Federalism and Federation

Introduction / Definition


Federalism is a philosophy, doctrine and arguably an ideology (Watts: 1998) that favors a distinct territorial pattern of government, one that combines the centralization of some political powers and the decentralization of others. The etymological origins of federalism derive from “foedus,” the Latin for “alliances” or “leagues” of states that joined together in pacts, covenants or agreements, typically for defensive purposes. Federalists promote “federal political systems,” which combine principles of “shared rule” with “self-rule” in their territorial designs (Elazar: 1987). Often they do so to resolve self-determination disputes. The best known “federal political system” in English is described as a “federation,” which will be defined below. It is important, however, to recognize that political systems that are not explicit “federations” are sometimes described as “federal political systems.” Some argue, for example, that “union states,” such as the United Kingdom of Great Britain and Northern Ireland, the Kingdom of Spain, and the Republic of India, are in practice, if not in precise legal description,  “federal political systems” (Stepan: 1998; Stepan, et al.: 2011). Such union-states are certainly examples of “territorial pluralism” (O'Leary and McGarry: 2012), but as we shall see their centralization of sovereignty precludes them being legally described as federations. Other political systems that are sometimes described as “federal” include “confederations” (Forsyth: 1981); “federacies” (Anderson: 2012; O'Leary: 2005; Stepan: 2013);  “cross-border inter-governmental institutions” (McGarry and O'Leary: 2005), and condominiums (O'Leary: 1993). Below, however, the focus will be on federations.

Any territorial entity with the right to self-determination, i.e. a place and its people that have the right to determine their political status, may exercise that right to become part of a federal political system. Sometimes it is recognized that the relevant entity also has the right to dissolve its membership of a federal political system, that is to secede, though lawyers attached to the federal system usually argue that the right does not exist unless it is expressly specified at the moment the system is formed, or stipulated within the federation’s constitution. There have been federal political systems which expressly embedded the right of secession, e.g. that of the former Soviet Union. Contemporary Ethiopia furnishes another example. There have, however, been other federal systems where the right of secession has been asserted but crushed in protracted civil war. Nigeria and the United States respectively violently blocked the secession of Biafra and the Southern Confederacy. Secession from federal political systems has been successfully defended in arms, usually with external assistance, e.g. the secession of Bangladesh from Pakistan, but sometimes without, e.g. the secession of Slovenia and Croatia from Yugoslavia, and most of the secessions from the former USSR. In some federal political systems the constitution is silent on the right of secession. In others, the preamble references self-determination or the voluntary status of the federal union: controversy then centers on whether the preamble is an integral component of the constitution. However the constitution of a federal system is constructed some will insist that a voluntary union owes its members the right of dis-union. Reasonable rules may be attached both to accession and secession. Canada’s Supreme Court, for example, has sought to regulate how a referendum may be held to enable a province to secede from Canada, and to specify the obligations on the relevant parties to conduct reasonable negotiations.


A federation may be defined as a political system in which at least two territorial levels of government share sovereign constitutional authority over their respective division and joint share of law-making powers; differently put, neither the federal government nor the relevant federative entities may unilaterally alter one another’s powers without a process of constitutional amendment in which both levels of government participate. A confederation in which the member-states can strip confederal institutions of their delegated powers without their consent is therefore not a federation. Likewise, a unitary state in which the central government may lawfully destroy the delegated rights or reconstitute the forms of local or regional governments without their express consent is not a federation. Joint participation in sovereign authority by both the federal government and the federative entities is the hallmark of a federation.

Within federations, sovereignty is constitutionally divided and shared between a federal government (which normally has federation-wide authority over certain functions) and what are variously known as regions, provinces, states, länder, cantons, republics, or entities. Regions are used below as the generic word for the federative entities. In a federation, the federal government and the regions respectively have exclusive responsibility for certain functions, though some powers may be shared. The division and sharing of powers is entrenched in a constitution, which requires the consent of the regions for amendments–though not always their unanimous consent. For example a constitutional amendment may require the support of a qualified majority of the federative entities.

“Non-centralization” emphatically distinguishes between federation and the devolution or delegation of authority (Elazar: 1994 9). Within federations powers rightfully belong to both the federation and its regions. By contrast, under devolved systems powers are delegated to provinces and in principle may be revised or reversed by the central government.  For this reason, among others, India, Spain, and the UK cannot be regarded as federations (pace Stepan, et al.: 2011). Spain’s constitution emphasizes the indivisibility of its sovereignty; India’s constitution has been described as an indestructible union of destructible states; and the United Kingdom has expressly reserved the unqualified legal supremacy of the Westminster Parliament over the devolved Scottish Parliament and the Welsh Assembly.

In federations, an impartial judicial tribunal normally decides constitutional disputes, though exclusively political means may also be used to resolve intergovernmental conflict, e.g. an inter-governmental conference of ministers.

Federations standardly have bicameral legislatures within the federal government. One chamber represents the federation’s demos, normally in proportion to population. In the chamber of the regions, by contrast, the smallest component units are usually overrepresented. Laws are made by the consent of both chambers, ensuring a “double majority”: the consent of the people as a whole and of the regions. In federations, regions usually cover the entire state’s territory, with exceptions sometimes being made for capital cities, or sparsely inhabited territories. The scope of self-government enjoyed by regions in federations varies: some federations are less “non-centralized” than others. In some federations the regions enjoy very extensive powers of self-government; in others the federal government has multiple opportunities to extend and even usurp powers it has been granted within the constitution––especially if it is aided by a compliant supreme or constitutional court.

Historical Evolution

Territorial pluralism was typical in the “agro-literate polity” (Gellner: 1983 9 ff.), in “pre-industrial societies” (Crone: 1989), and in the lands of “composite monarchies” (Elliott: 1992). In these political systems, normally governed by dynasties, national, ethnic, linguistic and even religious homogeneity was rare. The Ottoman and Habsburg Empires are examples. Indirect rule (or what we might call undemocratic autonomy) was commonplace before the modern era of nationalism (Hechter: 2000 Ch. 3; Mann: 1984). It is therefore territorial monism, not territorial pluralism that is modern. Territorial monism is the attempt to make all citizens within an extended territory culturally homogeneous in language, religion and presumptive ethnicity, and subject to a supreme sovereign center.

Territorial monism has never swept all before it though many have indeed tried to follow the Jacobin model of the unitary, centralized state in which sovereignty is “indivisible.” The empires that dominated the globe at the beginning of the twentieth century were not territorially monist, either at home or abroad. Subsequently federal, confederal and union-in-diversity arrangements were never eliminated from the repertoires of institutional design, either in the democratic capitalist or the communist worlds.

The United States was the first modern federation in which the federal government in principle could exercise federal government within its member-states on matters assigned to the federal government. Executive power was exercised by its President, legislative power by its Congress; and the Supreme Court eventually policed the separation of powers between the branches of the federal government and between the federal government and the states. The US model, though influential, is no longer typical in the design of federations.

Parliamentary federalism first originated as a conflict-regulating strategy, intended to address potential self-determination disputes, in the mid-nineteenth century, with the creation of federations in Switzerland and Canada (in 1848 and 1867), and in the re-making of the Austria and Hungary in 1867 under the dual monarchy. Since the 1970s several democratic states, including Belgium, the United Kingdom, Italy, and Spain, formally unitary or union states, have reconstructed themselves: some of their constituent nationalities or people now enjoy some territorial autonomy or self-government. Belgium is now a federation; the others are more decentralized than they were; and there are often variations (asymmetries) in the powers enjoyed by different regions. Territorial pluralism allows nationalities, big or small, some autonomy, and, in some variations, facilitates power-sharing within the central or federal government. “Territorial pluralism” has developed a significant constituency in the western academy, mainly in law schools and political science departments. Not surprisingly it is popular among political parties from minority communities. 

Theoretical Implications

Types of Federation.

Federations may usefully be distinguished in three ways.

  • First, by whether they are national or multi-national. A national federation may recognize just one nation in the federation; and may insist that because there is just one nation or demos a right of secession cannot apply. That is how the United States of America, after its civil war, is understood, as a national federation. The multi-national federation, by contrast, recognizes several nations; with the limiting case of the bi-national federation that recognizes two. Such federations are more likely to concede the right of secession.

  • Second, by the modes of decision-making practiced within the federal government. The majoritarian federation empowers a federation-wide majority with ample powers to make new laws and policies. The consensual federation, by contrast, requires qualified or super-majorities for the successful passage of laws and policies.

  • Third, by the powers granted to the federal government. The centralized federation strongly empowers the federal government, granting it exclusive powers over the military, foreign affairs, fiscal and monetary policy; and distinct emergency and residual powers. Where there is a clash of competences, either the courts rule in favor of the federal government, or the federal government has express legal supremacy. The decentralized federation, by contrast, grants the regions significant power to constrain the federal government or to participate in key federal policy-making; these regions retain residual powers, and when there are clashes over competences they may prevail over the federal government.

We may call the national, majoritarian and centralized federation an integrated federation. Its polar opposite is the multi-national, consensual and decentralized federation, or a pluralist federation. At their respective limits a national federation resembles a unitary state while the pluralist federation resembles a confederation.

Some academics and policy-makers reject pluralist federations, strongly preferring the integration or the assimilation of national, ethnic, religious and linguistic minorities. They fear that the pluralist form promotes ethnocentrism within “titular nationalities,” and discrimination against minorities within autonomous territories. They claim such federations sacrifice individual rights to group demands (e.g. Brubaker: 2006; Wimmer: 2002; Wimmer: 2003-4). Traditional socialists and social democrats worry that pluralist federations undermine solidarity and block redistributive transfers. Conservatives, political unionists, and analysts who prefer integration and the separation of powers to power-sharing, believe pluralist federations (which they invariably call “ethno-federations” are de-stabilizing (Roeder: 2007), and will promote secessionist break-ups (e.g. Bunce: 1999; Nordlinger: 1972; Snyder: 2000). The record of pluralist federations is certainly mixed. For every Switzerland, Canada, and Belgium there is at least one Austro-Hungary, Soviet Union, and Yugoslavia. Establishing whether pluralist federations may work and evaluating their virtues and flaws, are therefore key subjects for the political and legal sciences, and for the management of self-determination disputes (Weller, et al.: 2008).

Practical Applications

Federations are formed in three ways, and with permutations thereof (Stepan: 2001 Ch. 15). First, they can be formed from previously independent states, including from a confederation of independent states that decide to “come together.” This route was followed in Switzerland and the United States: one that some hope (and others fear) that the European Union is embarked upon (Heinemann-Grüder: 2002; Majone: 2005 Ch. 10; Majone: 2009 Chs. 2-3; Wistrich: 1994). Second, a federation can develop through the transformation of a unitary or union state, to “hold together,” as has happened in Belgium (Hooghe: 1991; Hooghe: 2004; Fitzmaurice: 1996). Canada’s birth was a hybrid of these two processes. It joined a number of previously separate British colonies, but re-divided upper and lower Canada into the separate provinces of Ontario and Quebec (McNaught: 1988.)  The third formation process is a federation “put together” by force (Stepan: 1999). Stepan has in mind the Soviet Union, established by the Red Army, and Yugoslavia, re-put together by Tito’s partisans. But Communists have not been alone in “putting together” federations. The British empire made them in Nigeria, in southern and east Africa, and in Malaysia (Suberu: 2001; Suberu and Diamond: 2002; Hicks: 1978; Trager: 1968). The US is sometimes held to have  “put together” two federations, one in Bosnia and Herzegovina, and one within one of its two constituent entities (Cohen: 2002; see also Holbrooke: 1998). The prospects for coercively put-together federations, under conditions of democratization and independence, are not good (Stepan: 2001; McGarry and O'Leary: 2005). Contemporary Iraq’s status is debatable: it can be read as “put together” under US tutelage after 2003; it can be seen as a “coming together” in which Kurdistan rejoined Iraq; and the Constitution of 2005 may also be defended as a “holding together” operation (Deeks and Burton: 2007; McGarry and O'Leary: 2007).

Table 1 presents the integrated and the pluralist federation as polar types (adapted from O'Leary: 2005). An integrated federation is precisely one in which minorities do not have homeland regions that they control as local majorities. In a fully integrated federation no minority has democratic control over its own homeland: “reservations” for the dependent and unequal micro-nationalities of tribal or nomadic peoples do not disturb the integrative structure of national federations. In the United States, Germany, Australia, and the Latin American federations of Mexico, Brazil and Venezuela, the federation-wide majority is a majority in every federal unit. The formation of federations in these places had nothing to do with the accommodation of minorities. Indeed, the establishment or maintenance of such national federations has often been accompanied by deliberate efforts to prevent minorities from becoming self-governing. In the southern states for a century after slavery was abolished, American federalism facilitated the racist disorganization rather than the egalitarian integration of African Americans–who did not obtain an enfranchised majority in any state.

Federal Traits Integrated Pluralist
Table 1. Integrated and Pluralist Federations (adapting O’Leary 2005)
Nation(s) recognized Mononational Multinational
Decision-making in the federal government Majoritarian Consensual (qualified-majority voting provisions)
Allocation of Powers, especially legal supremacy, fiscal and security

Centralized, especially characterized by

federal legal supremacy
fiscal centralization,
strong federal security powers

Decentralized, especially characterized by

regional legal supremacy,
fiscal decentralization,
strong regional security forces

Ethno-Territorial Structure All regions are dominated by the Staatsvolk or no regions are dominated by a minority nationality, ethnicity, religion or language group All major nationalities control their own homeland or the Staatsvolk is divided across more than one region whereas other minorities are mostly encapsulated in their homeland regions
Conflict-regulation goal Integration Accommodation

A fully pluralist federation, by contrast, is bi-national or multi-national (sometimes the expression “pluri-national” is preferred because it includes all cases where N is > 1). Its ethno-territorial structure expresses nationalist content in the strongest form: matching nation to homeland.  In a pure multinational federation each major nationality is encapsulated within its own homeland that is united in a single regional government. A weaker and fairly frequent federal form of a pluralist ethno-territorial structure is one in which the Staatsvolk, or the dominant people, is divided among multiple regions, whereas other smaller nations have their own regions. Canada and Russia reflect this pattern, though under President Putin the latter has rescinded its previous extensive decentralization of powers. There have been very few democratic and fully pluralist federations: Belgium and Canada are the clearest examples in the long-established democracies; Iraq and Bosnia and Herzegovina are more recent and unconsolidated versions. The multinational character of a pluralist federation may be recognized in the state’s constitution, in its flag and symbols, to emphasize that it is a partnership among nationally distinct peoples. Iraq’s 2005 Constitution stipulates in Article 3 that it is a “country of many nationalities.”  Article 4 makes Arabic and Kurdish the country’s two official languages, while Article 12 (1) stipulates that, “the flag, national anthem, and emblem of Iraq shall be fixed by law in a way that represents the components of the Iraqi people.” A fully pluralist federation involves distinct territorial autonomy for the partner nations. By contrast, the deliberate partition of minority nationalities across multiple regions is intended to prevent the possible formation of secessionist entities. Tito’s placement of Serbs and Croats outside the newly defined Serbian and Croatian regions reflected existing demographic realities but was also intended to inhibit secessionism within the Yugoslav federation. Nigeria’s military dictators decided to partition Hausaland, Yorubaland, and Iboland–after the defeat of Biafra–precisely to terminate the possibility of a future break-up.

Table 1 deliberately represents “polar types” of federation. It is possible, however, to conceptualize and find federations with both integrative and pluralist elements. Switzerland, for example, seems prima facie to be a pluralist federation, but in fact, and in its norms, it is not multinational. It is a federation of cantons, not of nations; it is multilingual and bi-religious (it was re-founded in 1848 by Protestant victors and Catholic losers after a civil war); it is multiethnic, not least because of its large number of diverse “guest-workers” who do not have full citizenship rights; but it is not multinational. Its speakers of German (in fact Swiss Deutsch), French and Italian are not legally regarded, and do not regard themselves, as part of any nation other than the Swiss. It is therefore useful to recognize multicultural forms of accommodation that fit between the integrative and pluralist federations. Such “multicultural federations,” however, have no consistent typological features. Situated between the pluralist and integrated types, they vary across all the dimensions highlighted in Table 1, and in other ways which there is no space to discuss here. They are more likely to have consensual, consociational, or qualified majority voting rules within the federal government; to ensure that the federal civil service is representative of the federation’s diversity; to create strong second chambers representing the constituent regions; and to have strong regional judiciaries and a regional role in the selection of federal judges; and to tolerate asymmetrical arrangements (Weller and Nobbs: 2010.)

A pluralist federation is distinct from a confederation, though the two are often conflated or confused. For example, Stepan et al (2011 10) regard a pure multi-national federation as highly “improbable”, and indistinguishable from a confederation. But if the European Union formally federalizes it would be a pluralist federation. A federation is a state with shared citizenship and a single international personality in which sovereignty is shared; while a confederation is typically a union of (independent) states, established usually for a limited set of purposes, such as defense or economic cooperation, and in which full sovereignty is retained by the member-states.  The federal governments of federations have a direct role in the lives of their citizens, marked by the presence of their agencies within each region, while confederal authorities normally interact indirectly with the citizens of member-states. Since confederations are generally looser unions rather than federations, they are more likely to have decision-making rules based on unanimity, and to permit opt-outs, and indeed secession. Some pluralist federations, notably Belgium, allow their constituent units a role in international relations. Both Canada and Belgium permit regions with French-speaking populations to sit in La Francophonie, the league of French-speaking state From the other direction, the European Union, which originated as a confederation, has developed some federal characteristics (Weiler: 1999). Since the Maastricht Treaty, there has been EU citizenship, and both the European Central Bank and the European Court of Justice make direct decisions affecting persons and organizations inside member-states. The EU’s formal normal decision-making rule within the Council and the Parliament has shifted from unanimity to qualified majority rule. 


Integrated federations do not accommodate nationalists and linguistic, ethnic or religious minorities according to the latter’s most-preferred terms. There have, however, been ‘sham’ or pseudo-pluralist federations, where the state is organized ostensibly as a federation, and smaller nationalities are majorities within federal regions, but there is no genuine self-government.  Such “federations” exist when federal regions are either not autonomous, i.e., the constitutional division of powers is ignored in practice; or where there are none of the democratic freedoms necessary for self-government. The Soviet Union was the most prominent sham pluralist “federation” in the twentieth century.  Though its state structure was federated from early on, real power lay in the tightly centralized Communist Party (the CPSU), which operated according to “democratic centralism.”  The Soviet Socialist Union Republics were not autonomous. The Soviets, in theory elected by local populations, were nominated and controlled by the CPSU. Key institutions, including the army police, and intelligence services, were controlled from Moscow. No effective judicial review arbitrated the division of functions between the center and the republics. While the Yugoslav federation was, at times, less centralized than its Soviet or Czech counterparts, it was a one party dictatorship and hostile to national self-government by its constituent peoples.  Communism, however, has had no monopoly, on sham federations. Nigeria and Pakistan during their long bouts of military dictatorship have been sham federations. Putin’s Russia, in which governors are appointed by the center, and regional parties banned from competing in federal elections, has moved substantially in this direction. Ethiopia, while structured formally as a multi-ethnic federation, is a one-party authoritarian state, run by the Ethiopian People's Revolutionary Democratic Front, and lacks genuine pluralism.


Secessionists argue that pluralist federations are unworkable, or at least that it is so for the one they live in. They believe that the nation-state is best; that is why they seek one of their own. Unitarists and integrationists also dislike pluralist federations. They claim that the pluralist forms promote illiberalism, artificially institutionalize narrow identities, and exacerbate conflict, leading, in many cases, to break-up. The late Eric Nordlinger (1972) rejected pluralist federations because they would allow the region’s “dominant segment to ignore or negate the demands of the minority segment,” and facilitate secession. Some American scholars of the communist federations (e.g. Philip Roeder (2007)) observe that they privileged ‘titular’ nationalities over the others within their regions. The popularity of these views among liberal political scientists in the US reflects their knowledge of how Southern whites used “states rights” to maintain slavery and later the ‘Jim Crow’ segregationist regime.  Invoking “states rights” is still often barely disguised code for white supremacism (Smith: 1997). Recently, a Swiss academic educated in the US worried that “ethnically-based” autonomy in Iraq “may heighten, rather than reduce the risks of gross human rights violations, especially for members of ethnic minorities living under the rule of the majority government in a federal unit” (Wimmer: 2003-4 123). This outlook informed his recommendation, shared by many American policymakers, that Iraq needed an integrated not a pluralist federation.

Such fears are understandable, but they are not persuasive, or decisive. Proper evaluation requires case-by-case study, and an assessment of the positive and negative cases. Nationalities of a smaller size are not a priori illiberal, just as those of a greater size are not a priori liberal. Some territorially dominant nationalities, such as the Flemish, provide uncomfortable levels of support to far-right and xenophobic parties, but this example does not make a rule, and it is unclear why they would become more liberal if they lost their territorial autonomy. Potential holders of autonomous territorial power should not be assumed to be like southern American whites used to be. Several contemporary nationalities, including the Catalans, Quebecois and Scots, talk of civic rather than ethnic nationalism, and appear reasonably tolerant of minorities in their midst. Quebec developed its own provincial charter of rights and freedoms, and it has been used to strike down discriminatory provincial legislation. The Kurds of Iraq have treated minority Arab-speaking Christians better than have Arab Iraqis, demonstrated in the flight patterns of Christian refugees. Kurdistan’s minority nationalities would not be better protected by a centralized Iraq, given Baghdad’s historic maltreatment of minorities.  Differently put, all territorial governments, at all levels, are capable of promoting or undermining human rights.

The rights of nationalities, minorities and individuals can be protected without denying autonomy to the relevant region, or providing the center radical powers of intervention through its courts or other public bodies. Under some conditions, pluralist federations may in fact be more consistent with the promotion of individual liberties than the alternatives proposed by their critics. Any reasonable understanding of human rights and equality requires respect for the culture of individuals, and such respect requires allowing nationalities and cultural minorities the power to protect and promote their culture––though not to promote illiberalism (Kymlicka: 2007). The rights of women and micro-minorities can be protected by regional bills of rights, courts, quotas and qualified majority-voting. Federalism may also provide an excellent check on centralized despotism.

The most apparently pervasive argument against pluralist federations is that they institutionalize and thereby deepen national, ethnic, linguistic and religious divisions (Brubaker: 1996; Brubaker: 2006; Luff: 1999; Roeder: 2007; Snyder: 2000; Snyder and Ballantine: 1996-7). Critics maintain they obstruct the solidarity needed to support redistributive programs; that they promote group-centered identities; and that they occasion instability and conflict, and facilitate the break-up of states. These views may appear plausible. A region has a territorially defined electorate.  Local parties will emphasize regional interests.  Pluralist federations lend material and symbolic support to national identities. They provide legislatures and executives, and thereby resources, that may be controlled by local nationalists. They will be able to call for referendums on independence. Its own ‘national’ government, legislature, and flag may strengthen the identification of the nationality with its region and its ‘map image’ showing territorial boundaries. Regions will typically seek to expand their powers, to haggle over the allocation of resources, and to blame problems, including economic downturns, on the federal government.  Such conflicts may be magnified when parties representing different national communities control the federal and regional units of government. On Roeder’s account (2007) almost all nation-states have their origins in previous autonomy arrangements. All communist federations broke up along the lines of their internal federal boundaries. Since 2000, Montenegro has seceded from Serbia, and Kosova has done so though it is not yet recognized at the UN. South Sudan seceded from Sudan in July of 2011. The international community holds Bosnia and Herzegovina together. There are significant secessionist movements in Quebec, the Basque Country, and Scotland.

Supporters of pluralist federations, however, contest the implicit premise of the critics that there would be solidarity, stability and territorial unity without territorial pluralism. Canadian, British and Turkish integrationists are not credible when they claim that the Quebecois, Irish or Kurds respectively would (and should) be happy with centralization and equal citizenship. Supporters of pluralist federations deny that nationalist mobilization flows from territorial pluralism. It is rather the other way around: territorial pluralism is an attempt to accommodate nationalist sentiment and mobilization. Why after all are regions recognized? Nationalist movements, all over the world, have contested centralized regimes especially when they have a democratic opportunity. There are many cases where pluralist federal designs are plainly a response to multinational and multicultural divisions rather than the cause of them. They argue that the critics tend to be highly selective in their choice of cases of break-ups.

It is also questionable to overlook the success stories. Two of the world’s oldest regimes, Switzerland and Canada, are in one case a multicultural federation with consociational features, and in the other a fully pluralist federation in practice.  They have endured from 1848 and 1867, respectively. India, the world’s largest democracy and its most successful post-colonial democracy, has developed territorial pluralism within its union-state, and its major failure in Kashmir arguably flows from insufficient pluralism rather than the converse. India’s comparative success is explained away by one thinker because of the unwillingness of its central authorities to recognize ethnicity (Snyder 2000: 287-96), which is an odd way to describe a union state which reconstructed its boundaries largely along ethno-linguistic lines in the 1950s. Belgium’s unity is brittle, as its crises over government formation attest, but more integrationist formulae do not have strong constituencies.  None of these states seem bound to fall apart, though they are not guaranteed to stay together. One of Ted Gurr’s main findings, after a large-N study of ethnic hotspots, was that territorial autonomy for minorities has been an effective remedy, and this was also Michael Hechter’s conclusion in another important study (Gurr: 1993; Gurr: 2000; Hechter: 2000), and that of Liam Anderson (Anderson: 2012; Anderson: 2012). Separate studies suggest that timely granting of territorial autonomy in conjunction with power sharing at the center facilitates stable arrangements (McGarry and O'Leary: 2009; Cederman et al 2015).


Given the mixed record of federations in regulating self-determination disputes, the proper question is not whether they are bound to fail or succeed, but rather under what conditions are they likely to succeed or fail?   

History matters, even if it is not destiny. Past coercion and conquest, totalitarian or authoritarian governments, and the historic maltreatment of nationalities, jointly and severally render free political accommodation more difficult. The democratization of coercive regimes may lead to the springtime of the oppressed nations, as they break free, rejecting the uncertainties of territorial pluralism for the potential security of independence. In such scenarios dominant nationalities, with poor track records of accommodation, find it difficult to make a persuasively credible commitment to federalism, which therefore stands a better chance if it emerges from past voluntary alliances, and in conjunction with gradual democratization.

The long-term survival of federalism is more likely when there are “nested” or “complementary” identities among the territorially concentrated nationalities, i.e. where there is some sense of allegiance to the whole state as well as to the national homeland (Stepan et al 2011). Nested or complementary identities are unlikely to flourish where the relevant regional nationality has been harshly treated by the state, or where violent wars have raged. Equally, complementary identities may be less likely to exist, or more difficult to nurture, if a region was forcibly incorporated, or reincorporated, into the state, rather than voluntarily acceding. The Baltic republics were the most eager secessionists from the former Soviet Union. The Red Army forced the former into the Soviet Union in 1940-45, and, before that, Tsarist armies forced them into Russia. 

The importance of complementary identities suggest that states should react early to popular demands for territorial pluralism, before relations between nationalities deteriorate.  The adoption or modification of territorial pluralism in the United Kingdom, Spain, Canada, Belgium, India, Indonesia and Switzerland suggests that politicians are sometimes capable of engaging in self-denying prophecies; they are willing to “re-shape” themselves to avoid forcible “down-sizing” (O'Leary, et al.: 2001).

With sizable nationalities, federations may work best when accompanied by consociational (or consensual) power sharing within federal institutions. Such provisions make it more likely that minorities retain a stake in the state.  Power sharing may be constitutionally entrenched, or, less attractively from the minority’s perspective, it may flow from political conventions or general election results that allow minorities to be pivotal players in legislatures. Inclusive federal governmental institutions and conventions have helped keep the Canadian, Swiss and Belgian federations together. Spain and India have been at their most cordial as union-states when minority nationalities have enjoyed influence in Madrid and New Delhi. The absence, or collapse, of inclusive federal or central institutions has, by contrast, contributed to breakdown and secession. Yugoslavia broke apart after its Federal Council was taken over by aggressive Serbian nationalists. Nigeria’s breakdown followed a coup led by Ibo officers and a counter-coup in which these officers were overthrown. Much of Nigeria’s post-1970 armed conflict, including sectarian warfare between Muslims and Christians and the rise of violent separatism in the oil-rich Delta area, has been traced to the lack of inclusiveness at the center (Suberu: 2001; Suberu and Diamond: 2002.) The breakdown of the West Indies Federation was linked to Jamaicans’ lack of representation and influence at the center, while the federation of Nyasaland, Northern and Southern Rhodesia, under-represented black Africans, and privileged white settlers (Hicks 1978). The Constitution of Iraq of 2005 has also arguably collapsed as a result of the failure of the newly dominant Shiite-majority to include both the Kurds and the Sunni Arabs within a genuine power-sharing center.

Federations with a strong dominant nationality, or a Staatsvolk, are less likely to break apart than those without (O'Leary: 2001), but federations that spread or divide their dominant nationality (not their smaller nationalities) across multiple territories are also more likely to endure (Hale: 2005; Hale: 2004). The former proposition holds partly because such dominant communities can deter or prevent secession, and yet may be secure enough to make territorial accommodations. What the Staatsvolk does with its dominance may be as importance as its mere existence. When it coerces, supports a totalitarian party or religion, maltreats minority nationalities, or supports centralist coups and putsches, it generates antagonistic rather than complementary identities. When it is liberal, democratic, rights-respecting, and open toward accommodation the dominant group may find co-operative partners.

The number of autonomous or federative units affects stability though not in obvious ways.  Two-unit, or dyadic, pluralist federations have an abysmal track record (Vile: 1982), with Serbia and Montenegro, and the two Sudans being merely the latest casualties.  This fact may be because divisions in dyadic federations converge on and reinforce the same axis of cleavage, with floating coalitions rendered very difficult. Belgium’s continuing survival may owe something to the fact that it is not a dyadic federation of Flanders and Wallonia: Brussels’ existence as a separate federal unit inhibits the secession of Flanders.  Canada’s stability and longevity may be helped by the fact that it is not a two-unit federation of Quebec and English Canada. The division of the English-speaking Canadian population into nine provinces has created a multiple balance of power, less threatening to Quebec, and makes possible inter-provincial alliances that cross-cut the ethno-linguistic cleavage. The advantages of a multiple balance of power are not, however, an argument for imposing such through a strategy of sub-dividing regional units dominated by distinct national communities. If a multiple balance of power develops organically, as in Canada, that is helpful. Some dominant communities may accept the division of their territory into several regions, provided their aspiration to collective self-government can be expressed within the union, or federal-level, institutions. Minority nationalities will, by contrast, stoutly resist, if they can, any imposed partition of their autonomous region.    

Economic prosperity, appropriately dispersed, enhances the prospects that a federation may survive and indeed thrive. It may allow redistribution from wealthy regions to the less wealthy, binding the latter to the state without incurring the rancor of the former.  But the example of India, until recently not an example of deepening prosperity, suggests that material–or shared–prosperity is a facilitative, not a necessary condition for stable territorial pluralism. The converse hypothesis is that severe distributive conflicts – over natural resources or fiscal or tariff or subsidy policies – especially if they coincide with national, ethnic, linguistic or religious allegiances may overload federations.

The prospects that federations may survive may be enhanced, even in otherwise unfavorable circumstances, where international agents, including nearby powerful states, have a strong interest in holding the relevant state together.  Bosnia and Herzegovina will not break up as long as NATO and the European Union are so determined.  Cyprus may re-federalize if it is in the clear interests of Ankara. If Iraq’s federation survives it will partly be because the United States, Turkey and Iran remain opposed to an independent Kurdistan. 


This analysis has attempted to provide a detailed exploration of federal means of addressing national self-determination disputes. Federalism offers no panacea for national, ethnic, religious and linguistic conflicts; politics never ends; and federations have their known pathologies or sore-spots: e.g.  “asymmetries;” over-and under-represented regions in the federal legislature, especially the second-chamber; the clash of ‘representativeness’ and ‘inclusion’ with abstract notions of ‘desert’ and ‘merit’; the clash of multiple citizen bodies as opposed to one; and the difficulties in managing just and viable patterns of fiscal transfers. Federations have, however, strong, and frequently more important advantages. They offer some prospects of accommodating multiple nationalities, religions, languages, and ethnicities, with their consent, in acts of joint self-determination. Pluralist federations reject, in principle, coercive assimilation and control (repression). More controversially, they reject integration, i.e. the idea that mobilized nationalities can be satisfied with individual (personal autonomy) rights without public accommodation of their nationality or culture. Integration is, arguably, the world’s dominant method of conflict regulation, but there is little evidence that it works quickly or in the long run in heavily multinational and multicultural places. Federations do not, guarantee stability but they may offer the best inoculation against secession available, as well as creative ways of achieving parity, proportionality and autonomy that respect the rights of peoples.

References / Further Reading

Amoretti, Ugo M., and Nancy Bermeo, Eds.  2004. Federalism and Territorial Cleavages. Baltimore: John Hopkins.

Anderson, Liam. "Autonomy and Federacy." In Federal Solutions to Ethnic Problems: Accommodating Diversity, 219-48. New York: Routledge, 2012.

———. "Ethnoterritorial Federalism: A Compromise Model?” In Federal Solutions to Ethnic Problems: Accommodating Diversity, 164-218. New York: Routledge, 2012.

———. "In Defense of Ethnic Federalism." In Federal Solutions to Ethnic Problems: Accommodating Diversity, 52-99. New York: Routledge, 2012.

Brubaker, Rogers. Ethnicity Without Groups.  Cambridge: Cambridge University Press, 2006.

———. Nationalism Reframed: Nationhood and the National Question in the New Europe.  Cambridge: Cambridge University Press, 1996.

Bunce, Valerie. "Peaceful versus Violent State Dismemberment: A Comparison of the Soviet Union, Yugoslavia and Czechoslovakia." Politics & Society 27, no. 2 (1999): 217-37.

Cederman, Lars-Erik, Simon Hug, Andreas Schádel, and Julian Wucherpfenniig. 2015. "Territorial Autonomy in the Shadow of Conflict: Too Little, Too Late?" American Political Science Review 109): 354-70.

Cohen, Lenard J. "Fabricating Federalism in "Dayton Bosnia": Political Development and Future Options." In Federalism Doomed? European Federalism between Integration and Separation, edited by Andreas Heinemann-Grüder, 116-45. Oxford: Bergahn Books, 2002.

Crone, Patricia. Pre-Industrial Societies.  Oxford: Basil Blackwell, 1989.

Deeks, Ashley S., and Matthew D. Burton. "Iraq's Constitution: A Drafting History." Cornell International Law Journal 40, no. 1 (Winter 2007).

Elazar, Daniel J. Exploring Federalism.  Tuscaloosa: University of Alabama, 1987.

Elazar, Daniel J. Federalism and the Way to Peace.  Kingston, Ontario: Queen's University Press, 1994.

Elliott, John H. "A Europe of Composite Monarchies." Past and Present, no. 137 (1992): 48-71.

Fitzmaurice, John. The Politics of Belgium: A Unique Federalism.  Boulder, Co: Westview, 1996.

Forsyth, Murray. Unions of States: The Theory and Practice of Confederation.  Leicester: Leicester University Press, 1981.

Gellner, Ernest. Nations and Nationalism.  Oxford: Basil Blackwell, 1983.

Gurr, T. "'Settling Ethnopolitical Conflicts'." In Minorities at Risk: A Global View of Ethnopolitical Conflicts, edited by T. Gurr. Washington: United States Institute of Peace Press, 1993.

Gurr, Ted Robert. People Versus States: Minorities at Risk in the New Century.  Washington, DC: United States Institute of Peace, 2000.

Hale, Henry E. "Divided We Stand: Institutional Sources of Ethnofederal State Survival and Collapse." World Politics 56, January (2004): 165-93.

———. "The Makeup and Breakup of Ethnofederal States: Why Russia Survives Where the USSR Fell." Perspectives on Politics 3, 1 (2005): 55-70.

Hechter, Michael. Containing Nationalism.  Oxford: Oxford University Press, 2000.

Heinemann-Grüder, Andreas, ed. Federalism Doomed? European Federalism between Integration and Separation. Oxford: Bergahn Books, 2002.

Hicks, Ursula K. Federalism, Failure and Success: A Comparative Study.  London: Macmillan, 1978.

Holbrooke, Richard. To End a War.  New York: Random House, 1998.

Hooghe, Liesbet. "Belgium: Hollowing the Center." In Federalism and Territorial Cleavages, edited by Ugo M. Amoretti and Nancy Bermeo, 55-86. Baltimore: John Hopkins, 2004.

———. A Leap in the Dark: Nationalist Conflict and Federal Reform in Belgium.  Ithaca, NY: Cornell University Press, 1991.

Kymlicka, Will. Multicultural Odysseys: Navigating the New International Politics of Diversity.  Oxford: Oxford University Press, 2007.

Leff, Carol Skalnik. "Democratization and Disintegration in Multinational States: The Breakup of the Communist Federations." World Politics 51, no. 2 (1999): 205-35.

Majone, Giandomenico. Dilemmas of European Integration.  Oxford: Oxford University Press, 2005.

———. Europe as a Would-be World Power: The EU At Fifty.  Cambridge: Cambridge University Press, 2009.

Mann, Michael. "The Autonomous Power of the State: Its Origins, Mechanisms and Results." European Journal of Sociology XXV (1984): 185-213.

McGarry, John, and Brendan O'Leary. "Federation as a Method of Ethnic Conflict Regulation." In From Power Sharing to Democracy: Post-Conflict Institutions in Ethnically Divided Societies, edited by SJR Noel, 263-96. Montreal: McGill-Queens, 2005.

———. "Iraq’s Constitution of 2005: Liberal Consociation as Political Prescription." International Journal of Constitutional Law 5, no. 4 (2007): 670-98.

———. "Must Pluri-National Federations Fail?” Ethnopolitics (Special Issue: Federalism, Regional Autonomy and Conflict) 8, 1 (March) (2009): 5-26.

McNaught, Kenneth. The Penguin History of Canada.  London: Penguin, 1988.

Nordlinger, Eric A. Conflict Regulation in Divided Societies. Occasional Papers in International Affairs. Vol. 29, Cambridge, Mass.: Center for International Affairs, Harvard University, 1972.

O'Leary, Brendan. "An Iron Law of Federations?  A (neo-Diceyian) theory of the Necessity of a Federal Staatsvolk, and of Consociational Rescue. The 5th Ernest Gellner Memorial Lecture." Nations and Nationalism 7, 3 (2001): 273-96.

———. "Power-Sharing, Pluralist Federation and Federacy." In The Future of Kurdistan in Iraq, edited by Brendan O'Leary, John McGarry and Khaled Salih, 47-91. Philadelphia: University of Pennsylvania Press, 2005.

———. "Sovereignty, Joint Authority and Shared Authority." In Northern Ireland: Sharing Authority, edited by Brendan O'Leary, Tom Lyne, Jim Marshall and Bon Rowthorn, 136-8. London: Institute of Public Policy Research, 1993.

O'Leary, Brendan, Ian S Lustick, and Thomas Callaghy, eds. Right-Sizing the State: the Politics of Moving Borders. Oxford: Oxford University Press, 2001.

O'Leary, Brendan, and John McGarry. "The Politics of Accommodation and Integration in Democratic States." In The Study of Politics and Ethnicity: Recent Analytical Developments, edited by Adrian Guelke and Jean Tournon. The World of Political Science: The Development of the Discipline Book Series. 79-116. Leverkusen, Opladen: Barbara Budrich, 2012.

Roeder, Philip G. Where Nation-States Come From: Institutional Change in the Age of Nationalism.  Princeton: Princeton University Press, 2007.

Smith, Rogers M. Civic Ideals: Conflicting Visions of Citizenship in U.S. History.  New Haven: Yale University Press, 1997.

Snyder, Jack. From Voting to Violence: Democratization and Nationalist Conflict.  New York: W.W. Norton, 2000.

Snyder, Jack, and Karen Ballentine. "Nationalism and the Marketplace of Ideas." In Nationalism and Ethnic Conflict, edited by Michael Brown, Owen R Cote, Sean Lynn-Jones and Steven E Miller, 61-96. Boston, Mass: MIT Press, 1996-7.

Stepan, Alfred. Arguing Comparative Politics.  Oxford: Oxford University Press, 2001.

———. "Federalism and Democracy: Beyond the U.S. Model." Journal of Democracy 10, no. 4 (1999): 19-34.

———. "Modern Multinational Democracies: Transcending a Gellnerian Oxymoron." In The State of the Nation: Ernest Gellner and the Theory of Nationalism, edited by John A. Hall, 219-42. New York: Cambridge University Press, 1998.

———. "A Revised Theory of Federacy and A Case Study of Civil War Termination in Aceh Indonesia." In Power Sharing in Deeply Divided Places, edited by Joanne McEvoy and Brendan O'Leary, 231-52. Philadelphia: University of Pennsylvania Press. 2013.

Stepan, Alfred, Juan J. Linz, and Yogendra Yadav. Crafting State-Nations: India and Other Multinational Democracies.  Baltimore: The Johns Hopkins University Press, 2011.

Suberu, Rotimi, and Larry Diamond. "Institutional Design, Ethnic Conflict Management and Democracy in Nigeria." In The Architecture of Democracy: Institutional Design, Conflict Management and Democracy, edited by Andrew Reynolds, 400-28. Oxford: Oxford University Press, 2002.

Suberu, Rotimi T. Federalism and Ethnic Conflict in Nigeria.  Washington D.C.: US Institute of Peace Press, 2001.

Trager, Frank N. "The Federation of Malaysia: An Intermediate Failure?” In Why Federations Fail: An Inquiry into the Requisites for Successful Federalism, edited by Thomas Franck, 125-66. New York: New York University Press, 1968.

Vile, Maurice J. C. "Federation and Confederation: The Experience of the United States and the British Commonwealth." In Political Co-operation in Divided Societies: A Series of Papers Relevant to the Conflict in Northern Ireland, edited by Desmond Rea, 216-28. Dublin: Gill & Macmillan, 1982.

Watts, Ronald L. "Federalism, Federal Political Systems, and Federations." Annual Review of Political Science 1 (1998): 117-37.

Weiler, Joseph H. The Constitution of Europe: "Do the New Clothes Have an Emperor?" and Other Essays on European Integration.  Cambridge: Cambridge University Press, 1999.

Weller, Marc, Barbara Metzger, and Niall Johnson, Eds. Settling Self-Determination Disputes: Complex Power-Sharing in Theory and Practice. Leiden•Boston: Martinus Nijhoff, 2008.

Weller, Marc, and Katharine Nobbs, Eds. Asymmetric Autonomy and the Settlement of Ethnic Conflicts. Philadelphia: University of Pennsylvania Press, 2010.

Wimmer, Andreas. "Democracy and Ethno-religious Conflict in Iraq." Survival 45, no. 4 (2003-4): 111-34.

———. Nationalist Exclusion and Ethnic Conflict: Shadows of Modernity.  Cambridge: Cambridge University Press, 2002.

Wistrich, Ernest. The United States of Europe.  London: Routledge, 1994.