John McGarry- Department of Political Studies, Queen's University, Kingston, ON, Canada
This article examines how and why regimes that dominate particular ethnic communities on behalf of a dominant one disguise themselves by claiming to practice accommodation (consociational power-sharing and territorial autonomy) or integration (equal citizenship with respect for private cultural differences). It also explains how to distinguish authentic accommodation and integration from the sham forms used by these regimes. The article seeks to help identify domination regimes that would otherwise be overlooked. This is important for academics. It is also important for international policymakers who seek to condemn domination and make it more difficult to maintain.
Introduction
Ethnic domination (hereafter, domination) is defined by political elites organizing their preferred ethnic community and disorganizing their rivals.1 It is clearly undesirable for those who espouse standard liberal or social democratic values. Unfortunately, it is also widespread. According to the Ethnic Power Relations (EPR) dataset, which uses a reasonable proxy for domination—the active, intentional, and targeted exclusion of an ethnic community from public power—42 of the world’s countries in 2021 had domination regimes (23%), and a total of 72 communities were dominated (Vogt et al., 2015).
One practice of domination regimes is the use of legitimation narratives that depict them positively and the dominated negatively. These narratives are used to consolidate support from outside states and international organizations and also from the polity’s dominant community.2 The narratives typically deny the regimes practice domination, instead claiming that diversity is managed in conformity with contemporary norms. Two such narratives are identified here. The first insists that the regimes practice “accommodation”; i.e., their polities are inter-ethnic partnerships built on consociational power-sharing or territorial autonomy.3 The second is that they claim to practice “integration,” i.e., equal citizenship that is blind to ethnic differences.4
Such narratives have not always been successful. When the apartheid regime in South Africa claimed in the 1960s and 1970s that it was granting self-government and then independence to its different African ethnic communities in their ‘homelands’, the policy was seen as a sham. It was rejected by the international community, albeit not until a decade or two after the policy was initiated.
In other cases, the claims have been successful. Western governments, including the United States and the United Kingdom, have cozied up to domination regimes, either because they accept the regimes’ narratives or because enough of their constituents do that their governments are not held to account for their support for such regimes. Academics have also been misled by the narratives, even some who are focused on uncovering ethnic discrimination. In assessing states at risk of genocide—the ultimate in negative discrimination—Barbara Harff accepted the Alawi-dominated Syrian regime’s integrationist narrative at face value, thereby reducing its risk score (Harff, 2003, pp. 55–6). This was at a time—2012—when the regime was initiating a civil war against an opposition drawn overwhelmingly from Syria’s dominated Sunni Arab majority, in which nearly 130,000 civilians were killed by regime forces (Syrian Observatory for Human Rights, 2020). Even the very impressive Ethnic Power Relations dataset, cited above, may have been misled in some cases. It codes Syria’s regime, and the Tutsi-dominated regime in Rwanda after 2020 as non-discriminatory toward their respective majorities, arguably because both regimes employ integrationist narratives (Vogt et al., 2015).5
If academics, human rights NGOs, external governments, and international organizations are to be able to identify such regimes and pressure them to reform, they must understand how sham forms of accommodation and integration are used to mask domination.
Domination, accommodation, and integration as distinct strategies for governing difference
Domination, accommodation, and integration are three distinct strategies for governing difference. The complete repertoire of strategies includes genocide, expulsion, assimilation, and territorial downsizing (McGarry and O’Leary, 1993).
The concept of domination used here is not the same as that which is often used in a milder and ubiquitous, sometimes trivial sense, e.g., where one community is said to ‘dominate’ others because it is demographically larger or economically stronger.6 Thus, French speakers in Quebec talk of the ‘domination’ of English speakers in Canada, even though Quebec enjoys substantial autonomy and even though French speakers enjoy formal and informal elements of power-sharing within federal institutions. The Scots and Welsh also speak of English domination in the UK and several small European nations speak of German domination within the EU. These experiences may be felt as domination, but they are not what I mean by the concept.
Rather, domination in this article results from a “strategy” by the elites of one community to organize their followers against other communities and disorganize the latter. It is not just intentional and targeted, but profoundly divisive and discriminatory, and often violent. Domination regimes can be identified by the existence of at least three practices or ‘stratagems’. First, they ethnically stack their key governing institutions with people from the dominant community (i.e., such people control and are disproportionally over-represented in such institutions, even if the people are a majority and would be able to dominate the institutions if there was proportional representation). Second, regime elites maintain an ethnically stacked security sector capable of intimidating and using discriminatory force against the dominated, and that will do so if domination is contested. Third, regime elites practice a stratagem of polarization to unify the dominant community against the others. Polarization involves not just divisive narratives that separate the dominant from the dominated but a range of culturally, symbolically, and economically discriminatory practices that favor the dominant. All three stratagems are applied persistently, i.e., over the life of the regime.
Beyond these three stratagems, of which the first is the most parsimonious way to identify a domination regime, domination regimes may resort to fragmentation, i.e., the division of the dominated into smaller segments (also known as ‘divide and rule’). They may practice co-option, i.e., the incorporation of members of dominated communities into regime institutions in ways that fall short of inclusion on the basis of equal citizenship. Co-opted individuals either occupy relatively powerless positions or are only nominally representative of their communities. The regime may also seek to make the dominated materially dependent to facilitate their compliance, or it may leverage already existing relations of dependence. It may engage in demographic engineering, policies aimed at increasing the size of the dominant community through immigration and pro-natalism, and at reducing the size of the dominated, through induced emigration or anti-natalism. 7
Finally, the focus of this article, domination regimes practice legitimation narratives that portray their actions in the best possible light, while portraying the dominated’s actions negatively.8 The functions of legitimation narratives are not only to justify the regime’s policies externally, with the aim of securing support or allaying opposition, but also to rally support from the dominant community.9 The false claim by domination regimes that they practice accommodation or integration is an important part of such narratives.
The concept of domination used here is related to several cognate terms employed in the academic literature, including “ethnic democracy” (Smooha, 2002), “nationalizing state” (Brubaker, 1996), “control” (Lustick, 1980), and “ethnocracy” (Yiftachel, 2006). Domination, as I employ it, is capacious both conceptually and in its empirical application (McGarry, 2025). It includes, as we shall see, domination by ethnic majorities or minorities, democratic or authoritarian regimes, communities that pose as ethnic or civic nationalists, and communities in every part of the globe. The cognate terms are (usually) narrower, either conceptually or in their applications by academics. Thus, ethnic democracy is limited to domination in procedural democracies and therefore to domination by majorities. Nationalizing states is limited conceptually to communities that make ethnic national appeals, although it is also broader, in that nationalizing states may practice assimilation as well as domination.10 Control and ethnocracy are in principle conceptually identical to domination but have been applied more restrictively in literature. Indeed, the applications of control, ethnocracy, and nationalizing states appear to have been completely limited, like ethnic democracy, to regimes that are based on majorities, operate in procedural democracies, and make ethnonational appeals. The geographic application of the four concepts has also been limited to Israel and Eastern Europe plus Northern Ireland and Sri Lanka. In contrast, domination is conceptually and empirically global. Examples include the Serb regime in Kosovo (1988–99), the Unionist regime in Northern Ireland (1921–72), the Sinhalese regime in Sri Lanka (1956-), the Tutsi regime in Burundi (1965–2000), the Sunni Arab regime in Bahrain (1970-), and many more (McGarry, 2025).
Accommodation involves political partnerships among communities that cohabit ethnically diverse polities. It can take two main forms. One is consociation, a form of power-sharing associated with ethnic communities. Its most important institutional pillar involves a share of executive power for the different communities, e.g., through a power-sharing cabinet in a parliamentary coalition, a collective or rotating presidency, or a hybrid semi-presidential executive (i.e., with both an executive president and a prime minister, with the latter accountable to parliament). Consociation also suggests a form of veto for the partners that makes power-sharing meaningful; proportionality in public offices, including the bureaucracy, judiciary, and security sectors; and some degree of community autonomy, either cultural or territorial. Consociations currently exist in Northern Ireland, Belgium, Switzerland, Bosnia-Herzegovina, South Tyrol, and Lebanon. Cyprus was once a consociation, during 1960–1963, and is likely to be one again if the island is reunified.
The second form of accommodation is territorial autonomy, through which minority communities acquire self-government in regions where they are local majorities and which they identify with as their homelands.11 Territorial autonomy may exist in federations in which autonomy is guaranteed by the federal constitution; federacies, i.e., asymmetric autonomy arrangements in states that are otherwise centralized or symmetrically decentralized, and in which the autonomy is guaranteed by the constitution or international treaty; or decentralized unitary states, where the autonomy arrangements are not federally guaranteed but remain at the discretion of the central authorities, though they may be protected by organic laws. Canada, Switzerland, Belgium, and Bosnia-Herzegovina are examples of federations that allow territorial autonomy to minorities; i.e., they are pluralist or pluri-national federations. South Tyrol, the Aland Islands, and Northern Ireland are cases of federacies, all of which are partly protected by international treaties. Great Britain (Scotland and Wales), Spain (e.g., Catalonia and the Basque Country), and India (e.g., Kashmir and Punjab) are decentralized unitary states in which territorial autonomy is rescindable by the central authorities.12 Accommodationists argue that integration favors ethnic majorities and is biased against minorities, at least those minorities that do not want to be integrated, i.e., that seek to have their identity and culture accommodated.
Integration is a strategy based on the idea—found in liberal, republican, and socialist thought—that ethnic identities are relatively unimportant and transient compared to identities based on the individual, the republic (nation), or class (McGarry et al., 2008). Liberal integrationists stress individual equality and an impartial or ‘difference-blind’ state that protects human rights and freedoms, including the right not to be negatively discriminated against on ethnic or racial grounds. Their republican and socialist counterparts emphasize the value of integration to enhance national unity and solidarity, respectively. In the integrationist view, domination regimes are seen as responsible for disunity, instability, and conflict. Integrationists are not assimilationists, or at least not coercive assimilationists. They celebrate the freedom of ethnic communities to practice their cultures in the private but not the public sphere: they believe a shared public culture is necessary. They encourage assimilation if it is voluntary because it helps promote integration. Integrationist scholars oppose accommodation; i.e., the public recognition of ethnic communities, because they see this as ‘dis-‘integrationist, and likely to promote instability and privileges for the recognized.
Integrationist nation-building is the dominant strategy for governing ethnic diversity in the West, particularly regarding immigrants. It is also the dominant strategy in post-colonial countries. France is perhaps the paradigmatic example of republican integration, with a relatively large public sphere that includes beaches and streets as well as public buildings. The United States and the UK are examples of liberal integrationist states regarding their immigrants. They have smaller public spheres (more liberal tolerance) than their French republican counterparts. Canada is relatively liberal in English Canada and relatively republican in Quebec.
International normative change and the masking of domination
Domination regimes frame their actions in ways that reflect norms that are prevalent internationally and regionally. Before 1945, when these norms were racist and imperialist, domination regimes openly justified their behavior in racist terms. The narratives they used ranged from defending civilizing missions—in which the dominated, portrayed as backward, barbaric, tribal, and child-like, were prepared for modernization and assimilation—to fully supremacist claims that divided polities into masters and subjects, e.g., slave systems in the US South or the Nazi empire in eastern Europe (1938–45). In these legitimation narratives, the claim was that ‘everyone’ was better off under the hierarchical order. The dominated were recipients of civilizational uplift, or ‘happy’ slaves, or, in the argument that the Nazis presented to Slavs in the territory of the Soviet Union, ‘liberated’ from Jews and Bolsheviks, or Jewish Bolsheviks.
Racist regimes had little incentive to mask domination in a world of racist empires. They were generally formal or explicit, with the subordinated community openly named as subordinate or excluded in legal codes as well as speeches. Under the penal laws of the “Protestant Ascendancy” in eighteenth-century Ireland, for example, Roman Catholics (a.k.a. “Papists”) had their religious practices formally proscribed. They were explicitly excluded from the franchise, parliament, bearing arms and armed service, most forms of public office, and the legal professions (O’Leary, 2020, Vol. II, pp. 193–8). The Ottoman Empire formally privileged Muslims in relation to other “people of the book” (i.e., the dhimmi, mainly Christians and Jews) and these people over others (Ye’or, 2013). The Nazis formally discriminated against “non-Aryans” through the Nuremberg Laws and other statutes (Friedländer, 1998, Vol. 1).
Norms supporting racist domination receded decisively after the victory of the Allies over the Axis powers in World War II. The defeat of Nazi Germany in 1945 and the revulsion caused by Nazi crimes helped contribute to a wave of decolonization from 1947 and to the success of the US Civil Rights movement from the mid-1940s till the mid-1960s. In turn, both latter movements helped consolidate the new norms. Domination did not disappear in this new context of self-determination, universal individual human rights, and the international convention on genocide, but the regimes that practiced it changed their narratives. Let me illustrate the pattern.
Domination in accommodationist guises I: sham autonomy
Several domination regimes have sought to disguise domination through sham autonomy. The autonomy is a sham because it does not entrust political functions to authentic representatives of the communities concerned, delegate significant powers, or cover the territory in which the relevant subordinated community seeks self-government, or all three at once.
Three cases of sham autonomy are worth attention.
(i) The Soviet Union was not a typical domination regime because its ruling elite were communists who rejected racism and ethnocentrism and sought a global socialist revolution.13 However, the Bolsheviks held their huge multinational polity together by institutions and policies that privileged its dominant Russian community, while nominally creating a multi-tiered decentralized republic (Connor, 1984; McGarry, 2018). In the 1920s, several Soviet Socialist Republics (SSRs) were acknowledged that possessed considerable autonomy on paper. Each had the right of self-determination, including the right of secession. Within most SSRs, there were also autonomous republics and oblasts, which usually recognized different nationalities to that of the titular people of SSRs.
The decision to create autonomous institutions was taken primarily to placate the major nationalities, particularly as several of them had sought to secede. By championing rights of self-determination, Lenin also wanted to attract support in the colonial world. But while the Soviet experiment entailed cultural privileges for the titular peoples of the ‘autonomous’ jurisdictions, particularly over language, museums, ethnography, and parts of schooling, these did not include political autonomy. The Soviet Union was governed tightly from Moscow through the Russian-dominated Communist Party. Rights of self-determination and secession were a myth in practice until Mikhail Gorbachev’s reforms inadvertently disabled the Communist Party in the late 1980 (Kotkin, 2008).14 Frequently, the party’s general secretary in an SSR came from a titular nationality, but his more powerful deputy was a Russian. Instead of exercising autonomy, Estonians, Kazakhs, Latvians, and Lithuanians were ruled and guarded largely by Russians, albeit Russian Communists. The dissolution of the Soviet Union did not result, as many have argued, from the failure of a genuinely plurinational federal state, as it had never been one (McGarry and O'Leary, 2009).
(ii) South Africa’s apartheid regime came to power in 1948, with the election victory of the Afrikaner-supported National Party. It reacted to the changing international normative environment by shifting from domination by race to the ‘accommodation’ of nations and ethnic groups. By the late 1950s, it established 10 self-governing homelands for different African ethnic communities. These were purportedly to be prepared for independence, a status that four of them were given between 1976 and 1981. None were recognized by other countries.15 The homelands, derided by their alleged beneficiaries as ‘Bantustans’, were palpably false exercises in self-determination. Most Africans did not seek ethnic autonomy, but democracy within an integrated (difference blind) South Africa. There was also support for African government in South Africa, associated with the Pan-Africanist Congress. Support for ethnic self-government among Africans was limited to rural Zulus. The homelands that were supposed to accommodate 68% of S. Africa’s population were restricted to 13% of its territory, much of it barren. Their boundaries did not come close to reflecting those that were sought by the Africans who were interested in autonomy. All homelands bar one—QwaQwa—lacked contiguous boundaries. Several comprised many non-contiguous pieces, separated to accommodate white farmers. Bophuthatswana had 19 segments and KwaZulu had 70 (Clark and Worger, 2016, p.75).
Rather than accommodating the large African majority, the Bantustans policy was designed to facilitate white domination by fragmenting Africans into numerous ethnic segments. African tribal elites were to be co-opted to assist in indirect rule. Over the longer-run, territorial down-sizing was foreseen, i.e., the transference of Africans’ citizenship from South Africa to the allegedly independent homelands, whether the Africans in question were born there or lived there. Africans remaining within ‘white South Africa’ were to acquire the status of guest workers, like Turks in Germany, before the latter became entitled to German citizenship. Approximately 3.5 million people were forcibly relocated to the homelands between 1960 and 1980, an exercise in demographic engineering (Abel, 2019). The creation of the Bantustans was an exercise in legitimation, aimed at convincing outsiders and white South Africans that the apartheid regime respected post-1945 norms regarding self-determination. The sleight of hand was an abject failure as it turned out.
(iii) The third example is the Palestinian “interim self-governing arrangements” in the West Bank and Gaza that emerged from the Oslo peace process of the early 1990s. A Palestinian Authority (PA) was established in parts of the West Bank and in Gaza from 1995. After 2006 Fatah and Hamas factions administered the former and latter, respectively, with active Israeli connivance. The exercise did not constitute real autonomy. The West Bank remained subject to Israeli military law, at least as their Palestinian inhabitants were concerned. The authority of the Palestinian Authority did not apply to 60% of the West Bank (Area C), a region that, along with Gaza, constitutes the minimum territory that moderate Palestinians are prepared to be self-determining in. Area C includes the West Bank’s most fertile land and has been the site of constant Jewish settlement since the Oslo Agreement, in contravention of international law. While Area C is territorially contiguous, the PA-administered areas (Areas A and B) are scattered and non-contiguous, resembling the South African Bantustans, or Swiss cheese. Israel reserves the right to intervene militarily in Areas A and B, sometimes cooperating with Fatah-led Palestinian security forces who perform the role of the co-opted.
Although Israel withdrew from the Gaza Strip in 2005, the latter has not enjoyed any form of authentic autonomy. It has been constantly blockaded by Israel by sea, air, and land and forced to rely on Israel for “food, water, electricity, trade, mail delivery, access to fishing, medical care and contact with the outside world” (Lustick, 2023). The elites who initially comprised the PA in 1995 were significantly representative of their community, but that is no longer true of their successors. They have been discredited by the “interim” autonomy arrangements, cooperation with Israeli security forces, internal corruption, and the failure to hold elections since 2005. At the time of writing, Gaza has been re-occupied and physically destroyed by Israel following the heinous attack by Hamas on Israel on 7 October 2023.
Domination in accommodationist guises II: sham power-sharing
Domination regimes have also engaged in building sham consociations, usually in the face of international and opposition pressure to end domination. Here are three examples.
i. In March 1978, the whites-only Rhodesian regime established a transitional Executive Council that comprised its leader, Ian Smith, and three Africans (Bishop Abel Muzorewa, Chief Jeremiah Chirau, and the Reverend Ndabaningi Sithole). After a new constitution was written by the Rhodesian Front (the whites’ party), the Executive Council gave way in April 1979 to what was also marketed as a power-sharing coalition, with Muzorewa as prime minister (Weitzer, 1990).
ii. In 1983, South Africa’s apartheid regime changed its constitution, adding legislative chambers for ‘Coloureds’ and ‘Indians’ to the whites-only legislature. The changes were described by the regime as involving power-sharing and by an integrationist critic of consociationalism as being inspired by consociational thinking (Taylor, 2008, p.97).
iii. From 1988 to 1993, Burundi’s Tutsi dictator, Pierre Buyoya, appeared to close the curtains on 23 years of Tutsi domination, which the regime had marketed as ‘integrationist’ (see below), by establishing a power-sharing executive with equal numbers of Tutsi and Hutu (Lemarchand, 1994).
All these experiments lacked two foundational requisites of genuine power-sharing: authentic community representatives and authentic sharing of power. In the Rhodesian case, the three black representatives were African moderates congenial to the white regime. The two main African opposition parties, ZANU and ZAPU, which, between them, won 87% of the African vote in their first democratically competitive elections of country in 1980, played no role in the negotiations that preceded the establishment of the transitional Executive Council in 1978, in the Executive Council itself, or in the new governing arrangements it gave way to in April 1979 (Weitzer, 1990, p.105). ZANU alone, led by Robert Mugabe, won 63% of the African vote in 1980 (Weitzer, 1990, p. 110). The African participants in the power-sharing arrangements were co-optees who lacked real power. During Rhodesia’s transitional Executive Council (1978–9), its three African members were excluded from the powerful War Council, which conducted the campaign against ZANU and ZAPU rebels, and from “any role in military decision-making.” The Council was described privately by its white member, Ian Smith, as a “façade” (Weitzer, 1990, p.105). Under the new 1979 constitution, the tiny white minority, less than 5% of the population, was to control the military, civil service, and judiciary for a decade. Unsurprisingly, Weitzer concluded, the new ‘Muzorewa-led’ government marked “no radical departure from the past.” Coercion on the part of the regime (i.e., discriminatory coercion against the African opposition) increased rather than diminished (Weitzer, 1990, p. 106).
Apartheid South Africa’s experiment with ‘consociationalism’ gave it a particularly bad name (Lijphart, 1985). The experiment entirely excluded African (black) representatives, even though black people were 70% of the state’s population. The Indian and Colored elites were unrepresentative: the overwhelming majority of Indians and Coloreds did not bother to vote in the referendum that ratified the new constitution (Worden, 2007, p.141).
One reason for the lack of enthusiasm among Indians and Coloreds was that these identity groups were inventions of the regime, rather than a bottom-up mobilization. The new institutions also lacked any appreciable power. The new tricameral legislature created in 1983 existed alongside a strong executive president—the State President—who was (indirectly) elected by simple majority vote through an Electoral College in which the white chamber had a significant majority, so much for executive power-sharing. Disagreements between the three chambers on legislation were to be decided through a conciliation process controlled by the State President and the white chamber (Welsh, 2009, pp. 219–20). The so-called concessions to Indians and Coloreds were aimed at dividing them from Africans.
In Burundi, the Hutu members of the ‘power-sharing’ executive were not Hutu leaders—which would imply the existence of followers. Rather, they were appointees of the Tutsi dictator, Buyoya, who could also fire them. As in Rhodesia and South Africa, the real opposition—in Burundi’s case, Frodebu—was banned. Its leader, Melchior Ndadaye—who in 1993 was to emulate his Zimbabwean counterpart, Robert Mugabe, by easily winning his country’s first democratic elections—described his Hutu predecessors in office as “marionettes serving their own interests and those of the (Tutsi) authorities” (Lemarchand, 1994, p.166). He was not mistaken.
The key portfolios in Buyoya’s cabinet—Interior, Foreign Affairs, and Public Works—had Tutsi ministers. The Hutu ministers were faced with an overwhelmingly Tutsi bureaucracy. Only the Ministry of Youth, Sports, and Culture had an administrative cadre that was (slightly) more than 50% Hutu, while the three key ministries referred to had cadres that were 92% Tutsi (Lemarchand, 1994, p.186). Even more decisively, the army, the ultimate arbiter of politics in Burundi, was “entirely under the command of Tutsi officers” and had “only a fraction of Hutu among the troops” (Lemarchand, 1994, p. 168). When Ndadaye was elected President in competitive elections in 1993, army officers assassinated him. Unsurprisingly, the army was described by Lemarchand as the “linchpin of Tutsi hegemony” and “custodian of Tutsi ethnocracy” (Lemarchand, 1994, p.158).16
Domination in integrationist guises
What may be more surprising, at least to integrationists, is that their own narratives—extolling equal citizenship, difference-blindness, and republican nation-building—have also been employed to mask domination by some ethnic communities over others.
After the Fourteenth Amendment (1868) banned racial discrimination in the United States, southern white regimes moved away from overt discrimination, i.e., from explicit and formal subordination of, and discrimination against, black people. Instead, they passed laws that were facially neutral, i.e., non-discriminatory on their face, but intended to discriminate in practice. Rather than legally remove the franchise, southern white regimes deprived blacks of the vote through poll taxes and literacy tests because black people were far more likely to be poor or illiterate than white people. Facially neutral laws were supplemented by (difficult to prove) discriminatory practices. White officials fraudulently counted votes or exercised discretion to allow illiterate whites to vote. White paramilitaries—the nightriders of the Ku Klux Klan—intimidated any remaining black voters from exercising their franchise.
The unionist-dominated regime in Northern Ireland (1921–72) used similar liberal arguments to deny they discriminated against Catholics, pointing to laws that banned religious discrimination. The laws did not prevent discrimination against Irish nationalists nor did they prevent the regime from discriminating in practice against Catholics in the funding of education; the allocation of public employment, housing, and other infrastructure; appointments to senior positions in the bureaucracy or judiciary; or in the way that ‘justice’ was meted out by its police force (O’Leary and McGarry, 2016, pp. 107–47).
The post-1948 regime in Israel has also traditionally argued that it respects the liberal individual rights of its Palestinian citizens, depicting its courts as impartial arbiters of these rights, while discriminating de facto, including by using a variety of facially neutral practices, e.g., tying welfare payments to army service while conscripting only Jews and co-opting Druze. Israel has also discriminated by handing over the allocation of public services—normally a matter for the state—to private (Jewish) agencies.17
These three cases, particularly the last two, avoided a broader republican integrationism in which a single community coterminous with the state was celebrated. Their narrow liberal-based integrationism coexisted with fairly explicit ethnonational, i.e., sectional, politics on the part of the regimes, along with justifications for privileging the dominant national community based on democratic majoritarian norms.
More recently, some majority elites interested in establishing (or re-establishing) domination, or at least in rolling back or blocking concessions to formerly dominated minorities, have moved toward a more republican form of integration, if somewhat half-heartedly. They emphasize the divisive, “disintegrating” effects of accommodating differences, in conjunction with sectional “dog-whistle” narratives that appeal to their supporters and disown the minorities in question.
In the contemporary United States, Republicans have sought to roll back the advances of black southerners (and black people in general) since the Civil Rights Act of 1964 and Voting Rights Act of 1965 by arguing that supporting black people through affirmative action is divisive as well as illegal under the Fourteenth Amendment. These arguments have been used to roll back race-based affirmative action for university admissions (Supreme Court of the United States, 2023). They have also been used to oppose one of the Voting Rights Act’s main mechanisms for increasing black representation in political life, i.e., the creation of “majority-minority districts”—districts that have a majority of voters from the black minority. Such “redistricting” decisions are said by Republicans to be racist. This is in spite of the fact that the alternative is majority–majority districts, which are not seen as racist, even though there is clear evidence that their voters, even when Democratic, clearly prefer white candidates (Neiman, 2020, p. 262). In ruling against majority–minority districts in Shaw v. Reno (1993), Justice Sandra Day O’Connor, a Reagan appointee, argued in classical republican integrationist language that such redistricting involved “racial gerrymandering” and would “balkanize us into competing racial factions” (Cited, Neiman, 2020, p.264).
There are other similar cases. Greek Cypriot hardline nationalists currently refuse to accept a bizonal and bicommunal federation in a reunified Cyprus on the basis that accommodating an autonomous entity governed by Turkish Cypriots would be divisive and “racist.” Apparently, the alternative of an integrated Greek Cypriot-dominated unitary state would not be. The white-dominated Australian Liberal Party, which, despite its title, is conservative and right-wing, recently asked Australians to vote No in a referendum that would have given an institutional “Voice” to Indigenous peoples because this would involve inserting racial distinctions into the Australian constitution. The Liberal leader, Peter Dutton, denounced the proposal because it would “permanently divide us by race” and “re-racialize the Constitution” (Bachelard, 2023). It has become de rigeur for such integrationists to cite Martin Luther King’s words that people should be judged by the strength of their character, rather than the color of their skin. One academic supporter of King recently lamented that his “dream” risked being turned into his “nightmare” (Lawrence, 2024).
The argument that chauvinistic majorities use integrationist narratives may not be particularly revelatory, though it should embarrass integrationists. Accommodationist scholars have argued for some time that integration favors majorities that have greater numbers than minorities and who are favored by rules that weigh only the rights of equal individuals (McGarry et al., 2008).
What may be more surprising is that republican integrationist narratives have also been used by minorities, including very small minorities, to legitimize their domination of large majorities, e.g., in Tutsi-dominated Burundi, before it professed support for accommodation (1965–1988); Ba‘athist and Alawi-dominated Syria (1966-); Ba‘athist and Sunni Arab-dominated Iraq (1968–2003); and Tutsi-dominated Rwanda (1994-).18 All these regimes have enthusiastically used inclusionary republican narratives focusing on a single nation of equal individuals while discriminating against the majorities in question. In each case, the key governing elites have been selected from the dominant minorities, or a section thereof. In each case, the security sectors of regime have been controlled by officers from the dominant minority, with the dominant minority also disproportionally represented in the rank and file. As we shall see, integration in each case has been used to polarize in code, between integrationists (the dominant) and those cast as ethnocentric disintegrationists (the dominated). Evidence is also available, though usually hidden, of discrimination against the majority in the allocation of important public resources, such as education.19
These regimes employ republican integrationist narratives for several reasons. Their elites and supporters may well find republican integration genuinely attractive, at least as a future (long-term) goal. In each case, the dominant minorities in question have been faced with ethnocentric majorities that have either dominated them in the past or threatened to do so in the future. The currently dominant Tutsi community of Rwanda was subjected to a massive genocide at the hands of the Hutu majority in 1994, and, before that, to over 30 years of Hutu domination,20 so it is hardly difficult to understand why its governing elites reject democracy or proportionality any time soon. In such contexts, a republican integrationism in which no group dominates is something that minorities are bound to find attractive over the long term after the majority’s ethnocentrism has been contained and dissipated by the minority’s regime.
In the meantime, republican integrationist narratives are used by the regimes in question to mask and facilitate domination, whether conceived as temporary or not. The narratives serve to maintain the esprit de corps and unity of the supporters of regime (the dominant minority) by convincing them that their cause—the fight against majority ethnocentrism—is legitimate. They may also serve to allay opposition among the large, dominated majorities—making co-option easier—in a way that would be more difficult if the regime took an explicitly ethnocentric (pro-minority) position.21
Importantly, integrationist narratives are used by the regimes in question to win external support and allay opposition. Integrationist nation-building based on equal citizenship is, after all, a globally dominant norm since the end of the Second World War and the onset of decolonization. It is the philosophy of two of the most revered politicians of the post-war era, Martin Luther King Jr. and Nelson Mandela. Maintaining ethnocentric domination, in contrast, risks comparisons with the apartheid regime in South Africa, something no government would want, particularly the two African regimes discussed. The Alawi regime’s embrace of Arab nationalism in Syria was not just popular among its Sunni Arab majority. It also helped Syria to become the military leader of the (largely Sunni-) Arab world against Israel, particularly after Egypt’s Sadat was seen as betraying the Arab cause by going to Jerusalem in 1977. It helped secure significant financial aid from oil-rich but militarily weak Sunni Arab Gulf states, particularly in the latter half of the 1970s and 1980s. The aid was used to strengthen the revenue-challenged Syrian regime’s security sector, ironically, against its Sunni Arab majority.22 It was also used to buy (co-opt) support for the regime from some of this majority.
Similarly, the post-1994 Rwandan regime has won significant backing from outsiders because of its integrationist narrative as well as its role in ending the 1994 genocide. Very substantial external financial aid has been forthcoming to Rwanda, particularly from the UK and the USA, and the aid has generally been given without conditions. The country remains formally aid-dependent, defined as deriving at least 10% of its gross national income (GNI) from overseas development assistance (OECD, 2021). The integration narrative has contributed to the diplomatic success of the regime. It enjoys excellent relations with the West, particularly the ‘anglophony’. Rather than being criticized by Western politicians, Rwanda has been praised as a “glowing story of successful postwar reconstruction” and even for its steps toward democratization (Longman, 2011, pp. 25–6; Reyntjens, 2013, pp.46–8). It was admitted into the Commonwealth in 2009, even though it was never a British colony. In 2013, the New York Times referred to president of Rwanda, Paul Kagame, as the “global elite’s favorite strongman” (Gettleman, 2013). The UK government has recently suggested that its asylum-seekers be sent to Rwanda, while Arsenal, one of the UK’s most prestigious soccer clubs, advertises Rwanda as a tourist destination on its jerseys (Uwiringiyimana, 2018).
Integrationist narratives allow domination regimes to hide ethnocentrism because they provide a rationale for refusing to gather statistical data that would be damaging. Thus, the regimes may decide to prevent the release of statistics on the ethnic composition of the regime, including its wider public sector or security forces, on the grounds that emphasizing the ethnic membership of its citizens is regressive and divisive. This is why data on ethnocentric discrimination in these regimes are often difficult to find. Even ethnically discriminatory violence by the security sector of regime can be hidden on integrationist grounds. In 1988, when the Tutsi military ruler of Burundi, Pierre Buyoya, was asked by a journalist for the ethnic make-up of the victims of security sector violence (the correct answer was that the overwhelming majority was Hutu), he replied that such details were irrelevant because ‘we are all Barundi’ (Cited, Lemarchand, 1994, p.9). Buyoya’s deflection suggested that it was the journalists’ preoccupation with ethnicity that was the problem, not the discrimination of the regime.
Perhaps most fundamentally, integrationist narratives have provided all of these minority regimes with ‘legitimate’ reasons for disorganizing and repressing political mobilization among their ethnic oppositions. This is because, according to integrationists, group claims are disunifying and regressive. Thus, the Rwandan regime bans Hutu parties and silences and jails Hutu representatives on the grounds that their politics are ethnocentric and even representative of “genocide ideology.”23 When the Arab Spring led to mass protests in Syria, largely from Sunni Arabs, in March 2011, the regime responded by accusing its opposition of “sectarianism” and “jihadism,” and cracking down on it on these grounds.24 From the perspective of the regime, such appeals did not just rally their its Alawi base, and give Syria’s secular Sunni Arabs grounds for pause, but also played well to outside audiences, including the United States, Russia, and even Iran, all of whom were strongly opposed to Sunni Arab Islamic terrorism.25 Russia and Iran were eventually to support the Assad regime militarily, saving it from Sunni Arab rebels. The United States did not support it but failed to intervene robustly against it (Phillips, 2020).26
Conclusion
This article has sought to reveal how domination regimes may disguise their discriminatory practices behind the normatively attractive guises of accommodation and integration.
It also seeks to facilitate the identification of domination regimes. This is necessary if external governments and international organizations are to hold them to account for their discrimination and incentivize reforms and alternatives to domination that work for all communities concerned.
Accountability and reforms are important not just because domination is morally unacceptable but also because it has the potential to degenerate into something even worse. While some have argued that domination may be considered a preferred alternative to worse strategies, such as massacres, genocides, and expulsions (Lustick, 1979, p. 336; Smooha, 2002, p. 481), such alternatives may be practiced by domination regimes.27 Indeed, domination, which hands control of decision-making and guns to just one part of a deeply divided polity, should be considered a high-risk factor for these alternatives.
Fortunately, there are other alternatives that should be preferred to domination. These include authentic integration, as chosen by South Africa’s dominant Afrikaner community from 1994, and authentic consociation, as chosen by Burundi’s dominant Tutsi community from 2000.
Data availability statement
The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.
Author contributions
JM: Conceptualization, Writing – original draft, Writing – review & editing.
Funding
The author(s) declare that financial support was received for the research, authorship, and/or publication of this article. The research was supported by the Social Science and Humanities Research Council of Canada (435-2018-0384).
Conflict of interest
The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Footnotes
1. ^The language of organizing and disorganizing as a central dynamic of ‘control’ was first used by my colleague, O’Leary (2020, Vol. II). ‘Preferred’ communities are usually those the elites belong to and identify with, but sometimes regime elites may instrumentally use communities they do not belong to or identify with to hold their lands together. For example, the Soviet authorities, led by the Georgian, Josef Stalin, and other communists, used Russians to maintain the Soviet Union (McGarry, 2018).
2. ^They may also be used to gather support or, more likely, blunt opposition from the dominated, or some of them.
3. ^Consociational power-sharing and territorial autonomy may be overlapping. Consociations include executive-level power sharing and either territorial or non-territorial (cultural) autonomy.
4. ^Integration and accommodation are distinguished by McGarry et al. (2008).
5. ^Both Harff and those responsible for the EPR dataset may also have been misled by these regimes’ use of ‘co-option’. Both regimes co-opted members of dominated groups into office as part of their legitimation narratives. For evidence of tokenism by the post-1970 Syrian and post-2000 Rwandan regimes, see, respectively, Bou Nassif (2021, pp. 130-35) and Wikileaks (2008).
6. ^Some academics believe domination is more widespread than I do, or even universal, because they use the term more capaciously. Thus, Agarin (2016, p. 86) thinks “all” European states, including all EU member-states, “are ethnocracies albeit of different form and disguise.” Wimmer (2004, p.36) argues that “ethno-national dominance” is “universal” in the modern world.
7. ^For examples of demographic engineering see McGarry (1998).
8. ^Legitimation narratives are not restricted to domination regimes, of course.
9. ^For a detailed discussion of the stratagems of domination regimes, see McGarry (2025, Chapter 7).
10. ^Brubaker (1996, p. 86) uses the term “dissimilation” to describe what I mean by domination.
11. ^As Moore (2015) argues, accommodating autonomy requires not just self-government for a people, but self-government in the place they identify with as their homeland.
12. ^Indeed, Kashmir’s autonomy was rescinded by the Modi government in 2019.
13. ^See footnote 1.
14. ^The Soviet Union remained a tightly centralized dictatorship from beginning to end, and its republics’ right of self-determination was a farce, as even the members of the Warsaw Pact discovered in 1956 and 1968.
15. ^The four were Transkei (1976), Bophuthatswana (1977), Venda (1979) and Ciskei (1981).
16. ^Lemarchand (1994, p. 171), the key authority on politics in Burundi, wrote that ‘the “control” aspects of the pre-1993 system clearly outweigh[ed] its “power-sharing “characteristics.
17. ^On the matter of discriminating against assistance to large families, for example, Israel’s first prime minister, David Ben Gurion advised that: “the Government is unable to deal with the problem, and the matter should be transferred to the Jewish Agency or some special Jewish organization. If the Jewish birthrate is not increased, it is doubtful that the Jewish State will survive’. Cited, Lustick (1980, p.109).
18. ^In none of these cases are the minorities more than one fifth of their state’s population.
19. ^For evidence of significant ethnic discrimination in governing institutions, the security sector, and the allocation of public resources in Burundi, see Lemarchand (1994, pp. 108-9 and 137-8). In Rwanda, see Reyntjens (2013). In Syria, see Bou Nassif (2021) and Van Dam (2011).
20. ^Most of the current Tutsi elites are from the Tutsi diaspora that returned to Rwanda as part of an invading army in 1990, or in the wake of that army’s victory over the Hutu-dominated genocidal regime in 1994.
21. ^Legitimation exercises are unlikely to convince anything other than a small minority of dominated communities.
22. ^Funds, training, and advanced weaponry went disproportionately to praetorian units whose job was to defend the regime against its internal opposition rather than to defend the state against its external enemies. See Bou Nassif (2021, p. 154).
23. ^For example, Victoire Ingabire, a Hutu opposition leader, was arrested in 2010 and subsequently jailed on charges of harboring ‘genocide ideology, when she publicly declared that Hutu lives had also been lost and that this should be recognized. She served 8 years of a fifteen-year prison sentence. See Thomson (2018, pp. 182-3).
24. ^The regime also released jihadists from jail, including leading ones, in an apparent attempt to substantiate the threat they posed. See Leenders (2015, p. 254).
25. ^Even conventional ‘Islamic’ regimes, such as Saudi Arabia, were threatened by jihadism, although Saudi Arabia joined the fight against the Assad regime.
26. ^See Phillips (2020).
27. ^Lustick wrote (1979: 336) that control, i.e., what I call domination, “may represent a model for the organization of intergroup relations that is substantially preferable to… extermination or deportation.” For a critique of this suggestion, see McGarry (2025).
References
Abel, M. (2019). Long-run effects of forced resettlement: evidence from apartheid South Africa. J. Econ. Hist. 79, 915–953. doi: 10.1017/S0022050719000512
Agarin, T. (2016). Extending the concept of ethnocracy: exploring the debate in the Baltic context. Cosmopolitan Civil Soc. J. 8, 81–99. doi: 10.5130/ccs.v8i3.5144
Bachelard, M. (2023). Dutton’s claim the voice will ‘re-racialise’ Australia is wrong. Here’s why : The Sydney Morning Herald. Available at: https://www.smh.com.au/politics/federal/dutton-s-claim-the-voice-will-re-racialise-australia-is-wrong-here-s-why-20230525-p5db7i.html
Bou Nassif, H. (2021). Endgames: Military response to protest in Arab autocracies. Cambridge: Cambridge University Press.
Brubaker, R. (1996). Nationalism reframed: Nationhood and the national question in the new Europe. Cambridge: Cambridge University Press.
Connor, W. (1984). The National Question in Marxist-Leninist theory and strategy. Princeton, NJ: Princeton University Press.
Friedländer, S. (1998). Nazi Germany and the Jews: Volume I: The Years of Persecution. London: Orion Publishing Co. 1933–39.
Harff, B. (2003). No lessons learned from the Holocaust? Assessing risks, of genocide and political mass murder since 1955. Am. Polit. Sci. Rev. 97, 57–73. doi: 10.1017/S0003055403000522
Kotkin, S. (2008). Armageddon averted: The Soviet collapse, 1970–2000 New York. Oxford: Oxford University Press.
Lawrence, A. (2024). ‘His dream’s been weaponized into his nightmare’: how Martin Luther King Jr’s words have been co-opted. Manchester, United Kingdom: The Guardian.
Leenders, R. (2015). “Repression is not ‘a stupid thing’: regime responses to the Syrian uprising and insurgency” in The Alawis of Syria: War, faith, and politics in the Levant. eds. M. Kerr and C. Larkin (New York: Oxford University Press), 245–274.
Lemarchand, R. (1994). Burundi: Ethnocide as discourse and practice. Cambridge: Cambridge University Press.
Longman, T. (2011). “Limitations to political reform: the undemocratic nature of transition in Rwanda” in Remaking Rwanda: State building and human rights after mass violence. eds. S. Straus and L. Waldorf (Madison: The University of Wisconsin Press), 25–47.
Lustick, I. (1979). Stability in deeply divided societies: Consociationalism versus control. World Polit. 31, 325–344. doi: 10.2307/2009992
Lustick, I. (1980). Arabs in the Jewish state: Israel's control of a national minority. Austin: University of Texas Press.
Lustick, I. (2023). Vengeance is not a policy: emotionally driven reactions from Washington won’t prevent future violence. Dismantling the Gaza prison could. Foreign Policy.
McGarry, J., and O’Leary, B. (1993). “Introduction: The Macro-Political Regulation of Ethnic Conflict” in The Politics of Ethnic Conflict Regulation eds. J. McGarry and B. O’Leary (London: Routledge), 1–47.
McGarry, J. (1998). 'Demographic engineering': the state-directed movement of ethnic groups as a technique of conflict regulation. Ethn. Racial Stud. 21, 613–638. doi: 10.1080/014198798329793
McGarry, J. (2018). ‘Connor's communist control polities’: why ethno-federalism does not explain the break-up of the Soviet Union, Yugoslavia and Czechoslovakia. Nations Natl. 24, 535–545. doi: 10.1111/nana.12447
McGarry, J. (2025). Domination: taking, keeping, and losing control over other peoples. Oxford: Oxford University Press.
McGarry, J., O’Leary, B., and Simeon, R. (2008). “Integration or accommodation? The enduring debate in conflict regulation” in Constitutional Design for Divided Societies: Integration or accommodation? ed. S. Choudhry (New York: Oxford University Press), 41–90.
McGarry, J., and O'Leary, B. (2009). Must pluri-national federations fail? Ethnopolitics 8, 5–25. doi: 10.1080/17449050902738838
Neiman, D. (2020). Promises to keep: African Americans and the constitutional order, 1976 to the present. New York: Oxford University Press.
O’Leary, B., and McGarry, J. (2016). The Politics of Antagonism Understanding Northern Ireland: Bloomsbury Publishing.
O’Leary, B. (2020). A treatise on Northern Ireland, Volume II: Control. Oxford: Oxford University Press.
OECD (2021). Geographical distribution of financial flows to developing countries 2008–2021: Disbursements, commitments. Paris: OECD.
Phillips, C. (2020). The Battle for Syria: International rivalry in the new Middle East. London: Yale University Press.
Reyntjens, F. (2013). Political governance in post-genocide Rwanda. New York: Cambridge University Press.
Smooha, S. (2002). The model of ethnic democracy: Israel as a Jewish and democratic state. Nations Natnl. 8, 475–503. doi: 10.1111/1469-8219.00062
Supreme Court of the United States. (2023). SFFA v. Harvard and SFFA v. UNC. Available at: https://www.supremecourt.gov/opinions/22pdf/20-1199_hgdj.pdf (Accessed November 19, 2024).
Syrian Observatory for Human Rights. (2020). Nearly 585,000 people have been killed since the beginning of the Syrian revolution. Available at: https://tinyurl.com/yxex83zv (Accessed November 19, 2024).
Taylor, R. (2008). “Ending apartheid: the relevance of consociationalism” in The failure of the Middle East peace process? ed. G. Ben-Porat (London: Palgrave Macmillan), 97–110.
Uwiringiyimana, C. (2018). Rwanda signs tourism sponsorship deal with Arsenal. Reuters. Available at: https://www.google.com/search?q=rwanda-signs-tourism-sponsorship-deal-with-arsenal-idUSKCN1IO1R4&rlz=1C1GCEU_enIN1087IN1087&oq=rwanda-signs-tourism-sponsorship-deal-with-arsenal-idUSKCN1IO1R4&gs_lcrp=EgZjaHJvbWUyBggAEEUYOdIBCDEwODlqMGo3qAIAsAIA&sourceid=chrome&ie=UTF-8 (Accessed November 19, 2024).
Vogt, M., Bormann, N.-C., Ruegger, S., Cederman, L.-E., Hunziker, P., and Girardin, L. (2015). Integrating data on ethnicity, geography, and conflict: the ethnic power relations data set family. J. Confl. Resolut. 59, 1327–1342. doi: 10.1177/0022002715591215
Weitzer, R. (1990). Transforming settler states: Communal conflict and internal security in Northern Ireland and Zimbabwe. Los Angeles: University of California Press.
Wikileaks. (2008). Ethnicity in Rwanda — who governs the country?. Available at: https://wikileaks.org/plusd/cables/08KIGALI525_a.html (Accessed November 19 2024).
Wimmer, A. (2004). “Dominant ethnicity and dominant nationhood” in Rethinking ethnicity: Majority groups and dominant minorities. ed. E. Kaufmann (London: Routledge), 40–58.
Worden, N. (2007). The making of modern South Africa: Conquest, apartheid. New York: Wiley-Blackwell.
Ye’or, B. (2013). The dhimmi: Jews & Christians under Islam. New Jersey: Associated University Presses.
Keywords: ethnic domination, ethnocracy, ethnic democracy, control, nationalizing state
Citation: McGarry J (2024) Disguising ethnic domination as accommodation or integration. Front. Polit. Sci. 6:1419727. doi: 10.3389/fpos.2024.1419727
Edited by:
François Rocher, University of Ottawa, CanadaReviewed by:
Matthijs Bogaards, Central European University, HungaryJorge Cagiao Y Conde, Université de Tours, France
Arjun Tremblay, University of Regina, Canada
Copyright © 2024 McGarry. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: John McGarry, am9obi5tY2dhcnJ5QHF1ZWVuc3UuY2E=