LEAD PAPER in a special issue on Ethnocracy

ETHNOCRACY: Exploring and Extending the Concept

James Anderson

Queens University Belfast


Ethnocracy means ‘government or rule by a particular ethnic group’ or ethnos, sometimes contrasted with democracy, rule by the demos or the people in general. The concept was primarily developed as national ethnocracy for regimes in contemporary national states which claim to be ‘democratic’. It was mainly pioneered by the Israeli geographer Oren Yiftachel to analyse the ethnically-biased policies and asymmetrical power relations of Jews and Palestinians in Israel/Palestine, but it needs to be further developed in other geographical settings and in different types of political context. Yiftachel himself extended it ‘down’ to city level and a specifically urban ethnocracy, and we can further explore how cities and city governance can sometimes moderate national state ethnocracy. Going beyond the national and the urban, and the particularities of the Israeli case, we can also enrich the concept by extending it to three other types of context each of which has its own specific dynamics: firstly, ‘back’ to imperial ethnocracy which often preceded and gave birth to national ethnocracy; secondly, it can be extended ‘forwards’ to the (usually mis-named) ‘post-conflict’ or power-sharing stages of ‘peace processes’, to what we might call shared or ‘post-conflict’ ethnocracy; thirdly, and more speculatively, it can be extended to contemporary religious-political conflicts which are at least partly transnational in character, to what could be called religious andpost-national’ ethnocracy. These five main types of ethnocracy and their inter-relationships can help tie together different features of ethnic and ethno-national conflicts. However questions remain: about, for instance, ethnicity and its sometimes problematic components; about where and how to draw the boundary between ethnocracy and democracy; and about rival concepts such as milder ‘ethnic democracy’ and harsher ‘apartheid’.

Professor Emeritus, The Senator George J Mitchell Institute for Global Peace, Security and Justice, Queen’s University Belfast. Corresponding author: J.Anderson@qub.ac.uk

ISSN: 1837-5391; https://epress.lib.uts.edu.au/journals/index.php/mcs

CCS Journal is published under the auspices of UTSePress, Sydney, Australia

© 2016 James Anderson. This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 Unported (CC BY 4.0) License (https://creativecommons.org/licenses/by/4.0/), allowing third parties to copy and redistribute the material in any medium or format and to remix, transform, and build upon the material for any purpose, even commercially, provided the original work is properly cited and states its license.

Citation: Cosmopolitan Civil Societies Journal 2016, 8(3): 5143. http://dx.doi.org/10.5130/ccs.v8i3.5143


Ethnocracy basically means ‘government or rule by an ethnic group’ or ethnos, and more precisely rule by a particular ethnos in a multi-ethnic situation where there is at least one other significant ethnic group. Ethnicity and group self-awareness can be specified in terms of religion, imputed ‘racial’ features, language, and/or a shared history and culture more broadly defined, components which vary and sometimes in problematic ways. Ethnocracy is often contrasted with democracy or rule by the demos, ‘the people’ in general in a given area or territory – demos originally denoted a place-based community, so democracy was spatial or geographical from the start. It too can be problematical – for instance ‘the people’ in Athens’ fabled democracy did not include women or slaves, and some people are still routinely excluded because of their youth or other reasons; decisions taken by simple majority vote can leave nearly half ‘the people’ dissatisfied; and if this is systematically repeated it becomes ‘majoritarian’ democracy or a negation of democracy (not the demos deciding but the same ‘50+%’ of voters). Nevertheless, despite such problems, democracy is seen as preferable to ethnocracy and indeed to most if not all the other alternative types of rule, though some of these overlap or mingle with democracy and ethnocracy.

These alternatives include autocracy (rule by a single dictator or tyrant as in Ancient Greece), aristocracy (… by big land lords as in feudalism), plutocracy (… by the rich), meritocracy (… by people chosen on merit), bureaucracy (… by officials), or technocracy (… by technocrats or technical experts who ‘know best’). There is indeed the growing complaint that our democracies are being turned into plutocracies (most obviously in the US) and more widely into technocracies, bringing the threat that present-day ethnocracies may face democratic deficits on several fronts simultaneously.

Ethnocracy was generally applicable to traditional empires where a coherent ethnic group conquered and spread its rule over culturally diverse populations, though significantly the concept did not apply to the later Roman Empire which granted citizenship to its inhabitants irrespective of their ethnicity or location within the empire. It still loosely applies to tribe-based politics, but it has been developed from a general imprecise label into an analytical concept, initially for understanding national state regimes. Ali Mazrui in Soldiers and Kinsmen in Uganda: The making of a military ethnocracy (1975, p. 4), defined ethnocracy more precisely as ‘a distributive system which allocates or divides political power primarily on the basis of ethnicity and kinship real or presumed’, and he used the concept to analyse the nature of General Idi Amin’s dictatorship.

However, it was primarily developed for analysing national ethnocratic regimes in contemporary national states which claim to be ‘democratic’. This was pioneered by the Israeli geographer Oren Yiftachel to analyse Israel’s ethnically-biased policies and the asymmetrical power relations between Jews and Palestinians. Particularly in Ethnocracy: Land and identity politics in Israel/Palestine (2006), he analysed the various ways in which the Jewish ethnos has trumped the demos of all the people, when in Israel ‘proper’ (pre-1967 borders) about 20% of the population are Muslim and Christian Arabs or Palestinians, with of course many more in the illegally ‘occupied’ Palestinian territories. Brave in my opinion to call his country an ethnocracy when the national state claims that Israel is ‘the only democracy in the Middle East’ and when this is perhaps the ‘legitimising narrative’ of its national government when criticised for its mistreatment of Palestinians. Moreover, Yiftachel extended the concept ‘down’ from the national level to the sometimes milder urban ethnocracy of cities and city government, along with colleagues like Haim Yacobi (in, for example, his The Jewish-Arab City: Spatio-politics in a mixed community, 2009). Drawing on urban theory and the distinctive ‘politics of urbanism’, we can elaborate more fully on state-city differences and why cities can sometimes moderate national ethnocracy.

We can also question how ethnocracy relates to perhaps alternative or rival concepts and categories, such as the more extreme-sounding apartheid on one side and the milder ethnic democracy on the other. Or are these better subsumed as variants of ethnocracy? When the states claim to be democratic (as most now do), there are questions too about where and how to ‘draw the line’ between democracy and ethnocracy: when, for instance, does democratic decision-making by simple majority voting spill over into majoritarianism and systematic discrimination against an ethnic minority; or how would we recognise if ethnocracy was actually ended, especially where it is mostly informal or de facto rather than de jure and explicit in state laws? Set the ‘democratic’ bar too low and anti-democratic discrimination goes unchallenged; set it too high and perhaps most contemporary democratic states would become defined as ethnocracies which rather dilutes the concept or renders it vacuous.

It can be explored in other national and urban contexts beyond the particularities of Israel-Palestine, in cases such as Belgium, Bosnia or Northern Ireland, for instance; or in this Special Issue, as well as Jerusalem, the cases of Australia, Beirut, Estonia, Fiji, Sri Lanka (as indeed Yiftachel – 2006, pp. 20-33 – shows with short excursions to Sri Lanka, Estonia and Australia). But more importantly, the concept can also be enriched by going beyond the national and the urban to other types of socio-political situation and each of these has its own distinctive dynamics. I suggest three further extensions. Firstly, we can extend it ‘back’ to imperial ethnocracy, which often preceded national ethnocracy, and which, despite having fewer, if any, pretensions to democracy, could paradoxically be less oppressive than successor national ethnocracies which claim democratic status. Secondly, more speculatively and briefly, we can extend it ‘forwards’ to the (usually mis-named) ‘post-conflict’ or power-sharing stages of ‘peace processes’, to what we can call shared or ‘post-conflict’ ethnocracy, and here democracy – or rather preventing or reducing damaging majoritarian democracy – is a central concern. Thirdly, it can be extended (some might think over-extended) to contemporary religious-political conflicts. Religion, with its well-developed institutions and transnational or universalist ethos, is often much more than a mere ethnic or ethno-national marker, and in what we might call religious and ‘post-national’ ethnocracy it can move centre-stage as ‘the’ motivator or goal in its own right. So the politics, like the religion, may aspire to be at least partly transnational in character. So-called ‘Islamic fundamentalism’ is the clearest example; it is also literally post-national to the extent it emerged from the wreckage of ‘Arab nationalism’; and here democracy may be conspicuously absent.

These five variants of ethnocracy – national, urban, imperial, ‘post-conflict’ and ‘post-national/religious’ – and their inter-relations help tie together different aspects of ethno-national conflicts involving contested states and divided cities1. The inter-relations vary: national and urban ethnocracies generally coexist in the same space and dealing with both together can point up the urban dimensions of ethno-national conflict and how cities have shaped the conflict (as well as vice versa). In contrast, relations between imperial and national ethnocracies are historically sequential with the former often though not always giving birth to the latter. ‘Post-conflict’ ethnocracy typically follows a conflict where national ethnocracy was an important cause; and similarly ‘post-national’ ethnocracies may have roots in imperial and national variants. In this historical perspective, and assessing ethnocracy’s inherently conflict-generating, de-stablising and changeable character, we need to see the concept in fluid and flexible terms. Unless constrained by internal forces, or more likely internal forces with crucial external support, ethnocracy is prone to intensification – becoming more extreme, more authoritarian and less democratic. It seems deluded ethnocrats have the rallying cry ‘If you’re in a hole, keep digging’!

National Ethnocracy

In our current era of national states, the dominant form of ethnocracy is national, in sovereign states and moreover states which usually claim to be ‘democratic’: like Israel, and Northern Ireland before the 1998 Belfast/Good Friday ‘power-sharing’ Agreement, and in other situations where claims to be a ‘democracy’ are even more questionable, like Apartheid South Africa. It is often assumed or implicit that ethnocracy applies only to contemporary national states with an element of democracy, but our understandings of the concept are sometimes over-determined by particular cases and this is too narrow. The concept applied to Idi Amin’s dictatorship over the national state of Uganda; and, as argued below, it can usefully be extended to forerunner imperial ethnocracies where democracy was weak or non-existent.

Oren Yiftachel (2006) honed his understanding in the particular context of Israel, which, like most contemporary ethnocracies, does indeed have a nominally-democratic regime, though not in the territory occupied since 1967, while within Israel ‘proper’ it systematically discriminates in favour of the dominant ethnos rather than ruling even-handedly on behalf of the demos of all the citizens. With liberal democracy now a global ‘norm’, ethnocratic regimes usually depend on ‘democracy’ claims for their legitimation, with subordinated groups having some, generally more limited, access to democracy.

Here it is important to stress that where formal democracy occurs in ethnocracies it is not merely ideological camouflage or ‘window-dressing’: on the contrary, in the current era the dominant ethnos typically demands democracy, actively wants it – at least for itself. Thus these national regimes are not simply authoritarian: they typically have parliamentary assemblies and periodic elections, for instance, and perhaps a (sometimes ostensibly?) independent judiciary, and a (supposedly?) free media. These can give the appearance of ‘ordinary democracy’ but they hide a ‘deeper structure’ which is profoundly anti-democratic in that the democracy applies only or mainly to politics within the dominant ethnos, not to the demos of all the people in the state territory. Rule is mainly or solely by members and representatives of that ethnos, it is they who are in charge, making the major decisions: ethnos and demos are conflated but the ‘democracy’ is disproportionately and sometimes exclusively available to the favoured ethnos. And it is this political cleavage between the ethnic groups which defines the character of the state as ethnocratic rather than democratic.

National ethnocracy basically works by (con)fusing the nation-state ideal of the people’s ‘right to self-determination’ in their own territory with ethnic ‘self-determination’ in territory which is ethnically shared. As Yiftachel (2006) points out, typical features include an essentialising, general and all-encompassing ‘ethnicization’ of society – seeing it as a hierarchy of ethnic groups; and using a variety of political, cultural, social, economic, moral and legal frameworks to determine the distribution of power and resources disproportionally in favour of the dominant ethnos. But this ‘ethnicization/hierarchization’ can spread ‘out of control’, extending or spilling over into exacerbating divisions within the dominant ethnicity, a further source of instability (e.g., the ‘ethno-class’ divisions within Israeli Jews between the Ashkenazim of European origin who established Zionism and for long dominated the state, as against the generally lower class Mizrahim of ‘oriental’ origin, though since the 1970s the latter have got a greater share of state-power through more extreme right-wing Likud governments2).

Labour and housing markets are major arenas for ethnocracy. Even where officially ‘open’ without any legal restrictions they tend to be segregated and stratified in terms of ethno-nation and ethno-class: indeed they are key arenas for de facto ethnocracy which is an important element even where ethnocracy is de jure, enshrined in law as in Israel, and they dominated in the Northern Irish case which was very largely de facto (though resting on a legal territorial framework designed to give a permanent Protestant majority – see below). For the working class dependent on state-provided housing, who got housed where depended on religion; and in the labour market workers were subject to segregation in terms of quite intricate job gradations (even within the same workplace) in terms of the pay, the reliability and the conditions (e.g. dirtyness or health hazards) of different jobs. In most cases the Protestant/unionist workers benefited, at least in relative terms (often the only terms recognised), while the workers as a whole were divided and put at a disadvantage. For example, dockers were segregated between those working on cross-channel boats which sailed to/from Britain on a daily basis and hence provided very regular employment, and dockers who worked on deep-sea cargo boats where employment was more spasmodic. Not difficult to guess which were the Protestant and which the Catholic dockers.

‘Ethnic Democracy’ or ‘Apartheid State’?

Officially Israel sees itself as a ‘Western liberal democracy’, though sometimes with an acknowledgement of ‘flaws’ in the treatment of Palestinians (even within Israel ‘proper’, never mind the ‘occupied’ territories of the West Bank, East Jerusalem and Gaza). Trying to come to terms with the harsh realities, other more critical Israeli analysts saw systematic ethnic bias and prefer the designation pioneered in 1990 by Sammy Smooha, ‘ethnic democracy’ (Yiftachel 2006, p. 87). This could fit another official characterisation of Israel as ‘Jewish and democratic’, intended to be read as unproblematic, but Smooha (2002) explicitly recognizes that Israeli democracy is limited, ambiguous and ‘second-rate’. Nevertheless he puts what we will see are questionable arguments: that it is ‘robust and stable’, and that Israel retains an overall democratic framework which guarantees basic civil rights and allows minorities to make incremental gains through democratic and peaceful struggle. He affirms its designation as a democracy, albeit an ‘ethnic’ rather than a ‘liberal’ one but a democracy not an ethnocracy.

Some analysts prefer the term ‘ethnic democracy’ because it is less critical of a regime (in some cases their own country), and it can be argued that it is appropriate where ethnocracy is comparatively mild, the discrimination not very severe. However, it cannot cover other contexts where democracy may be unimportant or non-existent, such as imperial ethnocracy; and in national ethnocracies it is an oxymoron where there is the anti-democratic reality of systematic majoritarian discrimination against an ethnic minority or minorities. In the Israeli case ‘democratic state’ is contradicted by a ‘Jewish state’ which systematically discriminates de jure against resident Palestinians. And that includes all the Palestinians under Israeli military rule in the West Bank as well as in ‘Israel proper’, for Israel-Palestine in reality is now a single state – already Israel’s ‘single-state solution’ – despite its fragmentation into separate, legally distinct territories reminiscent of Apartheid South Africa (a fragmentation which has undermined and now blocks the possibility of an independent Palestinian state and the so-called ‘two-state solution’). Therefore, and because of the political opprobrium attached to it, some analysts prefer to characterise Israel as an ‘apartheid state’ instead of using the more ‘academic’ concept of ethnocracy. However, the two terms are not mutually exclusive; ‘apartheid’ with its connotations of rigidly separate and grossly unequal development (implemented by the minority white population in the case of South Africa) is best treated as a more limited sub-category of ethnocracy to signal a particularly harsh form of it3.

As for ‘ethnic democracy’, the designation ‘Jewish and democratic’ would be akin to demanding a ‘Protestant and democratic Northern Ireland’, another unacceptable contradiction in terms which would cause outrage. Ireland was partitioned in 1920 with the border drawn so as to give Northern Ireland a built-in ‘2 to 1’ Protestant majority/Catholic minority, and in 1969 its ‘majoritarian democracy’ or ‘dictatorship of the majority’ led to the three decades of low-intensity warfare – ‘the Troubles’ – for which the official antidote is now formalised consociational power-sharing, or ‘post-conflict’ ethnocracy (below) (see McGarry and O’Leary 1995). A statement by its first Prime Minister that Northern Ireland had ‘a Protestant parliament for a Protestant people’ did indeed cause outrage at the time and is still quoted in outraged tones over ninety years later. With ethno-national division and systematic exclusion its over-riding feature, ethnocratic Northern Ireland barely qualified as a democracy, ethnic or otherwise, and certainly not a ‘Western liberal’ one (despite the ‘Catholic nationalists’ having full voting rights).

‘Ethnic cleansing’, expansion and conflict-generation

According to Michael Mann in The Dark Side of Democracy: Explaining ethnic cleansing (2005; the main title could perhaps more usefully refer to modern ethnocracy in general rather than only to ethnic cleansing), the removal of native Arabs from ‘Israel’ was most murderous in the late 1940s. It happened partly under nominal British Empire rule before the Zionist movement established its own independent national state: its regime was thus ethnocratic from ‘before the start’, and indeed intentionally so from Zionism’s beginnings at the end of the 19th century (see Anderson 2013, p. 206). But some national ethnocracies (e.g. the Turkey of the Young Turks; newly-independent Sri Lanka in the 1950s) were initially democratic in inspiration but can then slide into authoritarianism and ‘murderous ethnic cleansing’ (and indeed face accusations of genocide, as of Armenians and Sri Lankan Tamils4); though again some ethnocracies are established without significant violence (e.g., in Estonia vis à-vis its sizeable Russian-speaking population after 1990).

But Yiftachel notes that the establishment of Israel did involve ‘a massive ethnic cleansing’ – the 1947-48 Nakba or Palestinian catastrophe – in which over 700,000 Muslim and Christian Palestinians were removed into exile, many terrorised into leaving by selective massacres by the paramilitaries of Irgun led by Menachem Begin (later the founder of Likud), and the Stern Gang (whose leader Yitzhak Shamir also became a Likud Prime Minister). The Palestinians were killed or for the most part driven out according to pre-prepared ‘Plan D’ (or Dalet) and were replaced by Jewish settlers from Europe, and also from Arab countries (see Pappe 2007, pp.39-40, pp.86-126; Anderson 2013, pp.207-8).

Yiftachel also sees settler ethnocracies as premised on ethno-national expansion. This certainly fits the case of Israel and its territorial expansion, initially across Israel ‘proper’ and then in the West Bank, East Jerusalem and Gaza after their 1967 ‘occupations’. But we need to be careful not to over-generalise. Ethnocracy is not necessarily driven by a ‘settler society’, nor does it necessarily include ethnic cleansing, at least not of the murderous sort – post-1990 Estonian ethnocracy is a case in point where the ‘settlers’ are in fact members of the Russian-speaking and arguably relatively oppressed minority. Conversely, Northern Ireland does have a ‘settler society’ in some ways comparable to Israel’s at least at an ideological level (see below, and Anderson 2013) but the unionist regime was not territorially ‘expansionist’ in any real sense, and in general was more intent on defensively retaining what it had5, rather than expanding or deepening its ethnocracy. So ‘ethno-national expansion’, at least in a territorial sense, is not a necessary feature of ethnocracy. On the other hand it does occur in many cases. There is no ‘settler society’ involved in the Sri Lankan conflict in that both the northern Tamil and the dominant Sinhalese communities have co-inhabited the island for at least two millennia. Yet the post-war expansion of Sinhalese settlement into Tamil areas, often on land taken over by the army for supposedly security reasons, has been both an outcome of ethnocracy and an ethnocratic instrument to further weaken the Tamils politically. For example, it drove a ‘wedge’ of Sinhalese settled territory between the Tamils in the Jaffna and the Trincomalee regions (interestingly with advice from the Israelis, past masters in the use of territoriality for political purposes)6.

Yiftachel rightly characterises ethnocratic regimes as inherently conflict-generating: they give and ask for trouble, and hence are inherently unstable. Left to themselves they are prone to intensification, becoming more rather than less ethnocratic, ‘digging’ a deeper ‘hole’ for themselves: conflict generates a reaction and further conflict in a downward and often violent spiral. For example, Israel’s Zionist nationalism and the supporters and also the opponents, of the regime have all become increasingly ‘extreme’ over time. In Sri Lanka the ethnocracy of the majority Sinhalese escalated steeply to all-out war against the northern Tamil population with massive oppression and loss of life, but from relatively mild and almost trivial beginnings some fifty years earlier - largely instigated by an ambitious but thwarted Sinhalese politician, Mr. Bandaranayake who used pro-Sinhalese bias in order to become Prime Minister.

In the downward spiral even the ethnocratic regime itself can lose effective control through what can sometimes seem like self-inflicted political stupidity. In Northern Ireland the Protestant majority Unionist government generally defined all the Catholics as ‘disloyal’ supporters of a ‘United Ireland’, but by the 1960s this traditional objective actually had very little real support even in Catholic nationalist heartlands (a friend from nationalist West Belfast said only 5% of his secondary school class had supported it). Yet despite this, the regime persisted with its blanket assumption that Catholics were ‘disloyal traitors’, and after five decades of ethnocracy and then failed reforms (seen as ‘too little, too late’) that would become a ‘self-fullfilling’ prophesy, proof for many that Northern Ireland was ‘unreformable’, when the three decades of low-intensity warfare – the ‘Troubles’ – erupted in 1969. Political stupidity aside, the ethnocracy had a strong class dimension in keeping the working class divided and Protestants away from socialism, but most sections of society, nationalist and unionist, lost out. Ironically, perhaps none lost out more than the once relatively privileged Protestant workers who often failed to adapt to de-industrialisation and the new (un)employment circumstances.

Countering ethnocracy

Halting or breaking the downward spiral – reducing ethnocracy, reducing instability, increasing democracy – usually depends on the inevitable internal opposition getting external help and support in the form of effective pressures to force reforms. This often fails to materialise, or in sufficient measure. It happened to some extent in the case of Estonia with an easing of ethnocratic policies towards the minority Russians (many of whom did not have Estonian citizenship), under pressure from international human rights organisations and as a condition for Estonia being granted EU membership in 2004 (though pressure seems to have diminished again since membership was gained and the EU’s external leverage was reduced). It happened with Northern Ireland’s 1998 Good Friday Agreement, particularly through support and pressure from the European Union and the USA. But even in this comparatively fortunate case, external help took a long time to have effect and changing ‘internal affairs’ from the outside still has its limitations. It has still not even begun in Israel-Palestine where the main external player, the US, has the means but not the will to insist on meaningful reforms (and another important player, the EU, lacks both means and will). In Sri Lanka it has happened only to a very limited extent: the Tamils were defeated militarily and could exert less pressure themselves; but the victorious regime of Rajapaksa was deemed by the US and India to be too friendly towards China and was replaced by the less-overtly ethnocratic regime of Sirisena; in reality, however, it was not very different and now there is less external pressure to reform ethnocracy because the perceived geopolitical threat from China has diminished, at least for the time being7.

On the other hand, the tensions or contradictions of ethnocratic regimes and their inherent instability point towards how ethnocracy might be weakened or transcended by internal forces. Rather than simply (unthinkingly?) accepting the dominant terms of reference of ethnic or ethno-national protagonists, the relative importance of their ‘foregrounded’ ethnic divisions can be questioned. ‘Backgrounded’ or neglected social divisions which cross-cut and under-cut them can instead be encouraged by, for example, mobilizing around what some would argue are more important class, gender, and ecological issues. These are typically down-played and often actively obscured in ethnocracies, but this can lead to popular discontent and public protests against ethnic politics and its politicians, as happened in 2014 with urban riots in Croat and Bosniak parts of Bosnia largely around economic and unemployment issues. In Lebanon ethnocracy is now being confronted by a new left alliance, ‘Beirut Madinati’, which is organising not around ethnic issues but people’s living conditions, health and safety, affordable housing, environmental sustainability and social justice.

Looked at comparatively – at different national cases and at the same case getting better or worse over time – the methods and degrees of ethnocracy can vary widely on spectrums from ‘authoritarian’ to ‘democratic’, and de jure to de facto. Ethnocracy is relatively easy to recognise and assess where de jure, explicit in state laws, as in Israel or Apartheid South Africa; sometimes less easy to recognise or ‘prove’ where it operates de facto or informally in civil society. Thus in Northern Ireland ethnocracy was (or is?) almost entirely de facto (even in its pre-1960s heyday there were no laws explicitly discriminating in favour of Protestants, they were favoured more subtly)8; and tracing its decline is complicated and contested. Direct intervention from London from 1969, and especially its imposition of ‘Direct Rule’ closing down the local assembly in Belfast in 1972, set the scene for gradually – some would say very gradually – dismantling the structures of ethnocracy (e.g. removing local authority control of housing allocation; outlawing forms of ethno-national discrimination in the labour market). But at the same time the British Ministry of Defence set up the Ulster Defence Regiment which armed one section of the community – the unionists – for its battle against the IRA and in effect against the nationalist population more widely. One part of the state can be anti-ethnocratic while another is ethnocratic. However, by 1998 and the power-sharing Agreement, ethnocracy had undoubtedly been substantially reduced, though discrimination has been so embedded in local society over centuries that detailed research would be needed to identify when or if it has ended (perhaps it is now simply ‘shared’ – see below).

Similarly, in de jure terms we can easily conclude that Israeli ethnocracy has significantly more democratic content than Apartheid South Africa – hardly difficult, not least because the ruling ‘whites’ were a clear minority of the total population (and the designation ‘white and democratic’ would have been a totally unconvincing oxymoron). However when we turn to de facto geopolitics, or the social realities of labour markets, things are not quite so simple. Firstly, Israel as a geopolitical ally and supporter was indirectly implicated in Apartheid South Africa (as in Sri Lankan ethnocracy). Secondly, whereas the economy of Apartheid South Africa was totally dependent on the labour of the black majority and there was never any question of ‘ethnically cleansing’ or excluding them (except to the nearby phoney, separate ‘Homelands’ within South Africa), Israel in contrast consciously replaced whatever small dependence it once had on the cheap labour of Palestinians, employing new settlers instead, and latterly using temporary or unenfranchised migrant workers (increasingly a potentially important new ‘third party’ inserted into ethnocracies). From an Israeli economic viewpoint the Palestinians are dispensable (see Piterberg 2008). This trumping of economics by demographic considerations is unusual because oppressed minorities are typically seen as ripe for exploitation, and it goes some way to explaining the severity of Israeli ethnocracy.

Urban Ethnocracy

The extension of the concept to ethnically-divided cities and city government generally applies within the context of national state ethnocracy (see Bollens 2007)9. But the dynamics are different and outcomes crucially depend on the relations between state and city. On the one hand, urban centres can have considerable scope or relative autonomy to modify or moderate the national ethnocracy and associated conflict. The demographic ‘balance of forces’ in a particular city may be radically different (e.g. either more or less favourable to the minority) than that at state level; but in general the capacity to moderate stems from the specifically ‘urban’ character of cities. And so in Northern Ireland various power-sharing arrangements were pioneered in city and local governments long before they ventured onto the ‘national’ stage of the regional government assembly. On the other hand, city autonomy can be severely curtailed or nullified, especially when at the mercy of states which believe they are fighting for their own survival. Cities have always to be seen in their national and indeed wider international contexts, and they can either reduce the severity of ethnocracy and conflict or exacerbate it, depending largely on national circumstances and wider interference or support.

It is in the character of cities, and sometimes especially capital cities, to attract conflict: they are prime symbolic and material targets for violent attack, prizes to be won in the conflict of ethnically-divided societies. Few prizes come bigger than Jerusalem, central to three major world religions and the putative and disputed capital for Israeli Jews and for the Palestinians. But all cities are also relatively easy to disrupt (e.g. can be brought to a standstill by strategically placed car bombs, real or hoax). Antagonistic urban groups living at relatively high densities and in close proximity can make for intense and on-going conflict. Yet, other urban features point in the opposite direction – unless there is full scale conventional warfare and the city is subject to ‘urbicide’ (Coward 2004). This ‘killing’ or total destruction of a city is however often an exaggeration or in reality quite exceptional for ‘divided cities’ in ethnocratic conflicts (notwithstanding the all-too-frequent images of urban devastation from Gaza, Iraq or Syria). In most cases, the contending forces are, or have resident supporters, in the same shared city. For it to function reasonably ‘normally’ as a city there has to be at least some minimal social interaction and co-operation across and between the city’s different parts, activities and communities; and most of the city’s residents most of the time have a strong vested interest in that continuing – longing for ‘normality’, i.e. non-conflict, is often a popular refrain. This ‘urban functional imperative’ to co-operate can often undercut or at least modify ethnic divisions and conflict, and the message from urban ethnocracy is often relatively encouraging compared to its national counterpart. It is not coincidental that Northern Ireland power-sharing started at city level; nor that the anti-ethnocratic Lebanese Beirut Madinati alliance is city-based.

While generally applying to cities within national ethnocraticies, urban ethnocracy is a distinctly different type, not simply ‘national ethnocracy writ small’, just as cities are a qualitatively different type of entity from states. Urban ethnocratic regimes generally encourage ethnic segregation, but they have less control than states over issues such as economic development, migration, and private sector profit-driven housing provision (an ironic example is middle class Palestinians paying above average prices to move into new Jewish ‘colonial neighbourhoods’ which were intended to further Jerusalem’s Judaisation – see Yacobi and Pullan 2014). The relative ‘openness’ of urban housing and employment to economic market criteria can in some circumstances weaken ethnic priorities or allow a comparatively milder ethnocracy – as arguably occurred in what by today’s standards seems a very harmonious Jerusalem under Mayor Teddy Kollek, from 1965 until he was replaced by a right-wing Likud mayor, Ehud Olmert, in 1993 (see Dumper 1997).

As Yiftachel (2006, pp.189-190) puts it, the ethnocratic city

‘… offers a more open and porous spatial ethnic arena, where some of the ethnocratic boundaries and hierarchies may be undermined .... on the one hand, it is an officially open and accessible space, on the other, it is segregated, controlled, and hierarchical....it is increasingly part of a globalizing culture and economy characterised by international migration, development bursts, and discourses of human rights, civil society, and democracy…. Yet…it is a product of the nationalist, expansionist logic of purified ethnic space.... urban informality becomes a common planning strategy in such settings, allowing both orderly and modernist planning to proceed while containing the tensions and preserving the control over the unplannable elements in these cities.’

Here he is referring to ethnocratic urbanisation which has given rise to informal or illegal urban developments (e.g., unauthorised squatting, undocumented workers, unlicensed trade). However, these features – ‘grey areas’ or ‘gray space’ – are in fact more general to cities in the ‘developing world’ which are not ethnocratic, and they are not necessarily replicated in Western ones which are. But they do point to the importance of urban planning and land-use controls as tools of ethnocratic city management (see Bollens 2007). The features of ‘informality’ are prominent in the particular case of Israeli urban planning where a technocratic veneer masks a minimization of Palestinian power through the gerrymandering of municipal borders, and a strategy where illegalities are condoned or even facilitated as a way of containing ‘the ungovernable’. The actual result is unserviced and deprived communities where planning agencies and tools ostensibly supposed to promote social amenities and welfare are in fact used to contain and oppress the marginalised.

The distinctiveness of urban regimes is partly explained by Yiftachel in terms of the logic of capital accumulation as a driver of urban transformation vis-à-vis the logic and the limitations of ethnocratic control. That is a useful distinction but it does not fully account for the distinctiveness of ‘the urban’. After all, the logic of capital applies throughout the whole of society and to the state, not only to cities. Rather, the city differs from the state partly in terms of its formal government being subordinate to national state government, but also – as an element of that – what we can call its ‘partial de-politicization’. There is in capitalism a partial, institutionalised separation of ‘politics’ from ‘economics’, and one expression of it is a ‘division of labour’ between state and city. While the state monopolises the so-called ‘high politics’ of sovereignty, security and the exercise of legitimate violence, the modern city’s usual focus is more on economic, social and cultural issues and their associated ‘low politics’.

Seeing like a city not like a state

We can elaborate further on this by drawing on urban theory and the distinctive ‘politics of urbanism’ to discuss state-city differences more fully and how cities can sometimes moderate national ethnocracy. At the same time we have to recognise that ethnocracy itself can reduce the distinctiveness of urban politics and reduce their moderating effect.

However, first it helps to see urban politics in so-called ‘normal’ (i.e. non-ethnocratic) circumstances as fundamentally different from state politics, as Warren Magnusson argues cogently in his Politics of Urbanism: Seeing like a city, 2013. The contrast he draws is between a state politics of basically singular sovereignty in a more-or-less bounded national territory, compared to the essentially unbounded character of the city and its city-to-city networks (which often cross the territorial borders of the state). Specifically urban politics relate not to a single source of authority (like ‘the’ state) but to multiple sources, mainly economic and cultural as well as the conventionally political. And thereby the city opens up possibilities for comparatively greater flexibility and moderation in the exercise of ethnocracy.

To summarise his main points, Magnusson makes a persuasive case that:

‘urbanism is ultimately uncontainable, and that means that its politics always exceed the regulatory efforts of the highest authorities …. urbanism and statism are always at odds, since statism… is inspired by the ideal that everything can be contained or controlled by a higher authority …… To see like a city is to focus on what happens between people, what enables urban life, what questions arise within it, what solutions are developed, what conduct develops and to what effects. To see the political in these terms is to refer back to these practices rather than to the ones by which people are ostensibly “ruled”. The question of rule is secondary, because it only arises as a threat or a possible solution – in a context in which people are always engaged in making their own lives work, under conditions in which they are thrown together by their own aspirations or by the immediate necessities of life… If the state emerges in response to the problem of sovereignty, the city has different origins and a different political order is immanent in its own rationalities. If the state requires a monopoly of legitimate authority to persist as a state, the city as city does not. The city depends on the intricate pattern of government and self-government…and hence on multiple authorities in different but not necessarily inconsistent registers. The effect of the city is to hold sovereignty in suspense, defer its application, and perhaps render it redundant… One might argue that what typifies the city is not the imposition of an overarching authority, but the multiplication of challenges to existing authorities of all sorts… the city… is not organized on the sovereignty principle, but instead on the principle of self-organization which in turn implies a multiplicity of authorities operating under conditions of rivalry and inter-dependence.’ (Magnusson 2013, p.24, p.34, p.109, p.117, p.118)

This very interesting perspective10 points to the punchline that the city should be a major, even the major, site for attempting to transform ethnocracy and ethno-national conflict, as important as the state and in different, sometimes complementary ways.

Unfortunately however things are not quite that simple. The punchline has to be qualified because ethnocracy itself reduces what makes urban politics distinctive from state politics and hence reduces the city’s moderating effect. Ethnocracy tends to negate what in non-ethnocratic circumstances is the modern city’s ‘partial de-politicisation’ with respect to sovereignty – indeed this negation can be seen as one of the defining characteristics of the ethno-nationally divided city. Holding ‘sovereignty in suspense’, deferring or rendering it ‘redundant’, hardly applies to the highly-politicised city caught up in conflict about statehood. Far from sovereignty issues being removed from city to state level, they are often central to ethno-nationally divided cities, and right down to conflicts at neighbourhood level (for example, fights for local territory in Belfast are intense because they act as proxies for the struggle over national territory). This type of (‘high’) politicisation is indeed one important way of distinguishing the ethnocratic city from so-called ‘normal’ cities or modern ones in general.

Nevertheless, cities are generally very important as sites for potentially moderating or reducing ethnocracy. The associated conflicts, like contemporary populations, are now largely urbanised. The urban features identified by Magnusson, particularly the city’s ‘multiple authorities’ and the ‘multiplication of challenges to existing authorities’, continue to exist under ethnocracy.

Here Jennifer Robinson’s (2006) concept of ‘ordinary cities’ is useful, and in two respects. Firstly, she argues against the limiting, debilitating compartmentalisation of urban studies, divided up into different world regions or cities in ‘developed’ as distinct from ‘developing’ countries: the former tend to be privileged as the focus or source of urban theorising while the latter are relegated to residual categories. Instead all cities should be understood as ‘ordinary’, all have innovative potential, especially as all now interact with more global forces. This has clear relevance for urban ethnocracy for it can and does occur in cities across the globe – across the conventional ‘compartments’.

Secondly, compartmentalising has the strong implication that cities have to be understood and explained primarily in terms of their particular ‘compartment’ and factors particular to it. Instead they should all be treated as ‘ordinary’ in the first instance, all subject to the same or similar general processes, and at least partly understandable in terms of the same general insights. Thus ‘ethnocratic cities’ can be seen as ‘ordinary’, subject to many of the same general or ‘ordinary’ urban processes (e.g. (de-)industrialistion, (sub)urbanisation, gentrification, marketization and so forth) as many other, non-ethnocratic cities. And this especially needs to be remembered in cities riven by ethnic and national conflicts where there is a tendency to try and ‘explain everything’ in terms of the conflict, as caused by it or by the ‘other side’ (it sometimes seems as if ‘nothing normal’ ever happens in these conflict-divided places). Sometimes the real explanation is more mundane, to be found in ‘ordinary’ urban processes. At the same time, however, it is undoubtedly true that ethnocratic cities are shaped by (and shape) ethnic conflict, and to that extent they have to be seen as ‘extra-ordinary’. Their reality is both more complex and more interesting: they are simultaneously ‘ordinary’ and ‘extra-ordinary’ and they have to be understood in terms of how the two facets interact.

Imperial Ethnocracy

If urban and national ethnocracies clearly coexist in the same time-space, relations between imperial and national variants are more complex, combining coexistence and historical sequence. While national ethnocracies mostly stem from imperial ethnocracies, and traditional empires were eventually superceded by national states, imperialism and nationalism in fact overlapped and interacted with each other over two centuries, from the French Revolution to the post-World War II era, and each took on characteristics of the other (see Anderson and O’Dowd 2007). In studying ethno-nationally divided cities in contemporary ethnocracies11, we found that nearly all of them originated at the insecure and often contested frontiers of traditional imperial ethnocracies (e.g. Trieste in the Austro-Hungarian Empire, Mostar in the Ottoman, Danzig in the Prussian, Belfast and Montreal in the British, while the Muslim Uyghur city of Kashgar in China’s Xinjiang province – literally ‘New Frontier’ – is a contemporary ‘throw-back’). In frontier regions (e.g. in the Balkans, the Middle East) imperial control was often over-extended, or brutal methods substituted for efficiency, and /or there were interventions by rival imperialisms (e.g. Spain and France in Ireland; Austria-Hungary in the Ottoman Balkans; and the 19th century British almost everywhere). The ruling imperial powers typically politicised ethnic groupings, sometimes governing through them (e.g. in the Ottoman millet system where religious leaders became the political leaders of their communities); and imperial ‘divide and rule’ strategies based on a hierarchization of competing groups (as in Ireland) often spawned and exacerbated later ethno-national conflict.

Belgium and its capital Brussels are a rare exception to this historical sequence. There were no region-based differences of religion (most Belgians were at least nominally Catholic), but instead a general opposition of ‘progressive’ liberals to ‘religious’ conservatism. Here language-based regional differentiations and eventually national ethnocracies only emerged in the late 19th century over half a century after state freedom from the Austrian Habsburg Empire had been engineered, largely by external powers. Belgium’s ethno-national linguistic differentiations had no real political precedents in the Empire (who cared what the non-enfranchised majority of the people spoke?), even though the Empire had favoured French as its language of administration, followed by the upper classes irrespective of which region they lived in (see Hepburn 2004, pp. 136-140). Despite romantic theories of nationalism making language central, it mostly became important as a practical political issue and ‘marker’ only with the introduction of compulsory primary education in the later 19th century, for that immediately raised the issue of which language or dialect should be used in schools and by extension in society generally.

Another, perhaps only partial, exception to national ethnocracy having imperial roots is Sri Lanka where the British tended to favour generally better educated northern Tamils over the majority Sinhalese for state administration jobs. But it was at most a weak ethnocracy; the Sinhalese and the northern Tamils were broadly united in opposing British rule; and national Sinhalese ethnocracy only really started becoming a serious factor in the 1950s about a decade after national independence.

In some cases (e.g. Palestine, Ireland, and Australia, though not Sri Lanka or Lebanon) the imperial and subsequent national ethnocracies are further complicated by substantial ‘settler colonialism’. This produces its own dynamics in three-way relations between ‘settlers’ and ‘natives’ (e.g. between Zionist Jews and Muslim or Christian Palestinians), and thirdly the ruling ‘metropolitan’ or imperial power. It would be misleading to simply project back the contemporary two-way relationship (e.g. between Zionists and Palestinians) or play down (if not ignore) imperial ethnocracy. It, after all, was generally the key agency in creating both the settler colonialism and the later national ethnocracy. Imperial ethnocracy had its own overall ethnic biases, contrary to its commonly projected self-image of ‘honest broker’, though this ideology did have some partial basis in reality. The three-way relations involved ‘divide and rule’ strategies where imperial control depended on playing off ‘settlers’ and ‘natives’ against each other, mostly favouring the former though occasionally the latter; and this undoubtedly had a moderating effect, compared to the simpler and often harsher two-way ‘settler-native’ juxtaposition in the national ethnocracy which followed (despite the latter’s much greater lip service to democracy). This applied to a lesser extent in Australia because the indigenous people posed much less threat to the state than did the contending forces in Palestine or Ireland for instance, and there was therefore less need to partially accommodate them. But is it not significant that it was Australia’s own national federation which from 1901 formalized the ‘white Australia policy’ and the forced assimilation of so-called ‘Aborigines’?

In pre-1948 Palestine with its very substantial, active in-migration and a Zionist settler society, and earlier in Ireland with historical migration and now a ‘quasi- settler’ ideology or mentality among many Northern Ireland Protestants, the British imperial power played an absolutely key role in Jewish and Protestant populations achieving separate statehood. Thanks to the British, both achieved it despite being demographic minorities. They were favoured partly because Palestinian nationalism (initially at least) and Irish nationalism (mainly Catholic in composition) were seen as the greater threat to the British status quo. ‘Dictatorships of the majority’ in a democracy are bad enough but courtesy of the empire and non-democratic imperialism, Palestine and Ireland got demographically improbable ethnocracies based on ‘dictatorships of the minority’! (see Anderson 2013).

Ethnically biased ‘balancing acts’ by the British, and also by the French with respect to the Maronite Christians of Mount Lebanon (see Traboulsi 2007), ensured that minorities succeeded in carving out their state territories contrary to the wishes of the great majority of the respective pre-partition populations in Palestine, Ireland and Greater Syria. And they did it perhaps too ‘greedily’ in Northern Ireland and Lebanon, for the Protestants and Maronites would eventually lose their demographic pre-eminence. In Palestine the British made expedient WWI promises both to the Palestinians and to the Zionists, but they were committed to a Jewish ‘national home’ on Palestinian land through the Balfour Declaration of 1917 which was then written into their subsequent League of Nations Mandate. In the words of Jerusalem’s British Governor in the 1920s, they wanted to create ‘for England a “little loyal Jewish Ulster” in a sea of potentially hostile Arabism’ (a self-serving imperialist logic still alive in the USA, though Americans now increasingly have to question Israel’s ‘loyalty’).

Up to 1939, the British had seriously weakened the Palestinians politically and militarily before they were faced with ethnic cleansing in 1947-48, while in contrast half of the opposing Zionist forces had had military training from the British Army. Their ‘Plan D’ ethnic cleansing actually began before independence because the British effectively abdicated and left the Palestinians to their fate (see Pappe 2007, and Piterberg 2008).

Imperial power has to negotiate the three-way relations involving ‘settlers’ and ‘natives’, or it has to ‘divide and rule’ a plurality of indigenous ethnicities, but either way imperial ethnocracy is rarely a simple matter of unambiguously supporting one ethnic group. Instead, while seldom ‘even-handed’, it typically has to alternate in supporting different groups (and in consequence may be distrusted by all of them). National ethnocracies in contrast generally have no such ‘balancing’ constraints: national self-governing ‘democratic majorities’ typically have the simpler, less mediated, two-way relationship against the ‘minority’ – as in independent post-British Israel; and in Northern Ireland where the British government effectively left ‘local matters’ to Unionist rule – or rather ‘mis-rule’ – until things ‘blew up’, literally, in the late 1960s.

Although national ethnocrats are more concerned about projecting democratic appearances, their regimes of bias are often less restrained and more authoritarian than imperial forerunners. There is thus the general paradox of imperial ethnocracy often being ‘milder’ than its national ‘democratic’ successors. However, this paradox is only partial and has to be qualified. Some imperial powers clinging on desperately before departure in the face of local opposition resorted to extreme brutality and very ethnically-divisive tactics (e.g., the French in Indo-China and Algeria; the British in Malaya, Kenya, Cyprus and Aden). This, happening last, is often what is remembered locally, rather than more peaceful imperialism in its prime, and so a nasty lasting legacy was left for successor national regimes, including ethnocratic ones.

‘Post-Conflict’ or ‘Shared’ Ethnocracy

More speculatively, I’m suggesting two further extensions of ‘ethnocracy’: to the usually mis-named ‘post-conflict’ stages of ‘peace processes’, and, even more tentatively, to ‘post-national’ and religious ethnocracy. These variants are speculative or tentative largely because of their overlapping, ambiguous or shifting relations with national ethnocracy, and particular regimes or movements may move into and out of the categories.

‘Post-conflict’ or shared ethnocracy is a continuation of national ethnocracy but it constitutes a distinct category as it now involves two (or, as in Bosnia, three) distinct and conflicting ethnicities with access to state power though on a shared basis. In the new consociational, power-sharing arrangements they continue to compete but they are also under pressure to co-operate. So this variant has its own distinctive dynamics, or in some cases a lack of them: there may be mutual blocking of ‘the other’s’ initiatives, or ‘gridlock’ because of institutional guarantees which were originally designed to counter damaging majoritarian democracy and end violent conflict.

Typically power-sharing has to be imposed or at least strongly encouraged by external powers – dominant ethnocrats do not willingly agree to it. Examples include the 1995 ‘Dayton Agreement’ for the territorial and governmental relations between the Serbs, Croats and Muslim Bosniaks of Bosnia, and the 1998 Good Friday Agreement for Northern Ireland. They have had the great merit of very largely ending lethal conflict, but inevitably with negative aspects. Institutional guarantees have been necessary to get the armed fighters to stop fighting, but that almost inevitably builds dysfunctional rigidities, even ‘gridlock’, into the political system. It can help perpetuate, or indeed accentuate, the very ethno-national divisions which feed the conflict and were a cause of it in the first place, despite being well-intentioned. Old ethnocratic ways of thinking die hard anyway, but, harder if they are actively encouraged and perpetuated by consociationalism. Not surprisingly, its critics have latched onto this truth, some thinking to ‘demolish’ the very idea of consociationalism. But they in turn need to be reminded of the other truth that there once was an armed conflict which had to be stopped: here the framing of the original settlement obviously requires care and foresight but in reality there are no easy answers – undefeated fighters have to be given guarantees to persuade them to stop fighting.

Power-sharing between groups already defined in ethnic terms inevitably puts ethnic identity and ethnic politics centre-stage, and this tends to crowd-out other identities, other politics. It may confirm separate ethno-national electorates with each ethnic group voting for its own ethnic representatives, and little ‘cross voting’; and little scope for dealing properly with issues which straddle the ethnic divisions, such as those of social class, poverty, inequality, the environment, non-ethnic forms of discrimination, and so forth. These get systematically neglected which is what led to rioting in Bosnia and helped motivate Madinati in Beirut. The main democratic or inter-party competition tends to be within more than across or between the main ethno-national communities, and that often results in the more ‘extreme’ representatives (or self-styled ‘defenders’) of the ethnic groups winning their respective intra-ethnic competitions. Sometimes – as in the division of Bosnia into the Republica Srpska for Bosnian Serbs, while the other two ethno-national groups share the territorially separate Croat-Bosniak Federation – the continuing ‘post-conflict’ divisions are reinforced spatially and territorially. But even without territorial reinforcement there tends to be ethnically-divided and hence often dysfunctional government.

So in Northern Ireland people say that rather than having a ‘power-sharing’ government, we have a ‘sharing out of power’, or what can be seen as a sharing between two ethnocratic parties. The biggest parties from each ‘side’ – here the Irish nationalist Sinn Fein and the ‘Paisleyite’ Democratic Unionist Party (DUP) – ‘share out’ the most important government ministries between themselves, with often secret ‘horse-trading’. But that can break down especially before elections, and ‘stupid ethnocracy’ returns. (The DUP for instance behaving as if Northern Ireland is still the unionist ethnocracy of their memories or dreams, though alternatively, there’s the local joke that ‘Irish unionists are too stupid to know when they’ve won, and Irish nationalists are too smart to admit they’ve lost’.) The previous (actually quite clever) leader of the DUP did argue that unionists had ‘won the war’, the Good Friday Agreement of 1998 confirmed the continuation of the union with Britain, only a minority of Catholics now supported a United Ireland, and the hence the DUP should now be open and welcoming to Catholics. But that original ploy fell apart when (presumably for electoral purposes) he and his party helped ferment a petty, purely symbolic but toxic dispute about flying the British Union Jack over Belfast City Hall 365 days a year – a dispute which they predictably lost.

So, unstable ethnocracy can have a third outcome – not simply the continuation of lethal conflict, nor the ending of it in genuine democratisation and sharing, but something in-between: a consociational ‘shared post-conflict ethnocracy’ where conflict is (mostly) non-lethal but the reality (inverting the aphorism of Clausewitz) is that the politics is the war continuing ‘by other means’.

‘Post-National’ and Religious Ethnocracy

So far we have proceeded without looking at the slippery nature of ethnicity and the divisive problems that can arise from how it is defined and used for political ends. It seemed enough to ‘know’ that the Sri Lankan majority are Sinhalese who are Sinhala-speaking Buddhists while their minority Tamil opponents are Tamil-speaking Hindus. But while a language may symbolise an ethnic or ethno-national group identity, perhaps not all putative members actually use or even understand it. For some it is mainly or merely a tokenistic badge or marker of a particular political nationalism (despite often being in fact transnational). For others, however, language is mainly of value in its own right, an end in itself rather than a means to an end, and perhaps more important than nationalism – they are sometimes called ‘cultural nationalists’ or are language activists rather than ethno-national activists. Obviously the lines are blurred – indeed the same person or group may move from one category to the other (and back again) – but ‘ethnic politics in the service of language’ clearly has a different logic from ‘language in the service of ethnic politics’12.

The same applies to religion, and indeed more so. It is generally more important and more problematic as the basis of ethnocracy. It may be used simply as a marker of a particular ethnic group, but it is generally transnational or universal in ethos with strong objectives of its own. Most ethno-national conflicts are not centrally about religion per se – many participants are irreligious – nevertheless religion is often an important factor in shaping these conflicts – and their intractability. Religions have a material institutional existence rooted in everyday life, and they can erect strong social boundaries between groups (e.g. a religious equivalent of bi-lingualism is beyond most people’s imagination though some Indian Hindus managed it). While often supporting peace and reconciliation, religions also have substantial historical associations with violence. Ideologically they tend to deal with ‘matters of life and death’ and in dogmas of ‘absolute truth’ which can preclude a useful or peaceful discussion of disagreements. Arguably this can encourage violent conflict, sometimes with largely religious motivation (see O’Dowd and McKnight 2015)13.

So, not surprisingly, relations with nation and national ethnocracy are complex, ambiguous and shifting: religious ethnocracy is an unsettled variant developing through different routes. For example, it may come from combining religion with tribal conflict, as in Kaduna, Nigeria; it may ‘spill over’ from national ethnocracy, as in Israel-Palestine; or it may spring ‘post-national’ from the failures of secular nationalism, most obviously from the ashes of the pan-Arab national movement.

Religion-based ethnic conflict and ethnocratic ‘solutions’ may arise to some extent separately from national movements. This was the case in Kaduna, although the British Empire, by creating Nigeria with its north-south, Muslim-Christian split, and the unequal treatment of different ethnic groups by imperial ethnocracy and then the Nigerian state, provided the crucial context. The present sometimes violent conflict in Kaduna since the 1980s has been fuelled by hostilities between Pentecostal Christianity and Reform Islam but has centrally been about differential tribal access to socio-economic resources, (un)employment, and state positions and power (see Harris 2013). It is partly about getting/keeping control of the existing state, not an ethno-national struggle to re-shape or partition the state on religious lines, though it could conceivably transmute into such a struggle (perhaps similar to Sudan’s).

Religious motivations, objectives and justifications are increasingly mixed in with national secular ones within some national ethnocracies. We see this in Israel-Palestine where Orthodox and ultra-Orthodox religious radicals who comprise less than 30% of the total Jewish population have become increasingly powerful since the 1970s; they make up nearly all the West Bank settlers, and have been pushing towards a theocracy. And secular Palestinian politics have also experienced religious radicalisation (see Yiftachel and Roded 2010). Although the main conflict is still clearly ethno-national, religious factors sharpen it but also sometimes ‘spill-over’ to become more important than secular ones.

Where religious institutions have a long history of active involvement in sometimes violent politics (e.g., the Croatian Catholic Church), differentiating ‘religious’ from ‘national’ objectives, or distinguishing different types or historical stages of ethnocracy, can be difficult, a merging of priorities and continuity the apparently dominant realities14. Nevertheless, such differentiations are important for understanding ethnocracy. Thus in the case of some Islamic groups, the political reassertion of religion has been directly at the expense or as a consequence of failed or partly failed secular nationalist movements (e.g. secular pan-Arab nationalism, whose last representatives have included Saddam Hussein, Colonel Gaddafi, and now Assad in Syria, hardly a reassuring trio). The over-used term ‘post-national’ does seem appropriate for their radical Islamic successors or opponents. There is also the fact that the traditional Muslim concept of the ummah, or the whole community of Islam, is trans-ethnic, transnational and potentially universal in character, and this is to some extent reflected in the non-territorial politics of some jihadists who fight under such banners as Al Qaeda.

However the inverted commas around ‘post-national’ are retained because ours is still an era dominated by national states and nationalisms, and to a varying extent these groups (e.g. Hamas, Hezbollah, the Afghan Taliban, Al-Shabaab, Islamic State) are sometimes quite nationalistic or at least territorially particularistic in their objectives. Often there are tensions between religious objectives on the one hand, for example the widespread or ‘universal’ application of sharia law in opposition to ‘Western or Christian imperialism’, and, on the other hand, more immediate political objectives such as the creation of an Islamic coalition government in a particular territory, with the issues further complicated by groups acting as proxies to advance the geopolitical ambitions of particular regional powers, perhaps ‘dressed up’ in religious garb. The Shia Hezbollah of southern Lebanon got a lot of political kudos and leverage as the national defenders of Lebanon against Israeli invaders; and geopolitically it is an active proxy army for its major funder Shia Iran, more recently in the Syrian conflict.

So religion as a marker and component of ethno-national belonging may transmute into something much more substantial. The categories are fluid rather than fixed and we need to be alive to the possibility that movements or regimes may switch category. While the dominant historical trajectories have been for some ethnic groups to transmute into ethno-national ones, and religion to go from being a pre-national force in its own right to becoming a national marker, there is no reason to believe that history is one-directional. On the contrary, it clearly is not. While historical processes are never literally reversed or repeated (change is not merely cyclical or ‘circular’ for there are always new elements involved, and ‘spiral’ is perhaps a better metaphor), nevertheless there is widespread evidence of a reassertion or re-juvenation of religion as an end in itself, a political force on its own behalf across different faith groups, most notably in radical Islam but also in Christian fundamentalism.

The concept of ethnocracy and its five variants help unify different aspects and situations of ethno-national conflict and ethnically-biased regimes into a more coherent whole: national and urban ethnocracy in relation to contested states and divided cities; the cities’ origins at the edge of traditional territorial empires, and the often formative influence of imperial ethnocracy on later national or post-national developments; the continuation of more muted ‘post-conflict’ ethnocracy in consociational power-sharing situations; and the possibilities of a ‘post-national’ ethnocracy and religious motivations predominating over national ones. With all five types we find they are as much the cause as the consequence of conflict, often operating in zero-sum terms which are ultimately self-defeating and damaging for all sides. So we need to understand their conflict-generating and unstable character, and their inherent tendency to deepen conflict unless countered by internal and external forces. Understanding how they operate in terms of their specific dynamics in the different situations can help in reducing conflict or transforming it into something more positive. Less optimistically, the relevance of the concept may become more prevalent if reactions against perceived ‘outside forces’ continue to mobilise around the chauvinistic ‘purities’ of ethnic identity.


Anderson J. 2013, ‘Imperial ethnocracy and demography: Foundations of ethno-national conflict in Belfast and Jerusalem’, in Pullan W. and Baillie B. (eds.), Locating Urban Conflicts: Ethnicity, nationalism and the everyday, Palgrave Macmillan. Basingstoke, pp. 195-213.doi: https://doi.org/10.1057/9781137316882_11 and https://doi.org/10.1057/9781137316882.0019

Anderson J. and O’Dowd L. 2007, ‘Imperialism and Nationalism: The Home Rule struggle and border creation in Ireland, 1885-1925’, Political Geography, vol. 26, no. 8, pp. 934-950. doi: https://doi.org/10.1016/j.polgeo.2007.10.001

Bollens S. 2007, Cities, Nationalism, and Democratization, Routledge, London.

Coward M. 2004, ‘Urbicide in Bosnia’, in Graham S. (ed.), Cities, War and Terrorism: Towards an urban geopolitics, Blackwell, Oxford, pp. 154-171. doi: https://doi.org/10.1002/9780470753033.ch9

Dumper M. 1997, The Politics of Jerusalem since 1967, Columbia University Press, New York.

Harris C. 2015 ‘Violence in a religiously divided city: Kaduna, Nigeria – From the Shari’a Riots of 2000 to the post-election clashes of 2011’, in O’Dowd L. and McKnight M. (eds.), Religion, Violence and Cities, Routledge, London, pp. 24-39.

Hepburn A. 2004, Contested Cites in the Modern West, Palgrave Macmillan, Basinstoke. doi: https://doi.org/10.1057/9780230536746

Magnusson W. 2013, Politics of Urbanism: Seeing like a city, Routledge, London.

Mann M. 2005, The Dark Side of Democracy: Explaining ethnic cleansing, Cambridge University Press, Cambridge.

Mazrui A. 1975, Soldiers and Kinsmen in Uganda: The making of a military ethnocracy, Sage, London.

McGarry J. and O’Leary B. 1995, Explaining Northern Ireland: Broken images, Blackwell, Oxford.

O’Dowd L. and McKnight M. 2015, Religion, Violence and Cities, Routledge, London.

Pappe I. 2007, The Ethnic Cleansing of Palestine, Oneworld, Oxford.

Piterberg G. 2008, The Returns of Zionism: Myths, politics and scholarship in Israel, Verso, London.

Robinson J. 2006, Ordinary Cities: Between modernity and development, Routledge, London.

Smooha S. 2002, ‘Types of democracy and modes of conflict management in ethnically divided societies’, Nations and Nationalism, vol. 8, no. 4, pp. 423-431. doi: https://doi.org/10.1111/1469-8219.00059

Traboulsi F. 2007, A History of Modern Lebanon, Pluto Press, London.

Yacobi H. 2009, The Jewish-Arab City: Spatio-politics in a mixed community, Routledge, London.

Yacobi H. and Pullan W. 2014, ‘The geopolitics of neighbourhood: Jerusalem’s colonial space revisited’, Geopolitics, vol. 19, no. 3, pp. 514-539. doi: https://doi.org/10.1080/14650045.2013.857657

Yiftachel O. 2006, Ethnocracy: Land and identity politics in Israel/Palestine, University of Pennsylvania Press, Philadelphia.

Yiftachel O. and Roded B. 2010, ‘Abraham’s Urban Footprint: Political geography and religious radicalism in Israel-Palestine’, in AlSayyad, N. and Massoumi, M. (eds.) The Fundamentalist City? Religiosity and the Remaking of Urban Space, Routledge, New York, pp. 177-207.


The main basis for my work on ethnocracy was the research project “Conflict in Cities and the Contested State: Everyday Life and the Possibilities of Transformation in Belfast, Jerusalem and Other Divided Cities” www.conflictincities.org [and UK funding by ESRC Large Grant RES-060-25-0015 is gratefully acknowledged]. Wendy Pullan in Cambridge and Mick Dumper in Exeter led the Jerusalem team of researchers; a Queen’s University colleague Liam O’Dowd and myself led the Belfast team. We had PhD students in all three universities who did theses on ‘other’ ethno-nationally divided cities, including Brussels, Mostar, Nicosia, Beirut and Kirkuk; and we touched on other places like Sarajevo, Vukovar, Montreal, and historic Danzig (Gdansk) and Trieste. My thanks to all project members, and to Oren Yiftachel and Alexander Koensler and two anonymous reviewers for their helpful comments.

Statement of Conflict of Interest

I am a guest editor of this issue. The reviews of this paper were managed by my co-editor James Goodman and were conducted offline by anonymous reviewers.


My thanks to the University of Technology Sydney for a Distinguished Visitors Scholarship which helped to finance a three-month visit to Australia in 2014-15.



1 A version of this article was given as a public lecture at the University of Technology Sydney in November 2014. My interest in ethnocracy started a decade earlier researching ethno-nationally divided cities – see Acknowledgements for more details.

2 A recent example of damaging divisions within the dominant ethnos is the Israeli police discriminating against Jews of Ethiopian origin. (Guardian 31 August 2016)

3 Arguing informally, I have defended Yiftachel’s use of ‘ethnocracy’ on the grounds that ‘apartheid’ is a narrower, more specific or limited, though more politically-charged, category. He himself (Yiftachel 2006, p.92, pp.125-8, pp.303-4) suggested Israel experiences ‘a process of creeping apartheid’ which increases ethnocracy; and he preferred ‘creeping apartheid’ to ‘full apartheid’ because ‘Israel proper’ still retained substantial democratic elements. As argued above, however, a democratic element is not necessary to qualify as an ethnocracy – Idi Amin’s dictatorship or autocracy was an ethnocracy. Apartheid is best seen within the wider context of ethnocracy which is more analytically useful and flexible especially when extended to other situations beyond the national (and a similar argument applies to ethnocracy vis à-vis the more limited ‘ethnic democracy’).

4 On Sri Lankan genocide, see Balasundaram in this issue. In general, however, Mann’s (2005, p.17) term ‘murderous ethnic cleansing’ is preferable. ‘Genocide’, although a legal term, is problematic, requiring proof of intentionality to kill a particular group, and involving the question of whether it can apply to only a part of the group, and how large a part that has to be. It clearly applies to the Nazi genocide of Jews and also Roma, but not to the ‘murderous ethnic cleansing’ of Palestinians which primarily aimed to remove not kill them (see Pappe 2007, p.3).

5 The central issue/mechanism was countering the generally higher Catholic birth-rate and retaining the ‘2 to 1’ Protestant majority by using employment discrimination to force disproportionate numbers of Catholics to emigrate to find work.

6 In February 2015, shortly after the Mahinda Rajapaksa government was replaced by Maithripala Sirisena promising reforms, I interviewed and got a wealth of information from twelve Tamils – politicians, journalists, business and professional people - mainly in Jaffna in northern Sri Lanka but also in Colombo. Despite the ‘Tamil Tigers’ secessionist war ending in defeat in 2009, the Sri Lankan Army continues as an ‘army of occupation’ in the Tamil north and for reasons that have little to do with ‘security’ but everything to do with consolidating ethnocratic oppression and exploitation. The army is involved in commercial enterprises; much of the land it confiscated has not been returned to the rightful owners.

7 The Tamil interviewees welcomed Rajapaksa’s overthrow, but were not sufficiently confident about future developments to have their names made public. They were sceptical about Sirisena’s ability to deliver a meaningful reduction of ethnocracy because of his continued reliance on Sinhalese votes, and the continued presence of the effectively Sinhalese Army in Tamil territory. We visited a Tamil family living on infertile sandhills because its farmland remained confiscated six years after the war ended. To date the scepticism about Sirisena’s reforms seems fully justified.

8 However, non-explicit de facto ethnocracy is often obvious, as when Goan Christians find the supply of beef has been reduced and prices have rocketed because of the actions of the Indian Hindu Nationalist government – the Christians are beef-eaters while Hindus supposedly are not (Goan Herald 24 March 2015). It seems the ‘BJP’ regime of Narendra Modi is beginning to turn the world’s largest democracy into an ethnocracy. Sometimes, however, the biases are ‘hidden’ within the state apparatus and only emerge later, as when recently released British state papers for Northern Ireland in the 1980s confirmed suspicions that supposedly ‘neutral’ British civil servants were strongly pro-unionist: they believed they had ‘to depend on unionists’ to run Northern Ireland – see imperial ethnocracy, below (Irish News 31 August 2016).

9 Scott Bollens has carried out several detailed studies of urban governance in national ethnocracies. Urban ethnocracy also occurred in imperial ethnocracy but that is not dealt with here.

10 It is an anarchistic ‘anti-state’ perspective, and it can be argued that it seriously underplays the role of states in over-seeing if not dictating urban policies. Nevertheless it captures the fact that power in cities is much more de-centred, and the basic ways in which ‘ordinary’ or ‘normal’ urban politics are qualitatively different from those of the state.

11 See Acknowledgements below.

12 For example, the Gaelic League established in 1893 to restore the Irish language was initially neutral in party political terms believing the language belonged to all, Irish unionists as well as nationalists. But between 1912 and 1915 it was effectively taken over by its dominant nationalist component to further the specifically political aims of Irish republicans.

13 For our ‘Divided Cities’ project, (see Acknowledgements), we had an international workshop on ‘Religion, Violence and Cities’ with a Special Issue of Space and Polity organised by Liam O’Dowd and Martina McKnight (2015). It is perhaps not coincidental that our two least violent divided cities, Brussels and Montreal, are the two where language rather than religion constitutes the main dividing line.

14 The Croatian ultra-nationalists who led the 1991-95 war against Serb forces with active backing from the Croatian Catholic Church harked back to the 1941-45 Nazi puppet regime of the Ustashe fascists. As well as murdering several hundred thousand people, mostly Serbs but also Jews, Roma and others, the Ustashe attempted to forcibly convert Orthodox Serbs to Catholicism. After 1995 the mountain top from which Croat forces had bombarded the Muslim section of Mostar (its famous Old Bridge included) became the site of a huge mountain top cross, mimicking the Corcovada statue in Rio but with its own unsubtle message to the Muslims below.

Share this article: