Power-Sharing, Federalism, and Diversity

Learning Objectives

By the end of this module you will be able to:

Explain consociationalism and identify the structural conditions under which it stabilizes divided societies.
Compare federalism and territorial autonomy as tools for managing ethnic and linguistic diversity — including the secession paradox.
Describe the Ottoman millet system and evaluate its applicability as a template for pluralist religious governance.
Explain at least two indigenous governance models (Iroquois/Haudenosaunee, Inuit, Ainu) and identify the deliberative mechanisms they use.
Explain how ethnic stacking in security forces functions as a coup-proofing mechanism and its systemic risks.
Apply the intersectionality framework to analyze how proposed constitutional arrangements distribute power across overlapping social groups.
Evaluate Zomia as a case study in deliberate state-avoidance and what it implies for polity design.

Core Concepts

Ethnic conflict is not fate

The starting assumption of this module is also its most politically consequential: ethnic conflict is not an inevitable consequence of ethnic diversity. Many multiethnic societies maintain stable coexistence. Conflict becomes more likely under specific structural conditions — winner-take-all institutions, elite ethnic mobilization, resource scarcity that creates zero-sum perceptions, or colonial legacies that hardwire ethnic hierarchies into administrative categories.

This matters for design. If conflict were primordial and inevitable, institutional engineering would be irrelevant. Because it is contingent, the choices founders make about representation, autonomy, and resource allocation are load-bearing.

The colonial baseline problem

Before designing around existing ethnic categories, it is worth asking where those categories came from. Colonial classification systems — census categories, "Scheduled Tribes" in India, "hill tribe" designations in Southeast Asia — created administrative identities that post-colonial states inherited and institutionalized rather than dismantled. These colonial epistemologies continue to structure who gets recognized, who gets land rights, and who is treated as a contemporary political actor versus a historical artifact. Designing around inherited categories is not neutral: it can reproduce the power arrangements those categories were invented to maintain.

Arend Lijphart's consociational model is the most studied response to deeply divided societies. It rests on four structural pillars:

Grand coalition: executive power is distributed among representatives of all major groups, not just the plurality winner.
Segmental autonomy: each community governs itself in cultural domains — schools, media, religious institutions — without requiring majority approval.
Mutual veto: minorities can block decisions that threaten their vital interests, preventing majoritarian override.
Proportionality: government posts and public resources are allocated in proportion to group population, not winner-takes-all.

The mechanism is fundamentally one of elite cooperation: stable democracy in divided societies is maintained by ensuring that leaders of each group have credible reasons to keep each other in the coalition rather than defect to ethnic outbidding.

The limitation is also clear: rigid consociational arrangements may not durably resolve deeply intractable conflicts unless accompanied by trust-building processes that allow the power-sharing architecture to eventually transition toward more standard democratic procedures. Consociationalism is a holding structure, not an end state.

Federalism and territorial autonomy: the secession paradox

Federalism and regional autonomy reduce fears of domination by guaranteeing nondominant groups control over meaningful policy domains within defined territories. This increases system support among groups that would otherwise feel dominated by a central majority.

But the design faces a structural paradox: federalism can simultaneously serve as a solution to ethnic conflict and as a pathway toward secession if not carefully designed. Granting autonomy strengthens the governance capacity of regional groups — which may eventually make exit look more feasible than staying.

The design response is to ensure that the material and political interests groups have in remaining part of the broader state structure are explicitly reinforced. Autonomy must strengthen commitment to the overarching polity, not substitute for it. This requires attending to what the center offers that the region cannot replicate: fiscal transfers, military security, trade access, diplomatic recognition.

Federalism works when regions govern themselves. It unravels when they no longer need the center.

The Ottoman millet system: non-territorial religious autonomy

The millet system offers a structurally different answer to the pluralism problem. Rather than territorial autonomy (federalism) or executive power-sharing (consociationalism), the Ottomans organized diversity along confessional lines: each religious community — Greek Orthodox, Armenian, Jewish, and later others — administered its own courts, schools, welfare institutions, and communal leadership, with religious leaders responsible for tax collection and internal order.

The key feature is non-territoriality: millets were not geographic units. A Jew in Istanbul and a Jew in Sarajevo were both subject to the same millet jurisdiction for personal-status law, regardless of where they lived. This made the system work across dispersed populations without requiring any group to hold a contiguous territory.

Three design details worth noting:

It was a deliberate political innovation, not ancient tradition. Recent historiography shows the millet system was a later Ottoman creation introduced in the guise of established practice. It was invented to solve a governance problem, which means it can be reinvented.
Formalization came late. Before the Tanzimat reforms of the nineteenth century, what we call "millets" were informal arrangements — religious minorities granted significant autonomy without an overarching systematic structure. The formal millet system was a rationalization of ad hoc practice under modernization pressure.
The currency of control was taxation and order, not assimilation. The empire collected the cizye (poll tax) from non-Muslim communities and delegated internal governance entirely to confessional elites. It did not require religious conformity.

What the millet system did not do

The system organized religious diversity, not ethnic diversity. When ethno-nationalist movements emerged in the nineteenth century, the millet framework had no mechanism to address demands framed in ethnic rather than confessional terms. The collapse of the Balkan millets into nationalist conflicts is partly a story of a religious governance model encountering a political movement it was not designed to handle.

Indigenous governance: consensus without elections

One of the persistent errors in comparative political theory is treating electoral competition as the defining feature of democratic governance. Indigenous governance systems, documented across societies from the Muscogee Confederacy to Great Zimbabwe, operated through consensus-based decision-making in which consultation was paramount, leaders served as mediators guiding communities toward collective decisions, and leadership was earned or inherited within families, clans, and nations rather than determined by popular vote.

This is not a lesser version of democracy. Scholarly research on Indigenous North American governance demonstrates that non-elective systems of representation can achieve democratic goals — broad participation, accountability, consent-based decision-making — as effectively as or better than electoral systems. What indigenous governance optimizes is different: not preference aggregation through competitive voting, but collective deliberation until consensus is reached or dissent is acknowledged and accommodated.

Three concrete instantiations:

Ainu (Japan): Traditional Ainu society had no concept of private land ownership. Kotan (villages) held collective membership rights to settlements and fishing grounds. Judicial functions were not entrusted to individual chiefs; instead an indefinite number of community members sat collectively in judgment on criminal matters. Governance was distributed, not delegated.

Inuit / Nunavut (Canada): Nunavut's consensus government institutionalizes a hybrid of Westminster parliamentary structure with Inuit cultural values — all members elected as independents, no political parties, and consensus-building as the governing norm. This is a real contemporary example of deliberately fusing indigenous governance traditions with representative democratic institutions.

Haudenosaunee / Kahnawake: The Kahnawake Community Decision Making Process integrates traditional Haudenosaunee principles with modern deliberation, emphasizing both individual thinking and unanimity. This demonstrates that consensus-based processes are not historical artifacts — they are functional, contemporary governance mechanisms.

Ethnic stacking in security forces

Any treatment of diversity and institutional design must confront the security sector. Ethnic or sectarian stacking involves recruiting the officer corps disproportionately from the ruler's ethnic, tribal, regional, or sectarian group, ensuring that an alternative coup-coalition cannot easily form because officers from excluded groups cannot coordinate across communal lines.

The documented cases are instructive:

Assad's Syria: Alawite officers disproportionately represented at senior levels.
Saddam's Iraq: Tikrit-based and Sunni officers controlling security apparatus.

The mechanism is logical from a regime-survival standpoint: cross-ethnic conspiracy is harder to coordinate than within-group conspiracy. But the systemic consequences are severe. Ethnically stacked security forces create:

Institutional grievance: excluded groups read the composition of the military as a statement about their political standing.
Military ineffectiveness: promotion based on ethnic loyalty rather than competence degrades operational capability.
Fragility at transition: when the regime collapses, the security sector is a site of violent ethnic score-settling rather than a neutral instrument of continuity.

Design trap

Ethnic stacking solves a coup-proofing problem for the current ruler at the cost of state coherence for everyone who comes after. A new polity designer faces an inverse version of this trap: how do you build a security sector that no single ethnic group controls but that is still capable of coherent action?

Intersectionality as an analytical tool for institutional design

The intersectionality framework, developed through Patricia Hill Collins's work, offers a diagnostic lens for assessing whether proposed governance arrangements will reproduce existing hierarchies or disrupt them.

Collins's matrix of domination organizes how power operates across four domains:

Structural: laws, policies, institutions — the formal architecture.
Disciplinary: bureaucratic hierarchies and administrative practices that manage oppression without explicit legal rules.
Hegemonic: ideology and culture that normalize dominant group perspectives as common sense.
Interpersonal: daily routines and interactions that perpetuate subordination.

A constitutional arrangement can look egalitarian at the structural level while reproducing hierarchy through disciplinary or hegemonic mechanisms. Reserved seats for minority groups (structural domain) can coexist with civil service hiring practices that exclude them (disciplinary domain) and with cultural narratives that portray them as unfit for governance (hegemonic domain).

The design implication is that transformative institutional design requires shared analysis of how multiple oppressions are produced by interconnected systems rather than isolated phenomena. A constitution that addresses ethnicity but ignores how gender, caste, and class intersect with ethnicity will produce predictable failure modes.

Constitutional equality on paper is necessary but never sufficient. The question is which domains of power the text leaves untouched.

The Indian case is the most studied example of this dynamic. Patrick Heller's comparative research on Brazil, India, and South Africa shows that formal democratic procedures — elections, voting, representation — can coexist with persistent caste-based hierarchies that circumscribe basic rights of association. Heller terms this "social authoritarianism": routinized social exclusion rooted in categorical inequalities, operating beneath the surface of constitutional equality.

Social Dominance Orientation (SDO) — the individual-level preference for group-based hierarchy — is one of the strongest predictors of intergroup attitudes. High-SDO populations will consistently find ways to maintain hierarchies even when formal rules prohibit them. The design implication is that institutional rules alone cannot overcome deeply-held social preferences for hierarchy; structures must also create incentives and paths for cross-group contact.

This connects to the distinction between bridging and bonding social capital. Bridging social capital (weak ties across heterogeneous groups) reduces polarization and enables cross-cutting networks. Bonding capital (strong ties within homogenous groups) reinforces exclusive identities. A polity with strong bonding and weak bridging capital will tend toward ethnic or caste entrenchment regardless of constitutional architecture.

Zomia: state-avoidance as deliberate design

Zomia is James Scott's term for the upland region of Southeast Asia — spanning Northeast India, Myanmar, Laos, Thailand, and Vietnam — home to approximately 100 million people who historically maintained autonomy through deliberate cultural, agricultural, and political strategies designed to resist state incorporation. The Akha, Hmong, Karen, Lahu, Mien, and Wa peoples are among the groups involved.

Scott's argument, developed in The Art of Not Being Governed, is that these communities were not "pre-state" peoples awaiting modernization. They were "barbaric by design" — maintaining egalitarian political structures, mobile agricultural practices (swidden farming), fluid ethnic identities, and oral rather than written traditions as mechanisms to discourage state control and domination. Geographic settlement in marginal upland areas, mobile subsistence, multi-ethnic federations, and prophet-led rebellions were all components of a deliberate state-avoidance toolkit.

For a polity designer, Zomia poses an uncomfortable question: what does it mean to build a state for populations that have historically organized their survival around not being governed? And more generally: if communities have good reasons not to want state incorporation, what would a legitimate state have to offer them that they couldn't produce themselves?

Pancasila: a non-liberal democratic frame

Indonesia's national ideology offers a different lens on diversity management. Pancasila merges belief in God with social justice, democracy, and national unity in a framework that is explicitly non-liberal — grounding political legitimacy in cultural tolerance, religious accommodation, customary law, and spiritual republicanism rather than secular liberalism. Under Pancasila, popular sovereignty governs political procedure while religious values shape policy content — a procedural-substantive distinction that allows multiple religious traditions to contribute to governance without any single tradition overriding constitutional procedure. The catch: Indonesia's post-Reformasi experience shows that electoral democracy and liberal democracy are not synonymous — formal procedures can survive while minority rights and press freedom are eroded.

Indigenous land tenure as governance architecture

Any discussion of indigenous governance needs to grapple with land. Indigenous land tenure is fundamentally collective and communal rather than individual — encompassing lands, waters, and non-human entities in a comprehensive social frame of reference that is basic to indigenous identity and existence.

The scale is not trivial. Indigenous Peoples and local communities collectively hold approximately 20 percent of the Earth's land mass — but formal legal ownership covers only 10 percent of global land. An additional 8 percent carries government-recognized management rights, and at least one-third to one-half of the world's land remains under customary tenure without formal recognition. This gap between de facto governance and de jure ownership is a pervasive source of institutional conflict.

Research demonstrates that tenure security for indigenous lands strengthens stewardship and conservation outcomes — and that indigenous- and community-held lands outperform many state-managed protected areas on biodiversity metrics. This is not an argument from tradition or cultural preservation alone; it is an argument from governance effectiveness.

Ecuador (2008) and Bolivia (2009) went furthest in institutionalizing this recognition, incorporating indigenous Buen Vivir frameworks into their constitutions — recognizing indigenous epistemologies not as cultural heritage but as frameworks for governance and state legitimacy. Whether this constitutional recognition translates into durable institutional change remains contested.

Annotated Case Study

Rojava's Democratic Confederalism: multi-ethnic institutional design under pressure

Rojava (the Autonomous Administration of North and East Syria) is a live experiment in institutional pluralism operating under extraordinary constraints — active military conflict, no formal international recognition, and a diverse population including Kurds, Arabs, Syriac Christians, Yezidis, Turkmen, and Armenians.

Rojava's constitutional frameworks explicitly institutionalize multi-ethnic and multi-religious inclusion, with formal representation mechanisms for each community in governance councils and committees. The governing ideology — Democratic Confederalism — rejects Kurdish nationalism in favor of a "democratic nation" framework that prioritizes shared democratic participation over ethnic identity as the organizing principle of political community.

What makes this case analytically useful:

Representation mechanisms without territorial monopoly. Each ethnic and religious community has formal representation at every administrative level, not just in a single autonomous zone. This is structurally closer to the millet model (non-territorial community autonomy) than to federalism (territorial autonomy), but it operates within a shared polity rather than an imperial hierarchy.

The democratic nation frame vs. ethnic nationalism. Most multi-ethnic constitutional arrangements acknowledge group identities and allocate seats accordingly. Rojava goes further by explicitly rejecting ethnic nationalism as a legitimating principle — the political community is defined by participation in democratic structures, not by ethnic membership. This is a more ambitious claim, and more fragile: it depends on all groups accepting a shared framework rather than retreating to communal identity when under pressure.

Bridging institutions under stress. The co-chair system (requiring male-female co-leadership at every level) and the multi-ethnic committee structure create formal bridging mechanisms. Whether these survive the consolidation pressures of military governance is an open question.

What the case does not resolve:

The experiment remains unrecognized internationally and faces continuous external military pressure. It is not yet possible to assess long-term durability. It also operates within a particular revolutionary moment — the post-ISIS security vacuum — that may not transfer to peacetime institutional design. But as a demonstration that explicit, constitutionally-mandated multi-ethnic inclusion is organizationally feasible under extreme conditions, it is instructive.

Compare & Contrast

Three models for managing religious or ethnic diversity

Dimension	Consociationalism	Millet System	Federalism
Organizing principle	Elite coalition across groups	Confessional community autonomy	Territorial self-governance
Who gets autonomy	All major groups in executive coalition	Defined religious communities	Regional populations (may be ethnically defined)
Territorial?	No — cross-group executive	No — community-based, geography-independent	Yes — requires geographic concentration
Decision rule	Mutual veto + proportionality	Internal community law; imperial authority over external affairs	Federal law + regional law, divided by domain
Failure mode	Elite defection; outbidding; gridlock	Nationalist movements the system wasn't designed to handle; hierarchical ordering of communities	Secession; the autonomy-as-stepping-stone paradox
Historical example	Belgium, Lebanon, post-GFA Northern Ireland	Ottoman Empire (pre-Tanzimat informal, post-Tanzimat formal)	Switzerland, India, Nigeria, Canada
Indigenous parallel	Haudenosaunee confederacy structure	Millet = community autonomy without territory; close to indigenous customary jurisdiction models	Nunavut as territorial self-governance

The non-territorial advantage

Both the millet system and indigenous communal governance models can operate without territorial concentration — they govern people through membership in a community rather than residence in a zone. This is relevant for polity designers working with dispersed populations: diaspora communities, nomadic groups, or digital-native communities that do not map neatly onto geography.

Common Misconceptions

"Ethnic conflict is natural and ancient." The evidence consistently shows that ethnic conflict is politically constructed and contingent. Many of the "ancient hatreds" underlying contemporary conflicts are recent inventions, often shaped by colonial administrative categories that hardened fluid identities into fixed ones. Design matters precisely because outcomes are not predetermined.

"The millet system was a timeless Islamic institution." It was a political innovation of the Ottoman state, presented in the guise of tradition. Historiographic research has corrected the claim that it represented ancient Islamic governance practice. This is useful: it means the model can be adapted and reinvented without deference to a canonical original.

"Indigenous governance is pre-political or primitive." Indigenous governance systems developed sophisticated deliberative institutions — keystone institutions for broad power distribution, consensus processes, distributed justice. They are not waiting to be replaced by modern democratic systems; in some dimensions they are more democratic than electoral systems that aggregate individual preferences without requiring genuine deliberation.

"Federalism solves the secession problem." Federalism is a secession risk management tool that can itself generate secession risk if the center does not maintain compelling reasons for regions to stay. The paradox is structural, not accidental. Federalism must be designed alongside the affirmative bonds that make union preferable to independence.

"Formal constitutional equality is enough." Electoral democracy and liberal democracy are not the same thing. Competitive elections can coexist with social authoritarianism — the persistence of caste, ethnic, or gender hierarchies that render constitutional equality hollow. Formal rules operate within social structures that institutions alone cannot transform without deliberate attention to all four domains of power.

Key Takeaways

Ethnic conflict is a design problem, not a destiny. It becomes more likely under specific institutional conditions — winner-take-all structures, elite ethnic mobilization, colonial classifications that freeze identity — and less likely when institutions distribute power and resources credibly across groups.
The three major pluralism frameworks (consociationalism, federalism, millet-style community autonomy) solve different problems. Consociationalism is about who holds executive power. Federalism is about who controls territory. The millet model is about who governs community life regardless of where people live. None of them is universal; each generates its own failure mode.
Indigenous governance is a design library, not a museum exhibit. Consensus-based deliberation, collective resource tenure, distributed justice, and community self-governance are operational today — in Nunavut, in Kahnawake, in Chiapas. The design challenge is how to interface these systems with modern state structures without absorbing and dissolving them.
Security sector composition is a diversity problem. Ethnic stacking solves a coup-proofing problem for incumbents at the cost of long-term state coherence. Any polity design that takes seriously the management of diversity cannot treat the military as outside the architecture.
Intersectionality is an audit tool. A constitution that addresses ethnicity while leaving caste, gender, and class hierarchies untouched will reproduce them through disciplinary and hegemonic channels. Institutional designers need to trace how power operates across all four domains, not only through formal law.

Further Exploration

Consociationalism and Power-Sharing

Federalism and Autonomy

Ottoman Millet System

Indigenous Governance

Zomia and State-Avoidance

Intersectionality and Institutional Analysis