Federalism as an institutional doctrine

Federalism is, minimally, a method of allocating final decision-making authority over subjects (e.g., crime, healthcare, and immigration) in a governance unit (e.g., country). Faced with questions of the form “who can decide what when,” federal bodies, like the United States., Canada, Australia, and Germany, provide at least two entities (federal governments, provinces, cities, etc.) with final decision-making “powers” over at least one subject. No other entity is morally permitted to directly interfere (substitute decisions, fine, etc.) with the authority's decision-making regarding those subject(s). This is distinct from the unitary or centralized governance of, for example, France and Israel whereby one central entity possesses all final decision-making authority.¹ Beyond these basics, the meaning of and criteria for evaluating claims about federalism remain contested in law and political science.² The broader debates are then oft-ignored in mainstream political philosophy, resulting in conceptual confusion with important practical results discussed below.³

The following argues for adopting an “institutional” approach to federalism, rather than more common “ideological” approaches.⁴ A long tradition equates federalism with the US Founding Fathers' institutional proposals (Publius, 1788/1987). Yet, partly due to empirical developments, the dominant account outside philosophy now holds that federalism is a normative doctrine promoting a secure political organization combining “shared[-]rule and self-rule” and separates this federal “idea” from institutional forms that may realize it (Elazar, 1987; Watts, 2008). Philosophers often begin by accepting this ideological approach (as I discuss further in Da Silva, 2022).⁵ Popelier (2021, p. 33) even suggests “all” scholars view this combination as federalism's “normative core.” But philosophical strictures and practical realities demand a more circumscribed approach. The dominant ideological approach is too broad to be a distinct normative doctrine or cannot even apply to many paradigmatic federal bodies. Institutional approaches defining federalism by advocacy for the adoption of federations (defined below) for authority allocation are preferable.

To establish this, I first detail and defend criteria for evaluating philosophical accounts of federalism. I then elaborate the distinction between ideological and institutional approaches and apply the criteria to the dominant ideological approach and a new specification of an institutional approach inspired by Wheare (1946/1953, p. 11)'s classic, oft-critiqued account. I thereby demonstrate that ideological accounts fail to fulfill many normative adequacy criteria for a philosophical account of federalism and one can articulate a more action-guiding institutional account that avoids common critiques. I finally explain how this exemplary application grounds a broader case for institutional approaches to federalism and address pressing objections. The result is deceptively radical as it weakens a decades-old stark contrast between “federalism” and “federation.” The broader arguments challenge several orthodox positions in the study of federalism. However, the balance of reasons supports an institutional approach to federalism.

My arguments are conceptual and normative. I aim to clarify the concept of federalism in a way that maintains it as a distinct, action-guiding contribution to moral ontology. Rather than offer an alternative to leading accounts of why one should adopt federalism (e.g., Bellamy, 1996; Levy, 2007; Norman, 2006), I identify and specify the institutional normative concept they may support. I do not seek to defend all federations over other forms of governance. I instead demonstrate that advocacy for that institutional form is a normative doctrine one can justifiably adopt, albeit one with fewer necessary conceptual commitments than many claim. The concept is sufficiently distinct and compelling to warrant closer scrutiny if it does not best fit the “federalism” description. The concept matters, not the word. But findings below explain why one would adopt the institutional concept, why it deserves the label “federalism,” and other commitments a federalist should make. They thereby provide ample guidance for ongoing debates about the merits of federalism vis-à-vis other systems. The argument for institutionalism succeeds even if my specification is problematic. If, in turn, the general argument for institutionalism (rather than my neo-Whearean specification) fails, my analysis at least identifies burdens ideological accounts must meet and demonstrates that the dominant example does not meet them.

Existing practices and scholarship provide criteria for judging theories of federalism. A complete theory should be descriptively adequate. It should also serve federalism's professed normative functions, including guiding authority allocation within countries and judicial interpretation. It should do so in a way that is distinct, normatively compelling and explains its relationship to adjacent concepts. It should thereby assist the development of comparative law and politics and philosophical analysis of related concepts, like democracy. These are plausible conceptual desiderata even if no theory fulfills them all and can be evaluative criteria for comparing views.

Federalism is, in short, a normative ideal with numerous institutional forms. Federations are just one type of federal system that instantiate the combination of shared- and self-rule characteristics of the concept. Elazar developed this definition for the purposes of comparative political science and offers much more detail on which institutional forms plausibly qualify as “federal.” As discussed below, some institutional design features guide Elazar's own view. But Elazar is explicit in stating that all “federalist rather than monist or centralist” approaches to political life are connected by “some conception of the federal idea, some persuasive or ideology that endorsed federal solutions, some particular application of the federal principle, or some particular federal framework” (12–13). The basic idea that “federalism” does not simply refer to advocacy for particular institutional forms proved highly influential not only in political science but also in law and philosophy. Subsequent work in each builds on Elazar's framework and explicitly distinguishes the federal “idea” characteristic of federalism and particular federal institutional forms. Even the institutionalist Wheare's followers, like Watts (2008), use the federal idea—as contrasted with a particular federal form (e.g., federation)—as a framing device. It remains the starting point for most philosophical analyses. I accordingly present and then evaluate stylized versions of leading institutional and ideological approaches to federalism.

Applying the conceptual desiderata above suggests ideological approaches are less plausible and institutional ones less problematic than many suggest. Both face issues along some evaluative dimensions. But the strongest case for ideological federalism is weaker than many suppose; the balance of reasons supports a form of institutionalism. At a minimum, this application suggests the dominant ideological view is problematic and sets burdens for developing stronger examples.

I argue for this conclusion first by outlining the case for and limits of ideological federalism and then by explaining how institutional views fulfill the desiderata above better than many suppose. This section analyzes ideological federalism. However, I spend considerable time contrasting ideological approaches with their institutionalist rivals when specifying ideological federalism's strengths and weaknesses. This helps streamline presentation of institutional views below.

Using “federalism” to describe governance characterized by a combination of “rule” in the form of a set list of formal final decision-making powers for central powers and “rule” in distinct formal final decision-making powers for non-central units could save the term in a manner consistent with use post-1788. It also avoids many challenges with past approaches. I thus propose a thin institutional account of federalism inspired by, but distinct from, Wheare's classic. If it proves undesirable all things considered, it still helps motivate institutionalism generally.

In this proposal, federalism is advocacy for adopting the federation. Consistent with the institutional Wheare and rival ideological Elazar (along with most of their followers, like Watts, 2008), federations here require at least two levels of government, each of which has distinct domains of final decision-making authority, does not rely on the other to implement its authority, and is accountable to a public not completely overlapping with that of another level of government. There must be a constitutional division of powers and at least two entities must be able to wield powers independent of each other while being accountable to distinct entities.³⁰

The United States, Canada, and Australia are paradigmatic. Federal and provincial (though not municipal) governments in those countries each have distinct constitutional authority over particular subjects. While powers may overlap in particular cases, each level of government can only validly act in a domain to the extent it is within their defined area of constitutional jurisdiction. It is, for example, possible that insurance promotion within banks can be considered part of a federal banking power and a provincial insurance power. However, federal and provincial governments can each only regulate aspects of insurance promotion within their defined powers. If a law fundamentally concerns banking, only the federal government can pass it. If conflicts between valid federal and provincial laws arise, country-specific rules resolve the conflicts.³¹

Each government should also be accountable to distinct constituencies. Elections for members of each should be (at least largely) distinct. The federal government should have members from each province (viz., United States.- or Australia-style constituent state, Swiss canton, German lander), but members should not only be members of provincial legislatures. Provinces, for example, should have representatives in the federal government. Some representatives could also serve in the provincial one. But personnel should be non-identical and roles should not collapse. This is necessary to safeguard distinct decision-making and accountability loci. If the same persons fulfill most roles at both levels, the possibility of distinctiveness minimizes. There should also be means for a province to make decisions alone. Provincial “rule” should not rely only on representation within a body that also maintains members from other constituent units.

The view on offer is still firmly institutional: It takes institutional form as conceptually prior to any ideas that may follow from it and accepts the concomitant narrower range of genuinely federal phenomena. Yet it has fewer institutional requirements than many predecessor views. Its less demanding definition of a federation permits a thinner concept of federalism. The proposal does not require negotiations to resolve issues, as Elazar (1987) suggests, or that federations result from pre-existing groups contracting in ways that maintain their status in new constitutional arrangements (contra id.; Tierney, 2022). This already suggests without proving that federalism itself does not entail a loyalty requirement. Multiple entities must have distinct domains of authority. The concept of federalism itself is agnostic on whether they must work together to do so or when. If a requirement exists, it must follow from the nature of the constitutive institutional form: Loyalty requirement proponents must explain how it is necessary for federations to exist. The same is true of subsidiarity. Likewise, the concept of federalism does not require many features frequently correlating with it, be they commitments to democracy or judicial review. The federal form may entail each. However, the concept does not include them.

This view shares Wheare's commitments to federations as the defining form of federal governance and federalism as a normative doctrine promoting adoption of that form. It adopts Whearean understandings of the federal form as requiring multiple decision-makers with final authority and the independence of their decision-making bodies such that, for example, the federal government is not just a collection of representatives of provincial governments and provincial governments do not rely on representation in a centralized body for much of their authority. Following Wheare, it further commits to distinct domains of authority in which governments are free to make decisions absent direct interference and views the United States as a paradigm case.

The proposal is nonetheless distinct from and helps avoid common criticisms of Wheare and thicker institutional models. Popelier (2021), for example, critiques federation-focused institutional views for their undermotivated commitments to democracy and the superiority of American-style federations, failing to account for non-democratic federations like the U.S.S.R or the normative desirability of German or Swiss governance.³² Tierney (2022), in turn, criticizes institutional theorists for taking the United States as exemplary and thereby failing to account for federalism's distinct constitutional commitments to cultural diversity. Wheare himself admittedly only counts the United States, Switzerland, Australia, and Canada as clear cases of federal governance (1946/1953, p. 22). While some of his limitations were the result of historical circumstance, like the then-nascent status of the Indian constitution (47), discussing Germany as quasi-federal (26–27) appears curious. Yet descriptive issues in Wheare do not undermine his broader institutionalist project. Closer attention to cases he dismisses suggests nuances in the basic federal form, not a need to go beyond institutionalism. India has the basic federal form characterizing the proposal on offer. And complications with the German division of powers can be accommodated within the view. For example, Germany's recognition of concurrent powers is unproblematic where exclusive federal and exclusive lander powers each remain. For another, lander delegate representatives in the upper house are acceptable where the lower, legislative house is distinct.³³

Popelier and Tierney's concerns are not, in turn, inherent to institutional federalism and do not apply here. The proposed view does not, for example, consider the United States alone as exemplary of federalism, suggest federations are categorically normatively superior to other forms of governance, entail strong commitments to democracy, or rely on cultural homogeneity.³⁴ The United States is merely one paradigmatic federation and I merely present advocacy for federalism as a plausibly normative view here without relying on commitments to democracy or a monoculture. The view on offer may not directly make the “cultural diversity” Tierney requires characteristic of federalism. But I respectfully see little reason to think federalism requires such diversity.

The proposed view, in other words, avoids many descriptive criticisms lodged at Wheare's original view by relaxing its institutional and ideological requirements. It may not address concerns that federalism ought to have further ideological commitments, but this is a feature, not a bug. Any ideological commitments should follow from the form or be argued for independently.

Applying our desiderata to this view demonstrates that many empirical issues with institutional views are not inherent to institutionalism and some form of institutional federalism, if not the present one, offers much stronger prospects for fulfilling other desiderata for a complete theory.

The preceding suggests one can construct a non-ad hoc institutional account of federalism that meets many desiderata of a complete theory and avoids the most common empirical challenges to other institutional approaches, including Wheare's account. Viewing federalism as a doctrine advocating the institutional form of the federation can, in short, maintain federalism as a distinct, plausibly justifiable concept that provides a unique solution to authority allocation problems, permits comparative inquiry, and explains (at least many) important historical developments. It does so in ways that can also serve other normative roles, like guiding judicial interpretation. While the idea matters regardless of its name, the concept is plausibly described as “federalism.”

Simply outlining this option for how to address important issues and contrasting it with its dominant ideological rival adds to our philosophical and political ontology and points to a more general case for institutionalism. If one does not find my particular view fully compelling, the manner in which it meets many burdens for a compelling theory of federalism supports adopting some institutional approach. It further supports moving away from the dominant ideological view(s). At a minimum, the forgoing demonstrates how one could construct a plausible institutional approach. There is no longer sufficient reason to assume institutional approaches must be empirically problematic or normatively inert. The quest for a distinct federal form and normatively compelling concept supporting its adoption need not be quixotic. If the distinct concept does not fit one's preferred use of the word “federalism,” scholars should still analyze it independently from other forms of decentralization. But calling it federalism now appears apt.

The forgoing identified how a thin institutional approach can avoid common empirical critiques. It first demonstrated that no form of institutionalism, including one focused on federations, must take the United States as its only paradigmatic case. If one finds the choices above too narrow, admitting further examples could provide a still stronger institutional view so long as all paradigm cases shared basic features distinct from other forms of governance. The forgoing then demonstrated that a core set of features found in the United States and other paradigms still permit great variety. The particular proposal admittedly did not account for all purportedly federal theories and practices. But it demonstrated that institutional accounts could have broader descriptive value than many expect. Where appeals to descriptive adequacy helped motivate distinguishing “the” federal idea and institutions and ideological accounts had to abandon institutional capaciousness to be action-guiding, the case for institutionalism seems strong. Where problems remain, they largely point toward details more complete views must address.

Any worries that institutionalism makes federalism a mere idealization prove too much. Institutional approaches describe an ideal type. Federal practice in most countries no longer requires strict divisions of powers. Overlapping areas of jurisdiction in which two entities can act under different powers are common in many federations such that a federal government can pass a public health law under a criminal law power and a provincial one can pass a public health law under a hospital regulation power. Institutional approaches committed to strict sovereignty within domains risk mere stipulation and may not account for real practices. Per Neumann (2005/1955, p. 207), even a “superficial” study reveals no “element common to all … [federal states] except a juristic one … [whereby citizens are] subject to two jurisdictions.” However, even the juristic conception meaningfully applies to paradigmatic federations. Such application provides benefits above even where paradigms imperfectly fit the institutionalist ideal in practice. And ideological conceptions have the same problems identifying non-trivial common features in all bodies, so idealization arguments threaten to make any form of “federalism” imaginary or trivial. If one wants to abandon the search for necessary and sufficient conditions for federalism on idealization grounds, one should recall that even conditions that do not perfectly mirror real-world challenges can serve normative functions above and ideological theories are no better here.

Further pressing objections cannot defeat this general case for institutional federalism. For instance, many forms of institutionalism permit problematic outcomes absent external commitments to other values. Philosophers traditionally avoided federalism partly due to the legacy of Jim Crow.⁴¹ Federalism has admittedly been used as a screen for injustice. Absent additional normative commitments, one may understandably worry that federalism will prove unjustifiable. If human rights norms avoided Jim Crow-like outcomes, other problems could remain. Institutionalism may not, for example, preclude inequities across sub-units within a country and country-wide solidarity-undermining competition between sub-units. Yet any conception deserving the name “federalism” will permit some wrongs. Justifications for federalism can limit bad outcomes. State stability norms may, for example, preclude divisions of powers fostering secession. But federalism divides sovereignty. A meaningful version with broad descriptive adequacy will likely provide someone with power to decide badly. Equal opportunities across states are, for example, desirable. But potential inequalities are a cost of federalism. One should not expand concepts beyond recognition to avoid problems, like inequality, when one can address the causes head-on.

Descriptively, in turn, if institutional approaches cannot describe all purportedly federal states as “federal,” they still provide meaningful guidance on how to identify them and capture more developments, like configurations providing for municipal authority, than some suppose. Not being able to describe confederations as “federal” may be unintuitive, but the balance of reasons supports it. Commitments to exclusive spheres of jurisdiction may, for example, struggle to describe the E.U. as federal. However, this result makes sense. Members claim to maintain complete sovereignty and the E.U. operates differently than countries such one may consider E.U. law international. This raises questions about why the U.N. is not federal ideological thinkers do not answer. Moreover, even those who view the E.U. as (con)federal recognize its incomplete federalization (Hueglin & Fenna, 2015, pp. xiii–xiv). One should not assume the E.U. is federal. The strongest accounts of federalism that would capture it above raise several problems. Giving up on institutional approaches just to cover the E.U. thus appears undermotivated.

Finally, the need for further work to see if conceptions of federalism using other possible justificatory principles are desirable should not lead us to abandon the concept of federalism as advocacy for a unique institutional solution, federations. A concept, once again, need not explain all its possible conceptions. Problems only arise when a concept cannot coherently explain what falls underneath it. Ideological approaches either fail to set a limit or set it too narrowly to fulfill their aims. Institutional approaches at least provide clear, plausible criteria for what counts as “federal.” Where they also better serve federalism's intended functions and do not face as many empirical problems as many believe, returning to some institutional approach is warranted.

The authors declare no conflicts of interest.