constitutional Reform and the Introductory Clause of Section 91: Residual and Emergency Law-Making Authority” (1979) 57 Can Bar Rev 531 at s II argues for a parallel “local and private matters” residuary for the provinces. William R Lederman, “Unity and Diversity in Canadian Federalism: Ideals and Methods of Moderation” (1975) 53 Can Bar Rev 597 argues that POGG is primary among federal powers and all enumerated federal powers are examples of this primary power, not unique powers requiring a residuary. Yet POGG remains the canonical residuary power in Canadian constitutional law and Hauser is good law. Only the ‘gap’ branch is truly residuary. The national concern and emergency branches do not address topics fully outside specified heads of power, but rather topics requiring special responses. ‘Emergency’ and ‘national concern’ measures need not fill a legislative gap in which no one can legislate, but can, in practice, be exercised concurrently with provincial powers when a subject plausibly falls under both a provincial head and one of these additional POGG branches.

125. See e.g. Føllesdal & Muñiz-Fraticelli, supra note 3.

126. See generally supra note 124 for why this reading is contentious.

127. On historical issues, see Peter W Hogg, Constitutional Law of Canada, 5th ed, (Carswell, 2007) ch 17.

128. Bernard W Funston & Eugene Meehan, Canada’s Constitutional Law in a Nutshell, 4th ed (Carswell, 2013) at 77, discussing classics like Hogg, supra note 127. See also Jean Leclair, “The Elusive Quest for the Quintessential ‘National Interest’” (2005) 38:2 UBC L Rev 353; Sujit Choudhry, “Recasting Social Canada: A Reconsideration of Federal Jurisdiction over Social Policy” (2002) 52:3 UTLJ 163.

129. 2021 SCC 11 [References re Greenhouse Gas].

130. R v Crown Zellerbach Canada Ltd, [1988] 1 SCR 401 at para 33 [Crown Zellerbach].

131. See References re Greenhouse Gas, supra note 129 at paras 162-65.

132. Ibid at para 164. As Choudhry, supra note 128 at 227-28 notes, the provincial inability lemma introduced in Crown Zellerbach has its origins in Dale Gibson, “Measuring National Dimensions” (1976) 7:1 Man LJ 15.

133. References re Greenhouse Gas, supra note 129 at para 157 [emphasis in original].

134. Ibid at para 353, citing Crown Zellerbach, supra note 130 at 436.

135. R v Hydro-Québec, [1997] 3 SCR 213 [Hydro-Québec]. The majority in References re Greenhouse Gas admittedly distinguishes its decision from Hydro-Québec, albeit in a manner that is likely to be controversial. See References re Greenhouse Gas, supra note 129 at 82.

136. References re Greenhouse Gas, supra note 129 at 148-49 [emphasis added].

137. Leclair, supra note 128 at 365. For a classic case suggesting unwillingness suffices, see Johannesson v Municipality of West St Paul, [1952] 1 SCR 292. Crown Zellerbach, supra note 130 suggests unwillingness or inability suffices. Yet Choudhry, supra note 128 at 233-47 notes that provincial inability could have several meanings and outlines three distinct approaches, including what he takes to be the original interpretation whereby one can claim inability whenever governance is necessary, creating a risk that provinces will “race to the bottom” of non-service provision. References re Greenhouse Gas, supra note 129 may not fully resolve the matter. References to the necessity for “collective national and international action” ( ibid at para 12) and consequent view that the issue cannot be addressed by the provinces by its very nature suggest literal provincial inability is what matters. Yet discussions of the “risk” of “non-cooperation” as key to the finding of provincial inability (ibid at para 195) and application of the test suggests that the possibility of provincial unwillingness is key under the new test. Justice Brown disapproves of this in his dissent (ibid at para 445).

138. See Leclair, supra note 128.

139. See References re Greenhouse Gas, supra note 129 at paras 160-61.

140. See Reference re Greenhouse Gas Pollution Pricing Act, 2020 ABCA 74 at paras 129, 137-42.

141. See References re Greenhouse Gas, supra note 129. It did cite some articles that engage with the principle.

142. Ibid permits federal regulations that the provinces could pass under their own powers.

143. For one of the best discussions of this case law, see Eugénie Brouillet, “Canadian Federalism and the Principle of Subsidiarity: Should We Open Pandora’s Box?” (2011) 54 SCLR 601. I update that discussion somewhat here.

144. Føllesdal & Muñiz-Fraticelli highlight Canada Western, supra note 122; 114957 Canada Ltée (Spraytech, Société d’arrosage) v Hudson (Town), 2011 SCC 40 [Spraytech]; and Reference re Assisted Human Reproduction Act, 2010 SCC 61 [Reference re Assisted Reproduction] as key examples of the use of the term. See Føllesdal & Muñiz-Fraticelli, supra note 3. But only Reference re Assisted Reproduction engages with the principle in detail. Gauthier, supra note 45 also emphasizes cases where use is largely obiter.

145. See Spraytech, supra note 144. Recall also how the cases in supra note 122 were decided on interjurisdictional immunity grounds.

146. See Reference re Assisted Reproduction, supra note 144.

147. See e.g. Rogers Communications Inc c Châteauguay (Ville), 2016 SCC 23; Canada Post Corp v Hamilton (City), 2016 ONCA 767.

148. See Reference re Environmental Management Act (British Columbia), 2019 BCCA 181.

149. See Toronto (City) v Ontario (Attorney General), 2021 SCC 34. The dissent in that case cites Flynn, supra note 14 at para 116.

150. See Alberta v Latouche, 2010 ABPC 166.

151. These cases too will be subject to limitations. For example, if providing authority to religious groups would violate liberal norms, Cahill’s approach could be worrisome. See Cahill, supra note 4. This does not uniquely promote subsidiarity. Any problem here is more easily addressed by taking liberal-democratic norms into account at the ontological stage.

152. See References re Greenhouse Gas, supra note 129.

153. See Reference re Secession of Quebec, [1998] 2 SCR 217.

154. See Føllesdal, supra note 40 at 65ff; Allard-Tremblay, supra note 31.

155. See Føllesdal & Muñiz-Fraticelli, supra note 3 at 97. Allard-Tremblay, supra note 31 says it can add legitimacy to decisions.

156. See Føllesdal & Muñiz-Fraticelli, supra note 3 at 98. Barber thus supports a federal supervisory ‘backstop.’ See Barber, supra note 5 at 197.

157. See Christopher Alcantara, “Aboriginal Policy Reform and the Subsidiarity Principle: A Case Study of the Division of Matrimonial Real Property on Canadian Indian Reserves” (2008) 51:2 Canadian Public Administration 317.

158. Supra note 35 catalogues other functions.

159. See Føllesdal, supra note 4; Allard-Tremblay, supra note 31.

160. See e.g. Alberta (Attorney General) v Moloney, [2015] 3 SCR 327.

161. See Burbidge, supra note 50.