Comparative Federalism and Law
KeywordsEuropean Union Central Government Liberal Democracy Judicial Review Party System
A federal system of government is one where sovereignty is constitutionally shared and powers divided between an overarching central government and the governments of a set of constituent units, each of which enjoys a direct relationship with the people.
Federal systems of government, in the modern sense, have been in existence for over two centuries now – at least as long as the world has experienced liberal democracy – and historical antecedents go back much further. Federal systems of government also make up a good share of modern liberal democracies, particularly of the larger ones. With their attempt to use constitutional rules to enforce a system of shared rule they constitute a distinctive form of government, at least in principle fundamentally different from unitary systems, however decentralized those might be.
Federalism thus makes an important subject of study and the world’s federations an interesting set of potential comparators. Although much research into federal systems is case specific, comparative federalism of some sort has a long history, and increasing effort is now being made to carry out systematic comparison. Going back at last to pioneering federalism scholar Edward Freeman (1863), there is a strong tradition of examining diverse federal experiences, oriented to using typologies with some eye to generalization and considerable effort at conceptualization (notably Wheare 1946; Sawer 1969; Duchacek 1970; Davis 1978; Elazar 1987; Burgess 2005; Watts 2008; Hueglin and Fenna 2015). Much of that effort has been dedicated to “telling the federal story” and thus providing the essential basis for any attempt at productive comparison. They are typically structured around the main facets of federal systems: origins; constitutional assignment of responsibilities; judicial review; second chambers; fiscal relations; and constitutional amendment.
This entry explores the way federalism has been studied both as a dependent and as an independent variable. As a dependent variable, scholars have sought to explain how federal systems emerge; how they maintain or lose their “federalness”; whether they succeed or fail; and how they function. As an independent variable, there is ongoing interest in determining what difference federalism makes to what governments do and how they do it. Much of the discussion of federalism as a dependent variable involves the complex relationship between underlying conditions and character; institutional structures of federalism; the extra-federal institutional characteristics; and political organizations and processes. Meanwhile discussion of federalism as an independent variable revolves around two competing claims: that, on the one hand, federalism obstructs policy making and that, on the other, it facilitates policy making.
With its basis in a constitutional division of powers, federal has from the very beginning had a close connexion with constitutionalism and, in turn, with constitutional law and its interpretation. This has prompted questions at times about the impact of judicial review on policy making by governments and an ongoing interest in the effect of judicial review on the longer run evolution of federal systems. Do supreme courts, or certain types of supreme court, retard or facilitate adaptation? What has their role been in maintaining the elusive “balance” that is often seen as being important to federalism?
As with social science research more generally, a comparative perspective, and better yet, systematic comparative analysis, should be able to help us address all of these questions. The material covered in this overview shows that indeed the study of federalism has involved a healthy amount of general inference drawing as well as focused comparison between federal systems and some comparison between federal and nonfederal systems. As discussed below, however, potential for systematic comparison is severely constrained by the limitations of the material.
Federalism and the Comparative Method
In time, however, I came to realize that this was far too pretentious a project for one man. …each instance of a federalism ancient and modern is embedded in a set of unique local institutions, which themselves must be appreciated and understood. To acquire the information about history, the sensitivity to culture, and the linguistic competence to examine all these societies is more than any isolated scholar can do.
The obstacles Riker identified are practical rather than principled and thus presumably surmountable, either by someone more capable than he or through collective scholarship. Understanding any federal system requires a richly contextualized understanding, but once established, such an understanding should provide the basis for successful comparative analysis. The real challenge arises with the constraints on reliable comparison. The problem is twofold: a small number of cases and the distinctiveness of each case.
Methodological Challenges: Small N
Federalism is an ecumenical term, encompassing federal unions of a nongovernmental nature as well as governmental unions such as confederal systems that are not strictly speaking federal, as well as federal systems of government proper. Federal systems of government proper, which shall be the focus of this discussion, are distinguished from, on the one side, unitary systems or, on the other, confederal arrangements.
Traditionally, federations are categorized as such by reference to their constitutional framework and institutional form. This is as it should be, but it also carries the risk of mistaking appearance for reality and thus, as KC Wheare (1963, p. 20), the doyen of federal studies, reminded us: “it is obvious that the practice of the constitution is more important almost than the law of the constitution. It is usually convenient to begin with the law, as a basis of classification.”
On the one hand, some federations such as Malaysia or Austria are more federal in name than in substance (Rack 1996; Erk 2004; Case 2007). On the other hand, some such as Spain are more genuinely federal but give off sufficient mixed messages concerning their constitutional form to mislead some observers (Sala 2014). In a category of its own is the European Union (EU), which can be seen as a federation (Auer 2005), a federation-in-the-making (Annett 2010; Hueglin and Fenna 2015), or a distinctive model of “treaty federalism” (Hueglin 2013) – but which others see as not warranting the term at all (Wolinetz 2011). In between are those for whom the EU presents as an intriguing hybrid: a “jumbo confederation” (Lister 1996); an instance of “confederal federalism” (Kincaid 1999); a “quasi-federation” (McKay 2001); a “postmodern confederation” (Majone 2006); or even a “pluri-national federation in sensu cosmopolitico” (Dumont 2012). Questions about the precise status of the EU as some sort of federation have been brought to the fore by the challenges faced by the Eurozone in the global financial crisis and by the 2016 “Brexit” secession vote in the UK.
Thus, while it sometimes appears as if the world is awash in federations, with at least two dozen commonly said to exist, the search for genuine examples soon narrows the list down to a standard core and a handful of marginal cases. Federalism research, either comparative or case specific, overwhelmingly focuses on the classical core of the United States, Switzerland, Germany, Canada, and Australia. India, the leading federation of the developing world as well as the most populous and most diverse federation, merits a place on that list but has not often been included in comparative studies. The net can be broadened to include Austria, Belgium, Spain, South Africa, Argentina, Brazil, Mexico, Malaysia, Nigeria, and the European Union, but each has caveats attached. It is certainly impossible to do meaningful quantitative comparisons of federal systems with such a small N, but it is even difficult to do reliable qualitative ones.
Methodological Challenges: Great Variety
Even without the EU, variation among federal systems is considerable. This includes substantial differences in age, origins, design, and sociopolitical and economic contexts. If we mean federations in the modern sense of the term, as distinct from confederal arrangements, then federations range in age from the US system of the late eighteenth century to those such as Spain or South Africa created or being created in the late twentieth century. Having been established in 1949, the current German federal system is relatively young, but many of its features derive from an earlier incarnation in the late nineteenth century and indeed reflect a tradition going back to the middle ages.
The older federations – United States, Switzerland, Canada, Germany of the Second Reich, Australia – were all aggregative (though not necessarily purely so) in origin: formed by the voluntary union of formerly independent political entities. Much more recently we see federalism emerging out of a reverse process: unitary states such as Spain, Belgium, and South Africa devolving authority to accommodate diversity and local identities, often as a correlate of democratization. This contrast between aggregative and devolutionary origins is also referred to as “integrative” and devolutionary (Lenaerts 1990) or, less felicitously, “coming together” and “holding together.”
While the constitutional design of each federation is, in detail, unique, there is a broad distinction that separates two basic design types (Scharpf 1988, pp. 242–243; Hueglin and Fenna 2015, pp. 51–55). On the one hand are those federations, of which the United States is the archetype, characterized by a legislative division of powers in which the respective levels of government exercise full powers to make, implement, and administer policy in assigned domains and that rely on an elected, “Senate”-style, second chamber giving the constituent units an equality of representation in national decision-making. Originally these were envisaged as functioning in what KC Wheare called a “coordinate” or dualistic way with the respective levels of government autonomous in their own spheres.
On the other hand is the German model of joint or integrated federalism, echoed in the EU and partly in the Swiss case, characterized by an “administrative” or functional division of powers where policies are framed centrally but implemented and administered by the constituent units, and an appointive “Council”-style federal chamber providing the constituent units with direct representation in central decision-making. Rather than coordinate or dualistic in design, they are based on joint decision-making. For this reason, the integrated model is sometimes alternatively referred to as “cooperative federalism” – an unfortunate practice given the different way that term has widely been used in the study of federalism (see below). Since the division of powers has become much more overlapping, concurrent, and integrated in the American-style federations over time, the “representation of state governments in the federal legislative arena is a decisive factor in joint federalism” (Thorlakson 2003, p. 17).
Finally, there is the socioeconomic make-up of each federation. Fundamentally, there is the distinction here between those with pluri- or multinational societies and those that are regionally monocultural (sometimes, rather unhelpfully, called “territorial” federations). Unsurprisingly, the dynamics of the two types are characteristically different. Language differences, such as exist in Switzerland, Belgium, India, Canada, or Spain, contribute greatly to creating the former.
All in all, the combination of small N and large variety helps explain why some scholars prefer not to think of federal vs. unitary states in dyadic terms and instead to approach their functioning by way of the looser concept of “multilevel governance.” Whether this is justified or not, the reality remains that federal systems present difficulties for comparative analysis.
Having earlier remarked on the opportunity federalism offered for systematic comparative analysis, it is thus not surprising that Riker (1975, p. 131) later retreated to a much less sanguine view. “Mature federations are more unlike than alike”; thus “it is probably impossible to generalize about the operation of federalism.”
It is this sort of complexity that has caused scholars to flag the many obstacles to a reliable use of the comparative method in political science more generally. It is sometimes forgotten that pioneer theorist of the comparative method, John Stuart Mill (1974/1843), essentially rejected its application to the human sciences because of the “extraordinary” complexity of the variables. Since we will never be able to find two cases alike in all respects except the variable in question, we must rely on methods of comparison that cannot resolve problems of the “plurality of causes” (“equifinality,” in modern jargon) or intermixed causes – “chemical causation,” as Mill called it (on which see Ragin 1987). Mill (1974, p. 883) cautioned against “a specious semblance of conclusiveness” that comparative analysis can create in the social sciences. It is for this reason that “natural experiments” are the holy grail of comparative research since only they can provide some hope of holding the inevitably wide range of variables constant (Robinson et al. 2009). However, natural experiments are exceedingly rare.
Sober assessments of the comparative method more recently (e.g., Lieberson 1991) also focus on the difficulty of distinguishing multiple causes (“over determination”) or interaction effects. While Przeworski and Teune (1970) downplayed these difficulties, Przeworski (2007) eventually conceded that problems such as “endogeneity” force us to ask whether a “science of comparative politics” is possible at all. Such issues go to the heart of the matter in the human sciences, returning us to the question of whether the nomothetic enterprise of a “social science” is really feasible or whether we must settle for a much less ambitious idiographic approach of traditional “history” that treats cases as largely sui generis.
This retreat to a much more realistic view about the capacity of comparative analysis is a further reminder of the crucial value of in-depth case studies and the “process tracing” analysis they engage in to reveal complex causal mechanisms (George and Bennett 2005). Case studies occupy, Gerring (2004, p. 352) suggests, a seemingly precarious ontological position midway between ideographic and nomothetic extremes – but precisely there may well be the reality of the human sciences. It is this dilemma that leads to a plea for matched pairs approach to comparison, which in some ways offers the greatest potential for dealing with complex variables (Tarrow 2010). Any fewer cases would allow no comparison; any more threatens to compromise on empirical richness. Paired comparisons have been used widely in the federalism literature: juxtaposing Canada and the United States, Canada and Australia, the EU and the United States, Canada and the EU, the United States and Switzerland, and indeed almost any other permutation one can imagine.
Federalism as a Dependent Variable
Interest in federalism itself has focused on at least four different but related questions. The first is origins: why do some states emerge in a federal rather than unitary form? The second is survival, or more precisely, failure: why do some federations expire or dissolve? The third and fourth concern the dynamics of successful or established federal unions: how and why does the balance between centripetal and centrifugal tendencies change over time, and, more specifically, what role does judicial review play in the evolution of federal systems?
Why some states took a federal rather than unitary path is an interesting question that also has bearing on the understanding of federal practice. In addition to helping understand the past and interpret the present, it also helps with envisioning the future: under what circumstances might federalism be a “solution” to present woes, such as unitary states fractured by conflicting regional identities (Wibbels 2006). Most reflection and research has focused on aggregative federations since they are older and more classically federal. The obvious first step has been to map the experience of precursors, the various confederal unions that have existed over the centuries, and about which, thanks to Davis (1978), Forsyth (1981), and Lister (1999, 2001), much is now well understood.
It has long been recognized, as Wheare (1946, pp. 37–45) originally spelled out, that some combination of external threat and economic gain has been instrumental in the formation of the main aggregative federations. Riker (1964, p. 12) turned the strategic gain notion into a formal proposition, but shorn of its nuances the idea lost rather than gained utility. Going beyond Riker’s rather “commonplace” observation (Davis 1978, p. 137) has been difficult, though.
Empirically, an inviting comparison is between German and Italian unification in the 1870s. Although not at all a “natural experiment” (contra Ziblatt 2006: Chapter 1), since it does not control for other variables, these divergent forms of modern state formation make for a potentially illuminating contrast. Is there something about the fact that militarily powerful Prussia drew the German states into a federal union, while Piedmont presided over the formation of a unitary state in Italy? Wheare’s answer was that federalism requires an established capacity for regional self-government. Italy could not federate until the autocracies of Sicily and Naples were abolished. “But when this defeat was accomplished, nothing was left. There were no governments with authority rooted in the society they governed.” Ziblatt (2006) explored the comparison in more detail and arrived at the same conclusions.
Simple binary comparisons, as noted above, have their strengths; however, they also have their weaknesses. A critical weakness to the simple two-case comparison here is that while it helps us understand why modern Italy could not take a federal form, it cannot tell us why Germany did; “infrastructural capacity,” as Ziblatt calls it, would certainly seem to be a necessary condition, but there can be no valid inference from this comparison that it is in any way a sufficient one. A host of other conditions almost certainly have to be met.
Success and Failure
The flip side of explaining origins is explaining failure or the limits of federalism: presumably many of the same factors that make for federalism in the first place determine whether it will succeed or not. Wheare’s list of “requisites” also included such factors as previous existence of the constituent units as autonomous political entities and similarity of social and political institutions.
Because federal union has often seemed like the ideal solution to state-building challenges, a sense of disappointment has long surrounded those instances where a federal vision has proved insufficient to the task, including instances of “total failure” (Hicks 1978). The general answer has been to infer from the broad experience what conditions are conducive to success, such as the number of units or the relationship between unit boundaries, the distribution of ethnic or other potentially troublesome minorities (Lemco 1991), or the experience of prior independence (Markovic 1995). Of particular interest is whether federalism assuages or, alternatively, amplifies regional discontent when genuine differences of identity are at issue. Is federalism the “solution” in divided societies, as Elazar (1987, p. 8) argued? Inferences from multicase surveys (Amoretti and Bermeo 2004, pp. 457–489; McGarry and O’Leary 2009; Seymour and Gagnon 2012; Anderson 2015) suggest that while generalization is difficult, federalism does assist in “managing cleavages” and that the notion that federalism exacerbates identity issues can only be sustained by overlooking patently incompatible conditions and creating illusory counterfactuals. Federalism is not a panacea (Erk 2015).
Evolution and Erosion
Distinguishing “successful” from “unsuccessful” federations does not do justice to the quotidian but pervasive difficulties often experienced by the successful, established, federations. While the frictions of “cool” federalism are much less dramatic than those of its “hot” version (Sager 2005), they can nonetheless be a source of dissatisfaction. A widespread syndrome has been one of the creeping centralization and the resulting perception that federalism is being eroded in the established federations. Aggregative federations typically took shape through a process of debate and negotiation (misleading characterized by Riker as a “bargain”) in which efforts were made to lock the relationship between the two levels of government into the desired form through constitutional arrangements. The question is how those arrangements subsequently evolved and what role they and the institutions they established played in that evolution.
For those unions formed prior to or in the early years of industrial modernization, powerful centripetal tendencies were inevitable, particularly but not only because of the way the emergence of national welfare states in the mid-twentieth century effectively reversed major elements of the traditional distribution of responsibilities (Corry 1941; Peterson 1995; Fenna 2007). In an influential formulation, Sawer (1976, p. 64) described this “as involving three successive stages; first co-ordinate or dual federalism, secondly co-operative federalism, and thirdly organic or integrated federalism.” Attempts to map and compare centralizing trends show a tendency that spans quite different federal systems and indeed appears “ubiquitous” (Döring and Schnellenbach 2011), but which has nonetheless been variable (Krane 1988). An alternative reading sees periods of centralization and periods of decentralization (Chhibber and Kollman 2004, pp. 101–160).
This evolution, which in some instances has quite transformed the older federal systems, has more often than not taken place via subconstitutional means: typically, a combination of favorable judicial interpretation and the exercise of fiscal power. There does, however, seem to be a characteristic division between the continental federations Switzerland and Germany, and the New World federations, in this respect, with the former making much more regular and effective use of formal constitutional amendment (Hueglin and Fenna 2015, pp. 275–307). For the latter, one can agree with Rodden (2006, p. 38) that “federal constitutions are notoriously poor guides to the distribution of authority in modern federations.”
The institutional barriers to change in federal systems did retard this process of adaptation for some time, but once the dam was broken, it has seemed to some observers that there has been a very real possibility of moving from centralization to what Braun (2011) calls “overcentralization.” Some attempt has been made to compare the degree of centralization across federations using fiscal indicators, with, unsurprisingly, Canada and Switzerland ranked the least, United States and Germany in the middle, and Australia and Austria the most (Thorlakson 2003).
The efficacy of federalism’s constitutional constraints in resisting centripetal tendencies has been discussed at length by some scholars (Filippov et al. 2004; Bednar 2009), but with mixed results and, more to the point, little attempt at empirical validation and comparative analysis. Riker (1964, p. 101), who attached minimal significance to the institutional features, argued that “the decentralization of the two-party system is sufficient to prevent national leaders … from controlling their partisans by either organizational or ideological devices. As such, this decentralized party system is the main protector of the integrity of states in our federalism.” However, this observation was based on little more than impressionistic argument. More systematic comparison reinforces the suspicion that parties should be regarded as merely an intervening variable and the party system a consequence rather than a cause of centralization (Chhibber and Kollman 2004, p. 227).
An alternative take on this variable asks not whether the party system is centralized, but whether it is Left-dominated. Parties of the Left are congenitally centralizing and where they enjoy a dominant position they will drive federations in that direction. Or so argue Turgeon and Wallner (2013), who illustrate their point with a two-case contrast between Australia and Canada. That comparison provides useful prima facie evidence, but one would need other cases with the underlying characteristics of Australia but without its Labor Party to provide compelling demonstration.
One of the few attempts to address the general question of what drives centralization through an empirically richer multicase comparative analysis has been Erk (2007), who compares Austria, Belgium, Canada, Germany, and Switzerland. Erk found that having a federal society – i.e., one characterized by regional ethno-linguistic diversity – proves most resistant to centralization. This lent support to Livingston’s (1952) well-known proposition that “Federalism is a function not of constitutions but of societies.” Riker (1964, p. 111) similarly emphasized the role of “popular sentiment of loyalty” in determining which level of government will dominate. While confirming the importance of that societal diversity, Braun (2011) sought to supply a more nuanced reading, using much the same range of cases, but including variations in institutional structure. Similarly, Bolleyer (2009) explored some of these cases to determine how the varying ability of constituent units to cooperate might be explained and might affect federal dynamics.
Courts and the Constitution
How effective have federalism’s constitutional/institutional “safeguards” been in retarding those centralizing tendencies noted above? Of all the mooted institutional safeguards, it is the courts that have been most discussed. This, however, has overwhelmingly been done using an N of one – the United States. Can comparative analysis tell us anything about the contribution judicial review has made to the evolution of federal systems? Views about the significance of the courts range from Alexander Hamilton’s blithe and rather disingenuous reassurance in Federalist 78 that the judiciary are the “least dangerous” branch (following Montesquieu), to Dicey’s (1915, p. 170) dictum that federalism “means legalism – the predominance of the judiciary in the constitution.” In the United States, legal scholars have long debated whether the Supreme Court should or can be expected to police the federal division of powers (e.g., Wechsler 1954; McGinnis and Somin 2004).
Federalism and Constitutionalism
Federalism contributed significantly to the emergence and development of modern constitutionalism since the creation of federal unions inevitably entailed legal formalization of the negotiated arrangements. Along with “new beginnings,” it was one of the two paths to modern constitutionalism (Ackerman 1997, p. 775). In some cases, such as Canada or Australia, federalism was a sufficient reason for the drafting of a modern constitution; in others, such as the United States, both factors were at work. Constitutionalism has thus been an important element of federal systems, manifesting itself in questions about amendment and other forms of adjustment to changing conditions such as interpretation and reinterpretation by the courts.
Almost without exception, federations have established higher courts either implicitly or explicitly mandated to assess the constitutionality of government actions and thus adjudicate conflicts over the division of powers. Thus federalism has also provided a major impetus for the development of judicial review. There was a great deal of truth to Dicey’s observation – certainly more than there was to Hamilton’s. Because of its reliance on (and contribution to) constitutionalism, and because of its unique attempt at creating some kind of shared sovereignty, federalism almost inevitably created a privileged role for the courts. Indeed, as embodied in the landmark Marbury v. Madison (1803) decision of the US Supreme Court, federalism was instrumental in the rise of judicial review: the then-controversial and path-breaking notion that the judiciary may declare the actions of government unconstitutional.
Judicial review spread to the Anglo federations Canada and Australia but because of very different associations and conditions was slower coming to the European states. By the second half of the twentieth century, though, it had become the international norm (and of course not just in federal systems) a status reflected in the contribution made by the European Court of Justice to the consolidation of the European Union. Characteristic differences remain, with the New World federations reposing the task in a general court of appeal – the Supreme Court of the United States or of Canada or the High Court of Australia – while the German model is one of a specialist constitutional court, the Bundesverfassungsgericht. There is, however, one distinct anomaly: Switzerland, the one federation where the jurisdictional scope of the national government is not subject to judicial review at all (though actions of the cantons are subject to the Federal Court). Switzerland’s unique reliance on direct democracy has made judicial review of central government actions seemingly unnecessary.
The Judicial Safeguards of Federalism?
This net centralist/nationalist bias of federal high courts is largely explained by the strong institutional factors linking such courts to the political organs of the central government. This government is generally responsible for the existence of the court, its administration, its budget, its internal procedures, and most notably for the selection of its judges. (Bzdera 1993, p. 27)
Highly suggestive for Bzdera was the way the court least beholden to the central government, the Judicial Committee of the Privy Council in London during the formative years of Canadian federalism, played a pronounced decentralizing role. The Judicial Committee was an Imperial instrument serving as the final court of appeal for Canadian federalism until 1949. Entirely external and independent, it manifestly favored a federal rather than a centralizing interpretation of what was intended to be a quite centralizing constitution.
Of course, periodization is a crucial component of such an analysis: the building phase of a federal union is presumably rather different from more mature phases. Thus, obversely, we might expect courts to play a centralizing role in “constituting” an integrated union when faced with disparate elements. Thus Goldstein (2001), for instance, compares the role of the European Court of Justice with the US Supreme Court in its early years. More generally, comparison of the EU with other federations in their earlier days has been a natural avenue of research (e.g., Church and Dardanelli 2005; Jabko 2006).
If this argument about courts and centralization is correct, we should find an inverse correlation between the central government’s control over the court and the degree to which the court defends federalism. And some such correlation can be discerned (Aroney and Kincaid 2016). Centrally dominated supreme courts in Australia and the United States have assumed a deferentially centralizing role. By contrast, the supreme courts of the German and Swiss federations, constituted with much greater respect for the federal principle, have adopted a more balanced position – with the Swiss Federal Court (Tribunal fédéral/Bundesgericht) demonstrating that balance in the license it extends to the cantons (Lienhard et al. 2016). Of course, these are merely correlations and in no way control for the possible influence of a wide range of other factors. It is also not inconsistent with the recognition that judicial review functions via internal intellectual logics encapsulated in the notion of interpretive “doctrine” (Baier 2006).
Governing Federal Systems
An enormous amount of the work done on federal systems concerns the practical questions about the mechanics of cooperation and conflict when two constitutionally entrenched levels of government interact, as over the course of the last century they have increasingly had to do (Poirier et al. 2015). This has two main dimensions: fiscal and political–administrative, the latter coming under the rubric of intergovernmental relations, or “IGR.”
Comparative analysis of the fiscal relations between governments in federal systems is limited. One exception is Rodden’s (2006, p. 272) survey of the “murky semi-sovereignty” that characterizes the spending and borrowing position of the constituent units in federal systems and the possible “hazard” this poses for public finance. Very little comparative work has been done on what is perhaps a more important aspect of fiscal federalism, viz., vertical fiscal imbalance (VFI) and the use of conditional grants by central governments. The significance of VFI lies chiefly in the way it allows central governments to “encroach” on the constitutional prerogatives of the constituent units by use of the “spending power.” Some comparative work has been done on horizontal fiscal equalization (HFE), the subsidization of less well-off jurisdictions to maintain comparability of government services across the country (Béland and Lecours 2011, 2014). Equalization varies considerably from federation to federation – particularly in degree but also in manner. Some federations equalize thoroughly or aggressively (Australia, Germany); some equalize partially (Canada, Switzerland); and one does very little at all (the United States).
The growth and consolidation of intergovernmental relations in the traditional dualistic federations over the twentieth century has long been studied under the rubric of “cooperative federalism.” As Elazar (1991, p. 69) noted, “the adjective cooperative…does not imply that intergovernmental relations are always peaceful and friendly. … Cooperative refers to the fact that governments must cooperate, that is, work and function together.” The story of cooperative federalism encompasses a wide range of interactions including negotiation, conflict, coercion, and genuine cooperation. It can also involve practices of “collaborative federalism,” viz., the establishment of joint governance arrangements (Painter 2001).
Intergovernmental relations, however, are the most elusive subject in the study of federalism. Lacking the explicit legal form of constitutions and judicial interpretation, the explicit numerical quality of fiscal relations, or the tangible shape of constitutionally mandated institutions, they occupy a distinctly gray area. Comparison generates some systemic distinctions, such as that between “executive federalism” and “administrative federalism.” Parliamentary federations with a legislative division of powers generate high levels of executive-to-executive interaction that has been termed “executive federalism” (Watts 1989). Both US federalism, with its presidential separation of powers, and German federalism, with its administrative division of powers, give rise to more diffuse interaction, often at the bureaucratic levels, that has been termed “administrative federalism” (Fenna 2012; Hueglin and Fenna 2015, pp. 239–246).
Comparing the very different processes in the American and Canadian systems, Bakvis and Brown (2010) find that outcomes are nonetheless similar. It is reasonable to expect, though, that presidentialism and parliamentarism will have characteristically different effects on the way a federal system functions. Comparing four federations, Kelemen (2000) found that to indeed be the case for environmental policy: the inherent weaknesses of presidential separation of powers arrangements leads to imposition of tight regulatory frameworks on the constituent units in a way that does not tend to happen in parliamentary systems. Again, such findings are hostage to the shortage of comparative material. Presidential systems are generally rare in the world of successful liberal democracies and specifically rare in the world of successful democratic federations. This requires Kelemen to use the EU as a functional equivalent to the United States.
This evolution of modern federalism, with its increasing entanglement of the two levels of government and degree of convergence between the way powers are divided in the dual and integrated federations, has created a challenge for modern federal governance. It has also generated characteristic reform strategies in the two types of federations. Legislative federations have sought to strengthen shared rule in response to the increased sharing, while administrative federations have moved towards disentanglement (Broschek 2015).
Federalism as an Independent Variable
What difference does having a federal system make? Riker’s answer was “very little,” and then only to the style rather than the substance of what governments do. Davis (1978, pp. 212–213) was, if anything, more sceptical – declaring that the term “federalism” had become “phylogenetically senile.” On the other hand, bold claims and strong criticisms have been made about federalism over the years, including that it protects individual rights, provides a way to accommodate diversity, retards policy making, and facilitates policy making. Here the logical approach is, as Riker (1969, p. 139) suggested, a matched-pairs design, juxtaposing “most (otherwise) similar” federal and unitary systems. Short of a genuine natural experiment, the problem here of course is that the available material provides few opportunities for anything but the most approximate “matching.”
Federalism as Hindrance?
Does divided jurisdiction help, hinder, or make little difference to effective governance and the introduction of new policies? In particular, focus has been on the potential for market-modifying policies. The short answer seems to be “all of the above”: federalism has or can help, hinder, or make little difference to reformist policy, depending on circumstances.
Writing when US federalism was proving a significant obstacle to policy reform for a rapidly industrializing society, Dicey (1915, p. 169) declared that a system of divided jurisdiction “is unfavourable to the interference or activity of government.” The assessment was echoed less dispassionately in the years of the Great Depression, most famously by English socialist Harold Laski (1939). For Dicey, Laski, and others, the main culprit was misalignment between responsibilities and capacities combined with the power of the courts to veto any kind of adjustment through initiatives of the national government. At its extreme, this took the form of the perverse situation created in the Lochner era of constitutional interpretation in the United States that neither level of government was granted power to regulate the market (Kens 1998).
All of this discussion and handwringing implicitly assumed federalism of the American type that is a legislative rather than administrative division of powers. Jurisdictional impasse would self-evidently not be the same problem in federalism of the German type. But “joint-decision” systems may well fall prey to their own pathologies, and according to Fritz Scharpf (1988), they indeed are, but of a different kind. In this view, Germany and the EU are afflicted with tendencies to a policy and institutional reform inertia Scharpf called the “joint-decision trap.” This syndrome arises when unanimity among too many decision makers is officially or effectively required. The proposition has been explored through empirical work and comparative reflection (Héritier 1999; Benz 2011; Falkner 2011).
The question of federalism’s impact on market-modifying initiatives was looked at in empirical depth in a comparative analysis of the emergence of welfare states in Australia, Canada, the United States, Austria, and Switzerland (Obinger et al. 2005). That analysis corresponded to the approach recommended by King et al. (1994, p. 45) as “structured, focused comparison.” It confirmed the contribution federalism made to retarding the development of modern welfare states in some of these cases, but also noted the extent to which that was a one-off event and the extent to which other contributing factors have been at work. In the US system, for instance, it is easy to confuse the effect of federalism with the effect of extraneous institutional characteristics, notably the presidential separation of powers design of the national government (Greer 2010). In the United States, it is also easy to confuse the effects of federalism with the effects of some of the underlying circumstances associated with federalism, notably the conservative pulls of the South (Finegold 2005).
Some of the adverse dynamics of federalism in development of redistributive social policy also seem evident in the environmental field. Comparing Switzerland, Canada, the United States, and the EU, Weibust (2009) found that centralization of powers in this area rescued the country from debilitating collective action problems of a decentralized system. At the same time, other comparative analyses of the impact of federalism on environmental policy find little scope for generalization, partly because each federation is different – both in the nuances of its federal structure and in other, extraneous, factors (e.g., Harrison 2010). Meanwhile, Banaszek (1996) compared women’s suffrage in Switzerland and the United States – two federal systems with contrasting outcomes – and quite absolved divided jurisdiction from any particular outcome responsibility. In her comparison, it was not intra-federal variation that explained the contrasting outcomes, but entirely extra-federal factors, notably the misguided strategies chosen by the Swiss suffragette movement itself.
Federalism as Facilitator
On the other side of the ledger is the claim that federalism enhances policy making. In the “laboratory federalism” thesis, one of the great virtues of having several independent jurisdictions and two levels of government is the much greater opportunity for policy “experimentation” and learning. As first proposed by James Bryce (1893, p. 353) and later promoted by Justice Brandeis (1932) of the US Supreme Court, federalism creates “laboratories of democracy” wherein unsuccessful experiments are safely quarantined and whereby successful ones can be emulated. The thesis is attractive but does the vaunted experimentation and learning actually occur in practice – and to the extent it does occur, under what specific circumstances?
Theoretical doubts about the likelihood of such a beneficent dynamic have been widely voiced (e.g., Strumpf 2002), but much less has been done to test the proposition empirically. Some systematic research has been done in the United States, where most of the interest in the thesis is focused. Findings have highlighted the proposition’s rather idealistic nature, consistent with the judgments of the more theoretical literature on the subject. “The characterization of the 50 states as laboratories of democracy is an appealing image, but it is a standard that is rarely met in practice” (Karch 2007, p. 204). Little effort, however, has been made to move beyond the single-case perspective on this question. Two types of comparison are required: a comparison of federal and unitary systems, as Riker suggested, to provide some better indication of whether federalism does indeed exhibit experimentation and learning tendencies, and a comparison of different federations to give some indication of what intra- or extra-federal characteristics contribute to or negate those putative tendencies.
An illustration of the former is Chappell and Curtin (2013) who follow Riker’s (1969, pp. 139–140) advice and utilize the Australia–New Zealand comparison to explore the difference between a federal and a unitary system enjoying otherwise substantial similarities. They found that divided jurisdiction did indeed provide greater opportunities for progressive initiatives. As far as the latter is concerned, there is the distinct possibility, as Pierson (1995) argued, that “federalism” is too crude an independent variable and that it is particular features of specific federal systems that we must examine. Evidence suggests that, indeed, a great deal of contingency is involved (Wälti 2013).
As in the social sciences generally, meaningful comparison is the methodological holy grail of federal studies. Federations are a significant subset of liberal democratic governance types and should provide us with the scope to explore and test ideas about either the way federal systems change and evolve (federalism as dependent variable) or the way federal systems affect the practices of politics and policy making (federalism as independent variable). The challenges, though, are considerable: the number of cases is small, and the degree of variance is large. As the above survey demonstrates, comparative efforts have focused overwhelmingly on the usual suspects: that small handful of stable liberal democracies with well-established formally and substantively federal systems. The United States, Switzerland, Canada, Australia, and Germany are the analytic elect. Increasingly, the EU is being added to the mix – despite, as we noted, ambivalence about its federalness. Outlying cases, such as failed federations, become relevant and useful for specific purposes.
Riker’s slightly whimsical notion of testing the significance of federalism by focusing on paired comparisons contrasting on the independent variable has not, with the occasional nicely suggestive exception, proven fertile for the simple reason that such matched pairs do not exist in the real world. The notion is, as Mill (1974, p. 882) noted, “manifestly absurd” since, to paraphrase Mill, if two countries were alike in every respect other than the structure of their system of government, they would be alike in that respect too. Examining either the fate of federalism as a dependent variable or the impact of variations in federal and related governmental structures does not entail unitary–federal comparison. Here, the problem is paucity of material, with analyses forced to rely on only one case of separation of powers federalism (the United States) and only a modest number of parliamentary federations, or one case of integrated federalism (Germany) and some partial examples (Switzerland, EU) or only a modest number of dualistic ones. Nonetheless, interesting and suggestive work has been done using these limited resources.
- Amoretti UM, Bermeo N (eds) (2004) Federalism and territorial cleavages. Johns Hopkins University Press, BaltimoreGoogle Scholar
- Aroney N, Kincaid J (eds) (2016) Courts in federal systems. University of Toronto Press, TorontoGoogle Scholar
- Baier G (2006) Courts and federalism: judicial doctrine in the United States, Australia, and Canada. UBC Press, VancouverGoogle Scholar
- Bednar J (2009) The robust federation: principles of design. Cambridge University Press, CambridgeGoogle Scholar
- Brandeis J (1932) Opinion in New State Ice Co. v. Liebmann, 285 US 262Google Scholar
- Bryce J (1893) The American commonwealth, 3rd edn. Macmillan, London, 2 volsGoogle Scholar
- Burgess M (2005) Comparative federalism: theory and practice. Routledge, LondonGoogle Scholar
- Case W (2007) Semi-democracy and minimalist federalism in Malaysia. In: He B, Galligan B, Inoguchi T (eds) Federalism in Asia. Edward Elgar, Cheltenham, pp 124–143Google Scholar
- Chhibber P, Kollman K (2004) The formation of national party systems: federalism and party competition in Canada, Great Britain, India, and the United States. Princeton University Press, PrincetonGoogle Scholar
- Davis SR (1978) The federal principle: a journey through time in quest of a meaning. University of California Press, BerkeleyGoogle Scholar
- Dicey AV (1915) Introduction to the study of the law of the constitution, 8th edn. Macmillan, LondonGoogle Scholar
- Duchacek ID (1970) Comparative federalism: the territorial dimension of politics. Holt, Rinehart and Winston, New YorkGoogle Scholar
- Dumont H (2012) The European Union, a plurinational federation in Sensu Cosmopolitico. In: Michel S, Gagnon A-G (eds) Multinational federalism: problems and prospects. Palgrave Macmillan, BasingstokeGoogle Scholar
- Elazar DJ (1987) Exploring federalism. University of Alabama Press, TuscaloosaGoogle Scholar
- Elazar DJ (1991) Cooperative federalism. In: Kenyon DA, Kincaid J (eds) Competition among states and local governments: efficiency and equity in American federalism. Urban Institute Press, Washington, DC, pp 65–86Google Scholar
- Erk J (2007) Explaining federalism: state, society and congruence in Austria, Belgium, Canada, Germany, and Switzerland. Routledge, LondonGoogle Scholar
- Falkner G (ed) (2011) The EU’s decision traps: comparing policies. Oxford University Press, New YorkGoogle Scholar
- Forsyth M (1981) Unions of states: the theory and practice of confederation. Leicester University Press, LeicesterGoogle Scholar
- Freeman E (1863) History of federal government from the foundation of the Achaian league to the disruption of the United States, vol 1. Macmillan, LondonGoogle Scholar
- Freeman EA (1896) Comparative politics, 2nd edn. Macmillan, LondonGoogle Scholar
- George AL, Bennett A (2005) Case studies and theory development in the social sciences. MIT Press, Cambridge, MAGoogle Scholar
- Goldstein LF (2001) Constituting federal sovereignty: the European Union in comparative context. Johns Hopkins University Press, BaltimoreGoogle Scholar
- Harrison K (2010) Multilevel governance and carbon pricing in Canada, the United States, and the European Union. In: Courchene J, Allan R (eds) Canada: the state of the federation 2009 – carbon pricing and environmental federalism. McGill-Queen’s University Press, Montreal/Kingston, pp 111–128Google Scholar
- Hicks U (1978) Federalism: failure and success, a comparative study. Oxford University Press, New YorkGoogle Scholar
- Hueglin TO (2013) Comparing federalisms: variations or distinct models? In: Benz A, Broschek J (eds) Federal dynamics: continuity, change, and the varieties of federalism. Oxford University Press, Oxford, pp 27–47Google Scholar
- Hueglin T, Fenna A (2015) Comparative federalism: a systematic inquiry, 2nd edn. University of Toronto Press, TorontoGoogle Scholar
- Jabko N (2006) The politics of central banking in the United States and in the European Union. In: Menon A, Schain M (eds) Comparative federalism: the European Union and the United States in comparative perspective. Oxford University Press, New York, pp 275–291Google Scholar
- Kens P (1998) Lochner v. New York: economic regulation on trial. University Press of Kansas, LawrenceGoogle Scholar
- King G, Keohane RO, Verba S (1994) Designing social inquiry: scientific inference in qualitative research. Princeton University Press, PrincetonGoogle Scholar
- Krane D (1988) The evolutionary patterns of federal states. In: Lloyd Brown-John C (ed) Centralizing and decentralizing trends in federal states. University Press of America, Lanham, pp 39–62Google Scholar
- Laski H (1939) The obsolescence of federalism. New Republic 98(1274):367–369Google Scholar
- Lemco J (1991) Political stability in federal governments. Praeger, New YorkGoogle Scholar
- Lienhard A, Kettiger D, Bühler J, Mérillat L, Winkler D (2016) The federal supreme court of Switzerland: judicial balancing of federalism without judicial review. In: Aroney N, Kincaid J (eds) Courts in federal systems. University of Toronto Press, TorontoGoogle Scholar
- Lister FK (1996) The European Union, the United Nations and the revival of confederal governance. Greenwood Press, WestportGoogle Scholar
- Lister FK (1999) The early security confederations: from the ancient Greeks to the United Colonies of New England. Greenwood, WestportGoogle Scholar
- Lister FK (2001) The later security confederations: the American, “new” Swiss, and German unions. Greenwood, WestportGoogle Scholar
- Majone G (2006) Federation, confederation, and mixed government: a EU–US comparison. In: Menon A, Schain M (eds) Comparative federalism: the European Union and the United States in comparative perspective. Oxford University Press, New York, pp 121–148Google Scholar
- Markovic M (1995) The federal experience in Yugoslavia. In: Karen K, Ostry S, Simeon R, Swinton K (eds) Rethinking federalism: citizens, markets, and governments in a changing world. University of British Columbia Press, VancouverGoogle Scholar
- McGinnis JO, Somin I (2004) Federalism vs. states’ rights: a defense of judicial review in a federal system. Northwest Univ Law Rev 99(1):89–130Google Scholar
- Mill JS (1974/1843) A system of logic ratiocinative and inductive: being a connected view of the principles of evidence and the methods of scientific investigation, 2 vols. University of Toronto Press, TorontoGoogle Scholar
- Obinger H, Leibfried S, Castles FG (eds) (2005) Federalism and the welfare state: new world and European experiences. Cambridge University Press, CambridgeGoogle Scholar
- Peterson PE (1995) The price of federalism. Brookings Institution, Washington, DCGoogle Scholar
- Poirier J, Saunders C, Kincaid J (2015) Intergovernmental relations in federal systems: comparative structures and dynamics. Oxford University Press, Don MillsGoogle Scholar
- Przeworski A (2007) Is the science of comparative politics possible? In: Boix C, Stokes S (eds) The Oxford handbook of comparative politics. Oxford University Press, OxfordGoogle Scholar
- Przeworski A, Teune H (1970) The logic of comparative social inquiry. Wiley, New YorkGoogle Scholar
- Rack R (1996) Austria: has the federation become obsolete? In: Hesse JJ, Wright V (eds) Federalizing Europe? The costs, benefits, and preconditions of federal political systems. Oxford University Press, Oxford, pp 204–218Google Scholar
- Ragin CC (1987) The comparative method: moving beyond qualitative and quantitative strategies. University of California Press, BerkeleyGoogle Scholar
- Riker WH (1964) Federalism: origin, operation, significance. Little, Brown and Company, BostonGoogle Scholar
- Riker WH (1975) Federalism. In: Greenstein FI, Polsby NW (eds) Handbook of political science, vol 5, Governmental institutions and processes. Addison-Wesley, Reading, pp 93–172Google Scholar
- Rodden JA (2006) Hamilton’s paradox: the promise and peril of fiscal federalism. Cambridge University Press, New YorkGoogle Scholar
- Sager LG (2005) Cool federalism and the life-cycle of moral progress. Willilam Mary Law Rev 46(4):1385–1398Google Scholar
- Sawer G (1969) Modern federalism. C. A. Watts & Co., LondonGoogle Scholar
- Sawer G (1976) Modern federalism, new edn. Pitman Australia, CarltonGoogle Scholar
- Seymour M, Gagnon A-G (eds) (2012) Multinational federalism: problems and prospects. Palgrave Macmillan, BasingstokeGoogle Scholar
- Turgeon L, Wallner J (2013) Adaptability and change in federations: centralization, political parties, and taxation authority in Australia and Canada. In: Skogstad G, Cameron D, Papillon M, Banting K (eds) The global promise of federalism. University of Toronto Press, Toronto, pp 188–213Google Scholar
- Wälti S (2013) Federalism and public policy: do federalism, regionalism and hybridity make any difference? Evidence from environmental policy. In: Loughlin J, Kincaid J, Swenden W (eds) Routledge handbook of regionalism and federalism. Routledge, Abingdon, pp 116–129Google Scholar
- Watts RL (1989) Executive federalism: a comparative analysis. Institute of Intergovernmental Relations, Queen’s University, KingstonGoogle Scholar
- Watts RL (2008) Comparing federal systems, 3rd edn. McGill–Queen’s University Press, Montreal/KingstonGoogle Scholar
- Weibust I (2009) Green leviathan: the case for a federal role in environmental policy. Ashgate, FarnhamGoogle Scholar
- Wheare KC (1946) Federal government. Oxford University Press, OxfordGoogle Scholar
- Wheare KC (1963) Federal government, 4th edn. Oxford University Press, OxfordGoogle Scholar
- Wolinetz SB (2011) Comparing the incomparable: treating the EU in comparative context. In: Laursen F (ed) The EU and federalism: polities and policies compared. Ashgate, Farnham, pp 27–39Google Scholar