MU-M-01: The Mauritian Multicultural Model — Unity in Diversity as Statecraft (1968–2026)

Status: DRAFTWords: 10,451

1. Key Takeaways

The "Mauritian model" is real, but it is not what its official framing says it is. The state's language — unité dans la diversité, the rainbow nation (nation arc-en-ciel), the harmonious coexistence of "all the communities" — describes an aspiration. The operating system beneath it is harder-edged: a dense, mostly unwritten settlement of insurance mechanisms, distribution conventions, and deliberate silences that converts a plural society's permanent communal anxiety into manageable, routinised politics. Its components include the best-loser system's constitutionalised minority insurance (MU-I-01), a universal welfare state that distributes without ever targeting by community (MU-G-05), per-capita public subsidies to religious bodies [TBD-VERIFY: the subsidy mechanics], a Hindu-premiership convention with negotiated exceptions, a functional multilingual diglossia, and a public-holiday calendar that performs communal parity annually. The model is consociational in effect while formally being nothing more than Westminster majoritarianism plus eight compensatory seats — and that gap between form and practice is its most analytically interesting feature.
The model's core wager is that ethnic peace is a public good worth paying for in efficiency, transparency, and ideological coherence. Every component carries a cost that a textbook reformer would strike out: the BLS requires candidates to declare a community and freezes the demographic baseline at the 1972 census; universalism delivers pensions to millionaires so that no community can read targeting as favouritism; religious subsidies entangle a secular-in-form state with communal institutions; slate-balancing subordinates candidate merit to communal arithmetic; the census silence purchases peace at the price of data. The Mauritian answer, sustained across every government since 1968, has been that these costs are premiums on an insurance policy against the one catastrophe the island has actually experienced — communal violence — and that the policy has paid out: no communal killing on a 1968 scale has recurred in fifty-eight years.
The settlement was forged in fear, not in idealism, and its founding moment is the January 1968 riots, not the March 1968 flag-raising. The Port Louis communal violence weeks before independence — at least 25 dead [TBD-VERIFY: casualty figures range to 29 and above in some accounts], British troops restoring order, thousands displaced along communal lines — meant Mauritius began its independent existence having just demonstrated the failure mode its institutions would be built to prevent. Seewoosagur Ramgoolam's response was instinctively consociational: the 1969 coalition with his bitterest opponent, Gaëtan Duval's PMSD, which had campaigned against independence, brought the party of the anxious Creole and Franco-Mauritian minorities into government within two years of fighting it (MU-A-01, MU-H-PM-01). The pattern — co-opt the threatening pole rather than govern against it — became the settlement's standing operating procedure.
The model's deepest stress test came not from communalism but from the attempt to abolish it: the MMM's class politics of the 1970s. Paul Bérenger's Mouvement Militant Mauricien explicitly rejected the communal settlement, organising cross-communal class mobilisation under the slogan of enn sel lepep, enn sel nasion — one people, one nation (MU-H-PM-03). The establishment's response — the state of emergency, postponed elections, detained MMM leadership (MU-A-01, MU-B-01) — was the model defending itself by illiberal means. The long-run outcome was a double absorption: the MMM was domesticated into the alliance-rotation system it had vowed to destroy, while its mauricianisme — the ideal of a trans-communal Mauritian identity — was absorbed into the official rhetoric of the very settlement it had attacked. The model thus runs on two contradictory legitimating languages simultaneously: communal accounting in practice, post-communal nationhood in aspiration.
The 1999 Kaya riots are the model's biggest crack and its most important diagnostic. The death of the seggae singer Kaya in police custody on 21 February 1999, and the days of rioting that followed — several dead, including the singer Berger Agathe shot by police; police stations burned; the island's worst unrest since 1968 [TBD-VERIFY: casualty figures and damage estimates] — revealed what the settlement's accounting concealed: that the Creole population, and especially its Afro-Mauritian working class, had received formal inclusion (BLS seats via the "General Population" category, holiday parity, universal services) while remaining at the bottom of every socioeconomic distribution. The malaise créole — named by the Catholic priest Roger Cerveaux in 1993 — is the standing demonstration that the model insures communities as categories far better than it equalises people as members, and that a settlement built to prevent inter-communal violence has no comparable machinery against intra-society exclusion.
The model's mechanics run on deliberate ignorance as much as on deliberate counting. Mauritius has not asked an ethnic question in its census since 1972 — a silence maintained precisely so that the communal balance cannot become a live political fact subject to annual revision — yet the BLS still allocates seats by the frozen 1972 figures, candidates must still classify themselves into four constitutional categories, party slates are still balanced constituency by communal constituency, and cabinets are still composed with an unwritten communal distribution [TBD-VERIFY: the conventions of portfolio allocation by community]. The result is a polity that officially refuses to know what it unofficially counts at every level — a paradox the UN Human Rights Committee's 2012 ruling in the Rezistans ek Alternativ case forced into the open, and which the never-completed BLS reform debates (MU-I-01 §4) have left unresolved into 2026.
The comparative record is the model's strongest credential: among the sugar-colony plural societies — Guyana, Fiji, Trinidad, and (by extension) Sri Lanka — Mauritius alone avoided ethnic party-system breakdown, coups, or civil war. The literature's explanations form an inventory rather than a verdict: the absence of an indigenous population whose prior claim could trump all others (everyone in Mauritius descends from arrival); the "every community a minority" arithmetic, in which even the Hindu plurality is internally divided and no bloc can rule alone in comfort; the electoral system's coalition-forcing geography; the sustained growth that gave the elite bargain rising resources to share (MU-G-01); the smallness and density that make elites known to each other; and the early, repeated practice of co-opting rather than crushing the losing side. Brautigam, Carroll and Carroll, Srebrnik, and Eriksen each weight these differently [TBD-VERIFY: the precise claims of the key comparative works], but the consensus negative finding matters most: no single institution explains the outcome, which means the model is not exportable as a kit — a caution against the "lessons from Mauritius" genre that MU-N-01 documents.
The 2024 election can be read as the model's vindication or as the beginning of its obsolescence, and both readings are defensible. The vindication reading: a 60-of-62 alternation executed peacefully across communal lines, with the Hindu-premiership convention intact and the slate arithmetic observed, is the settlement doing exactly what it was designed to do (MU-C-01). The obsolescence reading: the scale of the sweep, the youth vote's apparent indifference to communal appeals, declining religious observance and rising intermarriage [TBD-VERIFY: intermarriage and secularisation data], and the fact that the campaign was fought on corruption, surveillance, and cost of living rather than communal protection, all suggest an electorate outgrowing the categories the constitution still requires candidates to inhabit. The model's 2030s test is whether it can retire its scaffolding gracefully — the post-communal scenario — or whether the categories persist as an ossified gerontocratic inheritance defended by those whose careers were built inside them.
The honest summary of the model in 2026: a success that cannot fully explain itself, built by politicians who never wrote it down, and now administered by a generation that no longer remembers the fear that built it. That description is not a dismissal. Unwritten settlements have outlasted written ones; the fact that the model's components were improvised, contradictory, and never codified may be precisely why they could bend through 1982's 60–0, the 1999 rupture, and 2024's landslide without breaking. But it also means the model has no canonical statement to defend it when its costs — the frozen census, the compulsory declarations, the communal slates — come under generational and international challenge. The document's conclusion is that the Mauritian model's future depends less on constitutional engineering than on whether the insurance logic it embodies still has buyers among voters born after 1999.

2. The Model's Components: The Unwritten Settlement

2.1 The Official Framing and What It Covers

The official idiom is consistent across six decades and every governing coalition: Mauritius as nation arc-en-ciel, unity in diversity, the island where — in the formula every National Day address rehearses — temple, mosque, church, and pagoda stand within sight of one another. The framing is not empty; it does real work, supplying a shared civic vocabulary that all communities can speak without conceding anything. But as a description of the governing machinery it is almost perfectly misleading, because it presents as spontaneous harmony what is in fact a maintained equilibrium — an interlocking set of insurance mechanisms, distribution rules, and strategic silences that this section inventories. The distinction matters analytically: spontaneous harmony needs no maintenance and survives neglect; a maintained equilibrium degrades when its mechanisms are allowed to ossify or its premiums go unpaid. Every stress episode in Section 4 is a moment when the gap between the framing and the machinery became visible.

The inventory that follows is deliberately a components list rather than a narrative, because the model was never designed as a whole. No founding document states it; no commission ever codified it; several components contradict one another. It accreted — a riot's lesson here, a coalition's bargain there — and its coherence is retrospective, imposed by analysts (this document included) on what participants experienced as improvisation.

2.2 The Electoral Insurance: The Best-Loser System

The settlement's only fully constitutionalised component is the best-loser system, documented in full in MU-I-01 and summarised here only for its function within the model. The BLS allocates up to eight additional National Assembly seats after each general election to the highest-polling defeated candidates of communities underrepresented relative to their population share — the four constitutional communities being Hindu, Muslim, Sino-Mauritian, and the residual "General Population" — using the 1972 census as the permanently frozen demographic baseline. Its function in the model is insurance in the strict sense: it guarantees minorities a floor of parliamentary presence regardless of how the first-past-the-post arithmetic falls, and thereby removes the catastrophic downside scenario (total exclusion) that drives minority politics toward extra-institutional channels in comparable plural societies. The premium is equally strict: every candidate must declare a community, the state must operate four ethnic categories it elsewhere refuses to count (Section 2.7), and the system embeds in the constitution exactly the communal classification the official framing claims to transcend. MU-I-01 §4 documents the litigation, the 2012 UN Human Rights Committee ruling, and the unfinished reform; for this document's purposes the point is functional: the BLS is the model's load-bearing contradiction, simultaneously its most criticised and most stabilising part.

2.3 The Welfare Component: Universalism as Communal Neutrality

The second component is the welfare state's design principle, analysed at length in MU-G-05: every major social programme — free education through tertiary level, free universal health care, the non-contributory universal old-age pension, subsidised rice and flour — is universal rather than targeted. MU-G-05 argues that this universalism is best understood as peace technology, and the argument belongs at the centre of the model: in a plural society where every distributive decision is read through a communal lens, means-testing or category-targeting would convert each benefit into a standing audit of which community receives what. Universalism pre-empts the audit. The pension that goes to every Mauritian at the qualifying age cannot be communalised; the free clinic serves whoever walks in. The fiscal cost — benefits flowing to households that do not need them — is, on this reading, not waste but premium: the price of removing the entire social-policy domain from communal contestation. The reading is borne out negatively by the exceptions: the rare interventions that did carry implicit communal address (Section 4 discusses the Creole-targeted responses after 1999) proved politically combustible in exactly the way the universalist design predicts.

2.4 The Subsidised Sacred: Public Money for Communal-Religious Infrastructure

Third, and least studied: the state directly subsidises religious bodies. Recognised religious organisations receive annual public subventions allocated broadly in proportion to their declared adherents [TBD-VERIFY: the precise mechanics — the allocation is conventionally described as per-capita against census religious-affiliation figures (religion, unlike ethnicity, is still asked in the census), disbursed to registered bodies; the statutory basis, the list of recognised bodies, and the current quantum require verification against the annual Budget Estimates]. The subsidy makes Mauritius secular in no conventional sense: the state is not separated from religion but evenly entangled with all of them. The logic is the model's standard logic. A genuinely secular withdrawal would be read by every community as defunding its institutions; targeted support would be favouritism; proportionate support to all is the only equilibrium that no community experiences as loss. The same logic extends to the subsidised cultural apparatus — the speaking unions and cultural centres established by statute for the major ancestral languages and traditions (the Hindi, Urdu, Tamil, Telugu, Marathi, Mandarin, Creole, and Bhojpuri speaking unions, among others [TBD-VERIFY: the full statutory list and dates]) — through which the state funds the reproduction of communal identity itself. The model does not merely tolerate communal institutions; it pays for them, on the theory that communities confident of their cultural survival make calmer political bargains.

2.5 The Premiership and the Conventions of Office

Fourth, the unwritten conventions governing high office. The premiership has been held by a Hindu — and, more narrowly, has circulated within a small set of Hindu political families (MU-N-01 documents the dynastic critique) — for all but twenty-seven months since 1968. The exception proves the convention: Paul Bérenger's premiership of 2003–2005 occurred only by explicit pre-electoral pact, the 2000 MSM-MMM alliance agreement under which Anerood Jugnauth served three years and then handed over to Bérenger while moving to the Presidency — a negotiated, time-limited, announced-in-advance suspension of the norm rather than its breach (MU-B-01, MU-H-PM-03). Around the premiership runs a wider lattice of balancing conventions: the presidency, the vice-presidency, the speakership, and the deputy premiership are conventionally distributed so that the top of the state, viewed as a whole, displays communal spread [TBD-VERIFY: the conventions are widely asserted in Mauritian political commentary — e.g., a Muslim or Creole president balancing a Hindu prime minister — but are nowhere codified and have exceptions; the pattern requires systematic tabulation]. The conventions are the model's elite-level mirror of the BLS: where the BLS insures minority presence, the office conventions insure minority dignity — visible occupancy of the state's ceremonial summit — while leaving effective power where the electoral arithmetic puts it.

2.6 The Multilingual Settlement: Functional Diglossia as Peace

Fifth, the language settlement, which resolves what has destroyed other plural polities by refusing to resolve it. English is the official language of legislation and the courts — the language, crucially, of no Mauritian community, which is precisely its qualification: a neutral medium nobody owns. French dominates the press, broadcasting, and prestige culture, the durable inheritance of the Franco-Mauritian century and the channel of the island's francophone identity. Kreol Morisien is the universal spoken vernacular, the actual common language of the entire population across every community. The ancestral languages — Hindi, Urdu, Tamil, Telugu, Marathi, Mandarin, Arabic, Bhojpuri — are maintained in schools and the subsidised cultural apparatus as identity markers rather than working media. The settlement is a functional diglossia (more precisely a triglossia with an ancestral fourth layer): each language holds a domain, no language is promoted into another's domain, and therefore no community's linguistic patrimony is ever put to a vote. The cost, again, is real and was borne by the most vulnerable: for four decades the language every child actually spoke — Kreol — had no place in the school system that examined them in English and French, an exclusion whose educational consequences fell hardest on the Creole working class and which Section 4.4 takes up as one of the malaise créole's structural roots. Kreol's entry into the primary curriculum in 2012 [TBD-VERIFY: introduction as an optional subject from the 2012 school year] was the settlement's first major renegotiation, and its slow upward creep — into examinations, and in 2025 into parliamentary use [TBD-VERIFY: the status of Kreol in National Assembly proceedings under the post-2024 government] — is among the few components still visibly in motion.

2.7 The Calendar, the Census Silence, and the Grammar of Parity

Sixth, the public-holiday calendar — the model's most visible annual performance. The fixed national holidays distribute recognition with arithmetical care: Thaipoosam Cavadee, Maha Shivaratree, Ugadi, Ganesh Chaturthi, and Divali for the Hindu traditions in their plural variety; Eid-ul-Fitr for the Muslim community; Christmas and the Assumption [TBD-VERIFY: which Catholic feast holds the official slot — All Saints' Day (1 November) and Assumption (15 August) both appear in lists across years] for the Christian; the Chinese Spring Festival for the Sino-Mauritian; plus the two commemorations that encode the island's two arrival histories — the Abolition of Slavery (1 February) and the Arrival of Indentured Labourers (2 November) — and the civic dates of Independence/Republic Day (12 March) and Labour Day (1 May). The calendar is communal arithmetic made festive: every community sees its sacred time given state recognition, every other community gets the day off for it, and the whole population thereby acquires a low-intensity familiarity with everyone else's festivals that anthropologists of Mauritius (Section 3.4) identify as one of the lived foundations of coexistence.

Seventh, and structurally opposite: the census silence. Mauritius last asked an ethnic question in the 1972 census and has refused to ask one since — a deliberate policy of official ignorance, maintained by every government, on the explicit logic that a decennial recount of the communal balance would convert demography into a permanent political campaign (the full treatment is in MU-I-01, including the absurdity that the BLS consequently allocates 2020s seats on 1972 numbers). The silence and the calendar together display the model's deep grammar: recognition without measurement. Communities are to be honoured, subsidised, scheduled, and balanced — but never freshly counted, because counting creates winners and losers and the model's entire purpose is to ensure that no community ever experiences itself as the loser of a definitive contest.

2.8 Ethnic Peace as Public Good

The components share one economic structure: each treats ethnic peace as a public good and accepts a visible inefficiency as the cost of supplying it. The BLS pays in liberal principle; universalism pays in fiscal targeting efficiency; the religious subsidies pay in secular principle; the office conventions pay in meritocratic principle; the diglossia paid in educational equity; the census silence pays in statistical knowledge. A reformer who removed any single component on its own-domain merits — and every component has been so challenged — would be defunding the insurance without pricing the risk. That, in compressed form, is the Mauritian establishment's standing answer to every reform proposal from 1968 to the present, and Section 7 asks whether it is still true.

3. The Intellectual Genealogy

The model has an intellectual paper trail older than the state. The 1960–61 Meade and Titmuss commissions — remembered chiefly for their economic pessimism (MU-N-01 §2) — both treated the plural society as a first-order governance problem, not a cultural backdrop. Meade's report judged communal division a standing threat to the stability any development strategy required, and its prescriptions assumed a state that would have to actively manage distribution across communities; Titmuss's social-policy survey, from the Fabian tradition that built the British welfare state, recommended the universalist social-provision architecture — and the family-planning programme — that Mauritius substantially adopted (MU-G-05 traces the lineage in detail). The genealogical point is that the welfare component of the model was prescribed before independence by outsiders and then retained for communal-peace reasons by insiders: Titmuss supplied the design, but the Mauritian political class discovered its second function — communal neutrality — in use. The model's characteristic pattern of repurposing (institutions adopted for one reason, kept for another) starts here.

3.2 Ramgoolam's Consociational Instinct and the MMM's Challenge

Seewoosagur Ramgoolam never used the word consociational and almost certainly never read Lijphart, but his practice between 1967 and 1976 amounts to a consociational toolkit assembled under fire (MU-H-PM-01). The independence settlement itself was a negotiated package with minority insurance built in (the Banwell process and the BLS, MU-I-01 §2); the January 1968 riots were answered not with majoritarian consolidation but with the December 1969 coalition that brought Gaëtan Duval's PMSD — the party that had fought independence with the slogan of Hindu domination to come — into government, Duval himself taking External Affairs. The 1976 election, which the young MMM won in seats, was answered the same way: a Labour-PMSD coalition assembled to hold the centre. The instinct — widen the cabinet, share the spoils, co-opt the pole that frightens your base's opponents — set the template every successor has followed: of the fifteen governments since 1968, nearly all have been pre- or post-electoral coalitions spanning communal electorates [TBD-VERIFY: exact count of single-party majority governments since 1968 — conventionally stated as zero or near-zero].

The MMM of the early 1970s was the model's intellectual antithesis and its most serious challenge — more serious, in conceptual terms, than any communal mobilisation, because it attacked the settlement's premise rather than its terms. Bérenger's movement organised dockers, transport workers, and cane labourers across communal lines, on the Marxisant theory that communalism was a bourgeois instrument for dividing a single working class, and its slogan — enn sel lepep, enn sel nasion — proposed to replace the communal nation-of-communities with a unitary Mauritian people (MU-H-PM-03). The establishment treated it accordingly: the 1971 state of emergency, the detention of the MMM leadership, the postponement of elections to 1976 (MU-A-01, MU-B-01) — the one sustained illiberal episode in Mauritian history, undertaken in defence of the communal settlement against its class-politics negation. The resolution was absorption in both directions. The MMM, from 1982 onward, entered the alliance-rotation system, balanced its own slates communally, and supplied a prime minister under the premiership convention's negotiated-exception rules; the settlement, for its part, adopted the MMM's mauricianisme as official aspiration — the unity-in-diversity rhetoric of Section 2.1 is, in large part, the domesticated ghost of the MMM's unitary nation. The model thereafter contained its own critique as a licensed internal current, which is one reason external challenges have found so little purchase.

3.3 The Consociationalism Literature and the Mauritius Puzzle

Within comparative political science, Mauritius entered the canon as a puzzle for consociational theory rather than a confirmation of it. Arend Lijphart's framework predicts that deeply plural societies achieve stable democracy through grand coalitions, segmental autonomy, proportionality, and mutual veto — and Mauritius displays the outcomes (permanent coalition government, proportional distribution conventions, communal cultural autonomy) without most of the formal apparatus: no constitutional grand-coalition requirement, no communal veto, no proportional electoral system (first-past-the-post in three-member constituencies, plus the BLS top-up), no federalism. Lijphart's own treatments classify Mauritius among the plural societies that approximate power-sharing through informal means [TBD-VERIFY: Lijphart's specific references to Mauritius — he discusses it in his later comparative work on electoral systems and power-sharing, and Mauritius appears in the consociational literature chiefly through secondary applications; the precise citations require verification]. The sharper formulations are in the case literature: Mauritius as "power-sharing without consociation" or "consociationalism by convention" — the demonstration that the consociational functions (minority security, elite cartel bargaining, proportional payoffs) can be supplied by party strategy and unwritten norms rather than constitutional design. The finding cuts both ways for theory: it vindicates Lijphart's functional claims while undermining his institutional prescriptions, since the Mauritian mechanisms could not be legislated into existence elsewhere — they are equilibria, not statutes.

3.4 The Local Tradition: Mauricianisme and the Anthropology of Coexistence

The model also has an indigenous intellectual history, running in the island's two analytic registers. The political register is the long mauricianisme debate: the recurring argument — MMM-originated but since taken up by civil-society currents, the Catholic social-justice milieu, and the post-2000 anti-communal parties (Rezistans ek Alternativ and Linion Pep Morisien above all, MU-I-01 §4) — that the communal settlement perpetuates what it claims to manage, and that a mature Mauritius should abolish the categories: no communal declaration, no BLS, no ethnic slates, one Mauritian people. Against it stands the establishment position, rarely theorised but consistently practised, that the categories are descriptions before they are constructions, and that abolishing the accounting would not abolish the anxiety — merely deprive it of institutional channels. The debate has run for fifty years without resolution and is the live wire inside every electoral-reform commission MU-I-01 §5 catalogues.

The scholarly register is dominated by the anthropology of Thomas Hylland Eriksen, whose fieldwork-based studies — Us and Them in Modern Societies (1992), Common Denominators: Ethnicity, Nation-Building and Compromise in Mauritius (1998) — made Mauritius the discipline's standard case of "communalism without conflict" [TBD-VERIFY: Eriksen's exact formulations]. Eriksen's central findings supply the model's micro-foundations: Mauritian ethnicity is omnipresent but managed — encoded in surname, neighbourhood, festival, and marriage pattern, openly discussed in everyday life, and yet systematically de-dramatised by shared idioms (Kreol as universal medium, the festivals' cross-communal observance, the workplace's enforced mixing). His "common denominators" thesis — that Mauritian unity rests not on shared origins but on accumulated shared instruments (language, schooling, consumption, the welfare state itself) — is the anthropological restatement of this document's components inventory, and his caution is also this document's: the compromise is a process, renewed or eroded daily, not an achievement that can be banked. Alongside Eriksen sit the historians and sociologists of the malaise créole literature (Section 4.4) and the Mauritian commentariat — Catherine Boudet, Sheila Bunwaree, Roukaya Kasenally, and others — whose running analysis of communalisme in the island's press constitutes the model's permanent internal audit [TBD-VERIFY: representative works].

4. The Model Under Stress: The Episodes It Survived

4.1 January 1968: The Founding Trauma

The riots that preceded independence by six weeks are the model's negative founding document — the demonstrated failure mode against which everything in Section 2 was subsequently built or retained. The violence of January 1968 in Port Louis, concentrated in the Plaine Verte and Roche Bois districts, ran principally between Muslim and Creole communities, with gang structures and turf economies entangled in the communal lines; it killed at least 25 people [TBD-VERIFY: official and scholarly casualty figures — accounts range from 25 to 29 and higher, with hundreds injured], displaced thousands along communal lines as families abandoned mixed neighbourhoods, and was suppressed only by British troops flown in from Singapore. Two features mattered for the model's formation. First, the violence was not the Hindu-minority confrontation the independence campaign's rhetoric had primed everyone to fear — a reminder, taken to heart by the founding elite, that communal risk in Mauritius is multipolar and cannot be managed by attending to a single cleavage. Second, the state had been weeks away from inheriting sole responsibility for order, and visibly could not have coped. The lesson Ramgoolam's generation drew was that the new state's survival required buying communal peace institutionally, because it could not enforce it coercively — the insurance logic of Section 2.8, learned at the only price that teaches it.

4.2 The 1970s: Surviving the Class Challenge

The model's second stress episode — the MMM's attempt to dissolve the communal settlement into class politics, and the establishment's emergency-powers response — is analysed genealogically in Section 3.2 and historically in MU-A-01 and MU-B-01. It belongs in the stress inventory for one additional reason: it established that the model could survive a challenge to its premises, not merely to its terms, and that its survival mechanism in such cases is absorption rather than victory. The MMM was not defeated; it was metabolised, and its ideals were annexed to the settlement's own legitimating rhetoric. The episode also fixed the model's one genuinely dark precedent — that the settlement's custodians, faced with an existential challenge, suspended elections and detained opponents — which the corpus records without euphemism: the Mauritian model's continuity owes something to a period in which Mauritian democracy was interrupted in the model's defence.

4.3 February 1999: Kaya and the Crack in the Model

The gravest stress episode of the independent era began on 21 February 1999, when Joseph Réginald Topize — Kaya, the creator of seggae, the fusion of Creole séga and reggae, and the most beloved cultural figure of the Creole working class — died in police custody at Alcatraz detention centre, days after his arrest for smoking cannabis at a public rally calling for decriminalisation. The autopsy findings were immediately contested; the Creole districts of Port Louis and the lower Plaines Wilhems read the death as a police killing of their artist by a police force they experienced as another community's instrument; and four days of rioting followed — police stations attacked and burned, roads cut across the island, the security apparatus briefly losing control of significant territory, and further deaths including the singer Berger Agathe, shot by police during the unrest [TBD-VERIFY: total casualty figures — commonly given as three to five dead in the initial riots, with subsequent retaliatory inter-communal incidents at Goodlands and elsewhere adding further deaths; property damage estimates; and the findings of the judicial inquiry into Kaya's death]. A second wave of violence days later took on a directly inter-communal character between Creole and Hindu groups in the north [TBD-VERIFY: the Goodlands and surrounding incidents], carrying the island closer to 1968 than at any time in thirty years before subsiding.

The episode is the model's most important diagnostic because of what kind of failure it revealed. The 1999 riots were not a failure of the inter-communal insurance machinery: no community had moved against another's institutional position; the BLS, the conventions, the calendar were all functioning. They were an eruption of precisely the variable the machinery does not measure — cumulative socioeconomic exclusion concentrated in one community — detonated by a state-violence grievance. The model, built to prevent communities from fearing one another, had no instrument for a community that had ceased to expect anything from the state at all.

4.4 The Malaise Créole: The Exclusion the Model Papers Over

The riots had been diagnosed in advance. In 1993 the Catholic priest Roger Cerveaux gave the condition its name — le malaise créole — in an intervention that broke a double taboo: it asserted publicly that the Creole community, and especially its Afro-Mauritian working class, occupied the bottom of the island's socioeconomic order across every indicator (educational attainment, employment, housing, incarceration, the inherited geography of the post-abolition coastal settlements), and it located part of the responsibility within the Catholic Church's own historical preference for its Franco-Mauritian and gens de couleur elites [TBD-VERIFY: the precise occasion and text of Cerveaux's 1993 intervention]. The subsequent literature — church-sponsored studies, the Mauritian social-science work of the 2000s, and the Truth and Justice Commission (2009–2011), which traced the continuities from slavery through the post-abolition land question [TBD-VERIFY: the Commission's key findings and the fate of its recommendations] — assembled the structural account: a community formed by enslavement rather than indenture, which had entered freedom in 1835 without land, which the sugar economy then re-recruited at the bottom or displaced to the coasts, and whose descendants the model's category system filed under the residual label "General Population" — a classification that, unlike "Hindu" or "Muslim", names no culture, funds no ancestral language, and anchors no diaspora.

The malaise créole literature constitutes the standing internal critique of the model, and its charge is precise: the settlement insures categories, not people. Recognition parity — the holiday, the BLS seat, the subsidy — flowed to the Creole community as to every other, while the distributive outcomes that universalism was supposed to equalise remained communally skewed, because universal services compound rather than correct inherited starting positions (MU-G-05 carries the welfare-side analysis). Worse, the model's grammar of recognition-without-measurement (Section 2.7) actively obstructed the remedy: a state that refuses ethnic data cannot document, let alone target, communally concentrated disadvantage — the census silence that keeps inter-communal peace also keeps intra-society exclusion statistically invisible. The post-1999 responses (a flurry of commissions, the Trust Fund for the Social Integration of Vulnerable Groups, the eventual Truth and Justice Commission, Kreol's 2012 entry into the curriculum) were real but conspicuously avoided explicit communal targeting, on the model's standard logic — and the malaise literature's standing rejoinder is that the avoidance is the problem.

4.5 The Recurring Tensions: Language, Census, Majority, and Money

Below the episode threshold, four chronic tensions run through the model's whole history. The language question: each upward step of Kreol — orthographic standardisation, the 2012 curriculum entry [TBD-VERIFY], examination status, the parliamentary-use question of the mid-2020s [TBD-VERIFY] — has been fought as a proxy war, with Creole-identity advocates reading resistance as the malaise in linguistic form, ancestral-language constituencies reading promotion as a threat to their subsidised curricular space, and the francophone establishment defending the prestige hierarchy; the diglossia of Section 2.6 is a settlement perpetually under renegotiation at its lowest tier. The census question: the 1972 freeze (Section 2.7, MU-I-01) is attacked simultaneously from incompatible directions — by liberals as the entrenchment of obsolete categories, by malaise analysts as the suppression of the data that would prove their case, and by minority communities quietly content that no recount can document the demographic drift that might re-open the BLS allocation. The majority-anxiety question: the model's official symmetry conceals the asymmetric fact of a roughly two-thirds Indo-Mauritian, roughly half-Hindu electorate, and minority politics has therefore always carried a structural anxiety about majoritarian capture — managed, but never dissolved, by the conventions of Section 2.5, and periodically inflamed by the sociological reality that the upper public service and the police have been disproportionately Hindu-recruited [TBD-VERIFY: the public-service communal-composition studies] while private capital concentrated elsewhere. Which is the fourth tension, the money question: the Franco-Mauritian community — some 2 per cent of the population — retained through every political era a commanding position in the sugar industry and its corporate successors, the historic grands groupes of the private sector [TBD-VERIFY: concentration estimates across eras]. The model's founding economic bargain — documented from the perception side in MU-N-01 and the economic side in MU-G-01 — left this position essentially untouched: capital kept the economy, the majority took the state, and the welfare system (MU-G-05) was the transfer that reconciled the two. The bargain's quiet persistence is among the model's least advertised components, and the malaise créole sits at its unpaid intersection: the community with neither the state nor the capital.

5. The Model's Mechanics in Practice

5.1 The Slate, the Constituency, and the Ticket

The model's day-to-day engine room is party candidate selection, documented in electoral-architecture terms in MU-I-01 §7 and summarised here for its systemic function. Mauritius's twenty three-member mainland constituencies each have a known communal profile — the rural Hindu heartlands of the north and east, the Muslim strongholds of Port Louis Maritime, the Creole and gens de couleur urban belt of the Plaines Wilhems, the mixed coastal seats — and every serious party or alliance composes each three-candidate slate to mirror its constituency's composition, balancing not only the four constitutional communities but the sub-communal layers beneath them (Bhojpuri-belt Hindi-speaking Hindus, Tamil, Telugu, and Marathi minorities within the Hindu category, and the caste distinctions — notably the Vaish plurality — that Mauritian electoral folklore treats as decisive in the rural seats [TBD-VERIFY: the conventions of caste arithmetic in ticket composition, asserted ubiquitously in Mauritian political commentary but nowhere officially documented]). The practice means that before a single vote is cast, every major alliance has already performed the model's central operation: the conversion of communal anxiety into shared tickets, such that a voter of any community in any constituency is normally offered a co-communal candidate by every serious bloc. Communal appeals consequently operate within the slates rather than between monocommunal parties — which is precisely the party-system structure whose absence destroyed Guyana and Fiji (Section 6).

5.2 Cabinet, State, and the Distribution Conventions

The same arithmetic ascends. Cabinets are composed with an unwritten communal distribution — every government since 1968 has included Muslim, Creole/General Population, and (nearly always) Sino-Mauritian ministers alongside its Hindu core, with certain portfolios carrying conventional communal associations [TBD-VERIFY: the portfolio conventions — Mauritian commentary regularly asserts patterns (e.g., the historical association of particular ministries with particular communities) that require systematic tabulation against the actual cabinet lists 1968–2026]. The presidency and the ceremonial summit follow the balancing logic of Section 2.5. Below the political level, the public service operates under a constitutional non-discrimination regime and a Public Service Commission, with no formal quota — but under standing communal scrutiny, in which appointments, promotions, and the communal composition of particular services (the police above all) are tracked by the press and the communal organisations as a running audit [TBD-VERIFY: the Equal Opportunities Commission's and earlier bodies' findings on public-service communal balance]. The result is a characteristic Mauritian hybrid: a formally meritocratic state operating inside an informally communal accountability system, in which the question "is the balance being kept?" is never official and never absent. The friction between the two registers is permanent — every government faces simultaneous accusations of communal favouritism and meritocratic failure, frequently over the same appointments — and the friction is itself functional: it keeps every community's vigilance engaged through institutional channels (parliamentary questions, press campaigns, Equal Opportunities complaints) rather than extra-institutional ones.

5.3 The Diasporas and the External Dimension

The model has an external circuitry. Each community maintains a homeland relationship that doubles as a domestic political resource: the Indo-Mauritian majority's relationship with India — temple networks, the Bhojpuri revival, state-to-state instruments from the IGCIC to the Modi-era visits (MU-F-02 carries the full account) — operates simultaneously as foreign policy and as communal-cultural maintenance; the Sino-Mauritian community anchors the China relationship; the Franco-Mauritian and francophone networks anchor the France/La Francophonie tie; Creole identity politics has drawn on the African Union connection and the créolité currents of the wider Indian Ocean and Caribbean. The external patrons reciprocally treat their communities as constituencies — Indian high commissioners in Port Louis have historically operated with unusual domestic reach [TBD-VERIFY: the conventions of Indian diplomatic engagement with Mauritian communal politics] — which gives the model a managed-risk problem documented in MU-F-01: communal balance at home requires patron balance abroad, and Mauritian foreign policy's studied multi-alignment is in part the model's external face.

5.4 The Stabilising Arithmetic: Every Community a Minority

The mechanics rest, finally, on a demographic fact that the model did not create but has assiduously preserved in political form: no community can govern alone. The Hindu category's roughly 48–52 per cent [TBD-VERIFY: the 1972 frozen figures and current estimates] dissolves on inspection into Bhojpuri-belt and southern blocs, caste layers, and the Tamil, Telugu, and Marathi communities whose distinctness Mauritian politics formally honours; the "General Population" spans Franco-Mauritian magnates and Roche Bois fishermen who share a census category and nothing else; the Muslim community divides along its own internal lines. Every actual governing majority has therefore been a coalition of fragments — which means every community, including the largest, experiences politics as a minority experiences it: needing allies, fearing exclusion, valuing insurance. The "every community a minority" arithmetic is the model's deepest stabiliser because it makes the insurance components self-enforcing: the majority that could in principle abolish the BLS or the conventions is itself a coalition whose members might need them. The arithmetic also marks the model's limit condition, to which Section 7 returns: it stabilises only so long as the fragments remain politically salient. A genuinely post-communal electorate would dissolve the arithmetic — and with it both the model's costs and its self-enforcement.

6. The Comparative Question: Why Mauritius and Not Guyana, Fiji, Sri Lanka

6.1 The Standard Comparison Set

The comparative question is the model's external examination, and the examiners chose the set: the sugar-colony plural societies created by the same nineteenth-century machinery of abolition and indenture. Guyana, Trinidad and Tobago, Fiji, and Mauritius share an almost laboratory-grade common origin — plantation colonies whose African-descended populations were joined, after abolition, by Indian indentured labour in numbers that produced bi- or multi-polar demographies, all inheriting Westminster constitutions at decolonisation. Sri Lanka, though not a pure member of the set, supplies the catastrophic outer bound of an Indian Ocean plural society's failure. The divergence is stark. Guyana's party system fossilised by the early 1960s into an Indo-Guyanese/Afro-Guyanese binary (PPP/PNC) that has organised every election since, passing through rigged elections, authoritarian rule, recurrent electoral violence, and the 2020 count crisis. Fiji's indigenous/Indo-Fijian cleavage produced the coups of 1987 and 2000 — each triggered directly by Indo-Fijian-supported governments taking office — mass Indo-Fijian emigration that reversed the demographic balance, and two decades of military-tutelary politics. Sri Lanka's majoritarian drift — the 1956 language act, the standardisation policies, the 1983 pogrom — produced a twenty-six-year civil war. Trinidad, the set's second-best outcome, kept electoral democracy but with a durably ethnicised two-party structure. Mauritius alone produced neither an ethnic party binary, nor a coup, nor a war, nor an emigration-driven demographic reversal — and the comparative literature has spent four decades asking why [TBD-VERIFY: the key works — Deborah Brautigam's institutional analyses of Mauritius; Barbara Wake Carroll and Terrance Carroll's articles on Mauritian state-society accommodation and the "not so plural society"; Henry Srebrnik's "Full of Sound and Fury: Three Decades of Parliamentary Politics in Mauritius" and related pieces; the comparative chapters treating Mauritius–Guyana–Fiji; precise titles, journals, and dates require verification].

6.2 The Explanations Inventory

The literature's explanations sort into six families, none sufficient alone. (1) No indigenous claim. Mauritius had no pre-colonial population: every community descends from arrival (colonisation, enslavement, indenture, migration), so no group can deploy the trump of prior ownership that structured Fiji absolutely (indigenous land title and paramountcy doctrine) and Sri Lanka mythopoetically. Mauritian communal claims are all claims of equal standing among arrivals, which are negotiable, rather than claims of original right, which are not. This is the explanation with the strongest claim to being a necessary condition — and it is also pure historical luck, unavailable to any policy-maker. (2) The fragmented demography. The "every community a minority" arithmetic of Section 5.4 contrasts with Guyana's and Fiji's near-binary structures, where a single cleavage could organise the whole party system and every election approximated an ethnic census. Multipolarity forced multi-communal coalitions; bipolarity rewarded ethnic consolidation. (3) The electoral-institutional choices. The three-member constituency, the cross-communal slate logic it generates (Section 5.1), and the BLS insurance floor (MU-I-01) gave minorities institutional stakes that Guyana's winner-take-all list system and Fiji's communal-roll architecture — which separated the electorates and thereby hard-coded the cleavage — never provided. (4) The growth lubricant. The Mauritian elite bargain distributed a growing product (MU-G-01): the EPZ boom arrived within fifteen years of independence and full employment within twenty, so the settlement's premiums were paid out of increments rather than out of any community's existing share. Guyana's bargain, by contrast, collapsed amid economic failure and emigration; the counterfactual discipline here is genuine — the Mauritian model itself nearly failed in the stagnant, riot-prone early 1970s, and its consolidation tracks the boom. (5) The elite-bargain culture. The repeated practice of co-opting the defeated (Section 3.2), the alliance-rotation system in which every major party has governed with every other [TBD-VERIFY: the full alliance matrix 1968–2024 — PTr-PMSD, PTr-MMM, MMM-MSM, MSM-PTr, and the 2024 PTr-MMM-ND alliance substantially complete it], and the smallness that makes the political class a single acquainted milieu, produced what the Carrolls characterised as an accommodationist state-society style [TBD-VERIFY]. (6) The smallness and density itself: an island of 1.2 million with no hinterland to retreat to, no territorial base for separatism, and total mutual exposure — every community lives within an hour of every other — raises the cost of conflict and lowers the cost of monitoring the bargain.

The honest synthesis, which the better literature states explicitly, is conjunctural: Mauritius is the case where six favourable factors coincided, several of them (no indigenous claim, multipolar demography, smallness) unavailable by choice. Brautigam's institutional account, the Carrolls' accommodation account, and Srebrnik's parliamentary-politics account each promote a different subset, but none claims a single master variable [TBD-VERIFY: each work's specific argument], and the econometric "miracle" literature's parallel finding (MU-N-01 §2) — that Mauritius is good evidence for everyone's theory — recurs here in political form.

6.3 The Export Claims and Their Limits

The model has been marketed — by Mauritian diplomacy, by development institutions, and by the "lessons from Mauritius" genre — as a transferable template for plural societies, and the comparative analysis above states the limits of the claim precisely. What is arguably transferable: the design principles — minority insurance over majoritarian purity, universalism over targeting in communally legible domains, recognition parity (calendar, subsidy, language domains) as cheap and potent, co-optation over exclusion of defeated blocs. What is not transferable: the conditions — the absence of indigenous claims, the multipolar fragmentation, the growth windfall, the scale. A Fiji or a Sri Lanka adopting the BLS in 1970 would have insured the wrong risk; the instrument presupposes the multipolarity it protects. The model's own custodians have generally been more modest than its marketers, and the corpus's judgment follows them: Mauritius demonstrates that a plural society can convert communal anxiety into routinised politics, and demonstrates equally that the conversion ran on conditions no policy can replicate. The demonstration of possibility is itself valuable — against the literature that treats deep pluralism as fate — but it is a demonstration, not a kit.

7. The Model's 2030s Test

7.1 The Eroding Categories

The model's deepest assumption — that the four constitutional communities describe real, stable, politically salient groups — is eroding from below, on three measurable fronts. Secularisation: religious observance and institutional religious authority are declining across communities, most visibly among the under-35s, weakening the communal organisations (temple federations, the mosques' committees, the Church) that historically delivered both social discipline and bloc political signals [TBD-VERIFY: survey data on religiosity trends — Afrobarometer rounds and local survey work; the census religion question's trend line]. Intermarriage: cross-communal marriage, statistically marginal in the founding generation, has risen steadily [TBD-VERIFY: intermarriage rates — civil-status data is not published by community, precisely because of the census silence, so the evidence is survey-based and partial], producing a growing cohort whom the four-category system cannot classify without violence to fact — the constituency from which the Rezistans ek Alternativ refusal-to-declare litigation drew its force (MU-I-01 §4). Self-identification drift: successive survey waves find rising shares, again concentrated among the young, identifying as "Mauritian" before any communal label [TBD-VERIFY: the survey series and figures]. None of this abolishes communal identity — festivals thrive, ancestral-language enrolment persists, and communal voting cues remain detectable in constituency returns — but the direction is consistent: the categories are becoming cultural repertoires people draw on rather than containers people live inside, and the model's machinery was built for containers.

7.2 The 2024 Election's Two Readings

The 10 November 2024 alternation (MU-C-01, MU-E-01) is the natural experiment on which both futures claim evidence. The model-vindication reading: the settlement performed flawlessly under maximum stress — an incumbent swept out 60–2, concession within hours, no communal incident, the premiership passing between Hindu-led alliances under the convention, the slates of both blocs communally composed in the classical manner, and the BLS allocation proceeding routinely. On this reading 2024 belongs with 1982 and 1995 in the model's trophy cabinet: the system's capacity to absorb total political rupture without communal rupture is precisely its design success. The category-rejection reading: the content of 2024 points the other way. The campaign was fought and won on corruption, the Missié Moustass surveillance revelations, and cost of living — issues with no communal valence; the youth vote broke on generational rather than communal lines [TBD-VERIFY: constituency-level and survey evidence on the 2024 youth vote]; the 60–2 sweep itself demonstrates an electorate moving as a national bloc rather than as communal segments, since no communally segmented electorate produces near-unanimous swings; and the winning alliance's platform carried electoral-reform commitments touching the declaration requirement [TBD-VERIFY: the Alliance du Changement's 2024 manifesto commitments on electoral reform and the BLS]. On this reading the model's machinery executed perfectly a politics that no longer needs it — the scaffolding standing after the building has set.

Both readings are defensible because they answer different questions: the first asks whether the machinery still works (yes), the second whether the risk it insures still exists at its founding magnitude (unclear). The model's custodians have historically refused to run the experiment of removal — Section 2.8's standing answer — and 2024 changed the politics of that refusal without resolving it.

7.3 The Cross-Cutting Inequality Dimension

The model's second 2030s problem cuts across the categories rather than between them. Mauritian inequality — the Gini drift of the 2010s, the relative-poverty concentrations, the wage stagnation beneath the high-income headline [TBD-VERIFY: the Statistics Mauritius household-budget-survey series and Gini trend] — increasingly sorts the population by class position in the post-sugar services economy (MU-G-03, MU-O-03) rather than by community alone, even as its worst concentrations remain communally legible in the Creole coastal and suburban poverty the malaise créole literature mapped (Section 4.4). This is the configuration the model handles worst, as 1999 demonstrated: distributive grievance that is partly communal and partly not, which the universalist machinery cannot target and the communal machinery cannot represent. The MMM's 1970s wager — that class would eventually trump community — may be vindicated two generations late, in a form no one wagered on: not a unified working-class movement but a diffuse anti-system disaffection (the 2024 sweep's energy, the emigration of the young [TBD-VERIFY: emigration rates among under-35 graduates], the drug crisis documented by the Lam Shang Leen Commission) that engages neither the communal channels nor, durably, the party-alliance system itself.

7.4 Adaptation or Ossification: The Two Scenarios

The forward question is the model's capacity for managed self-retirement. The post-communal scenario: the settlement's custodians progressively decommission the compulsory machinery while retaining the insurance principle — making communal declaration optional permanently (generalising the 2014 temporary fix, MU-I-01 §4), replacing the BLS's frozen-census arithmetic with a party-list proportional top-up that protects minority presence without classifying anyone [TBD-VERIFY: the reform-commission proposals along these lines — the Sachs Commission (2001), the Carcassonne and Sithanen papers, and successors], letting the conventions soften into mere prudence, and keeping the cheap components (calendar, subsidies, language domains) that recognise without compelling. On this path the model ends as successful scaffolding should: removed after the structure cures. The gerontocratic-communal scenario: the machinery persists because its operators' careers are built inside it — the party gatekeepers whose power is the slate arithmetic, the communal organisations whose subsidies and standing depend on the categories, the dynastic leaderships (every prime minister since 1968 from two families plus Bérenger, the 2024 government led by a man first elected to the office in 1995) for whom the communal-alliance game is the only politics they have practised — while the electorate's exit accelerates: emigration, abstention [TBD-VERIFY: turnout trend across 2014–2024], and anti-system volatility of the kind the 60–2 sweeps of 2014's reverse and 2024 already register. On this path the model does not fail as 1968 feared; it is simply abandoned, surviving as a ritual the political class performs for itself.

The constitutional-reform linkage will likely decide which path is taken, and MU-I-01 §4 documents its mechanics: the BLS question — kept alive by the UN Human Rights Committee's 2012 ruling that the compulsory declaration violates the ICCPR, patched but not resolved by the 2014 temporary provisions, and renewed by the post-2024 government's reform commitments [TBD-VERIFY: the status of electoral-reform legislation in the 2025–2026 parliament] — is the single point where the unwritten settlement meets a written instrument that can actually be amended. Every previous reform attempt died of the model's own veto structure: the parties that must vote the change are the parties the current arithmetic elects. Whether the 2024 supermajority — the rare parliament with the votes to amend and a mandate read by many as post-communal — spends that capital on the settlement's renovation is, at the 2026 vantage, the model's live test.

8. Conclusion: The Settlement That Was Never Signed

Three findings organise this document's analysis of the Mauritian model as an idea.

First, the model is an insurance system, not a harmony, and its components only cohere when read through the insurance logic. Taken separately, the best-loser system, the universalist welfare state, the religious subsidies, the office conventions, the diglossia, the holiday calendar, and the census silence look like an inventory of anomalies — each violating some principle (liberal, fiscal, secular, meritocratic, statistical) that a well-ordered state is supposed to honour. Read together, they are a single design: every component removes one category of communal catastrophe from the space of political possibility, and every component's cost is the premium on that removal. The design was never drawn; it accreted out of the fear of 1968, the absorption of the 1970s challenge, and six decades of coalition practice — and its unwritten character is not a deficiency but its operating mode, since an equilibrium of this kind survives by bending where a codified settlement would have to break. The corpus's framing of "ethnic peace as a public good" is therefore not a metaphor: Mauritius taxes itself, in efficiency and in principle, to supply a good whose value is only visible in the comparative counterfactual.

Second, the model's record is genuinely exceptional and genuinely incomplete, and the two facts are causally linked. The comparative examination of Section 6 returns the strongest available verdict: among the sugar-colony plural societies, Mauritius alone converted communal anxiety into routinised politics, and no rival explanation dislodges the settlement's contribution to that outcome. But the settlement's design choices produced its characteristic blindness: a machinery built to insure communities as categories has no instrument for exclusion within the society — and the malaise créole, named in 1993 and detonated in 1999, is the standing demonstration. The model keeps inter-communal peace partly by refusing the measurements (the ethnic census, the targeted programme) that intra-society justice would require; its greatest crack came not from the conflict it insures against but from the inequality its insurance conceals. Any honest statement of the model must carry both clauses, and the official nation arc-en-ciel framing, which carries only the first, is best understood as one of the model's components rather than as its description.

Third, the model's future turns on whether an insurance policy can be gracefully retired by its beneficiaries — a problem its founders never had to solve. The categories are eroding from below; the 2024 election performed the machinery flawlessly while suggesting the electorate may have outgrown it; the inequality dimension cuts across the lines the machinery reads; and the reform lever (the BLS question, MU-I-01 §4) sits before the rare parliament with the arithmetic to pull it. The post-communal and gerontocratic scenarios of Section 7.4 are both fully consistent with the record, and the corpus declines to predict between them. What the record does support is a narrower judgment: the model has survived every previous test by adapting informally while changing nothing formally — and the 2030s test is the first in which that very strategy is what is being tested, because the generational challenge is directed at the formal categories themselves. A settlement that was never signed cannot be formally renegotiated; it can only be re-practised or abandoned. Which of those the post-1999 generation chooses — and whether the political class gives it the choice — is the open question on which this document, like the model it analyses, must end.

Primary Sources Consulted:

Constitution of Mauritius (1968, as amended), First Schedule (the best-loser system and the four-community classification); the Constitution (Declaration of Community) (Temporary Provisions) Act 2014.
J. E. Meade et al., The Economic and Social Structure of Mauritius (Report to the Governor, 1961; Methuen edition); Richard Titmuss and Brian Abel-Smith, Social Policies and Population Growth in Mauritius (1961, reissued 1968).
Arend Lijphart, Democracy in Plural Societies (Yale, 1977) and later comparative work on power-sharing and electoral systems [TBD-VERIFY: the specific Mauritius references].
Thomas Hylland Eriksen, Us and Them in Modern Societies: Ethnicity and Nationalism in Mauritius, Trinidad and Beyond (Scandinavian University Press, 1992); Common Denominators: Ethnicity, Nation-Building and Compromise in Mauritius (Berg, 1998).
Deborah Brautigam, "The 'Mauritius Miracle': Democracy, Institutions and Economic Policy", in Richard Joseph (ed.), State, Conflict and Democracy in Africa (Lynne Rienner, 1999) [TBD-VERIFY: related Brautigam pieces on institutions and coalition-building in Mauritius].
Barbara Wake Carroll and Terrance Carroll, articles on Mauritian state-society accommodation and democratic institutionalisation (Journal of Modern African Studies, Journal of Commonwealth and Comparative Politics, 1990s–2000s) [TBD-VERIFY: precise titles and dates].
Henry Srebrnik, "'Full of Sound and Fury': Three Decades of Parliamentary Politics in Mauritius" (Journal of Southern African Studies, 2002) and related pieces on ethnicity and electoral politics in Mauritius [TBD-VERIFY].
Arvind Subramanian and Devesh Roy, "Who Can Explain the Mauritian Miracle?", in Dani Rodrik (ed.), In Search of Prosperity (Princeton, 2003) — for the institutions-versus-rents debate's communal-settlement dimension.
Report of the Banwell Commission on the electoral system (1966); the Sachs Commission report on constitutional and electoral reform (2001); the Carcassonne report (2011) and the Sithanen consultancy paper on electoral reform (2012) [TBD-VERIFY: formal titles and dates].
UN Human Rights Committee, Narrain et al. v. Mauritius (Communication No. 1744/2007, Views adopted 2012) — the Rezistans ek Alternativ compulsory-declaration case.
Truth and Justice Commission (Mauritius), Report (4 vols., Port Louis, 2011) — slavery, indenture, and the land question [TBD-VERIFY: key findings cited].
Commission of Inquiry on Drug Trafficking (Lam Shang Leen Commission), Report (Port Louis, 2018).
Roger Cerveaux's 1993 malaise créole intervention and the subsequent literature [TBD-VERIFY: the originating text/occasion], including church-sponsored and academic studies of Creole exclusion (e.g., the work collected around the Journal of Mauritian Studies and the post-1999 commissions).
Judicial and inquiry records on the death of Joseph Réginald Topize (Kaya) and the February 1999 riots [TBD-VERIFY: the inquiry's findings and casualty record]; contemporary press coverage (L'Express, Le Mauricien, Week-End, February–March 1999).
Contemporary records of the January 1968 Port Louis riots: colonial-office and Mauritian government records, and the historical literature on the independence transition [TBD-VERIFY: casualty figures across sources].
Statistics Mauritius: census volumes 1962 and 1972 (the last ethnic enumerations), Housing and Population Census reports 1983–2022 (religion and language series), Household Budget Survey and inequality series [TBD-VERIFY: Gini trend figures].
Government of Mauritius, annual Budget Estimates — the subventions to religious bodies and the statutory cultural centres and speaking unions [TBD-VERIFY: the subsidy mechanics, statutory list, and quantum].
Electoral Supervisory Commission and Office of the Electoral Commissioner, election reports 1967–2024 (the BLS allocations and candidate community declarations).
Equal Opportunities Commission (Mauritius), annual reports and inquiries touching communal balance in employment and the public service [TBD-VERIFY: relevant findings].
Afrobarometer survey rounds for Mauritius (2012–2024) — national-versus-communal identification, religiosity, and institutional trust [TBD-VERIFY: the specific items and trend figures].
Catherine Boudet, Sheila Bunwaree, Roukaya Kasenally, and the Mauritian academic commentariat's work on communalisme, ethnic politics, and democratic quality (Afrique contemporaine, Politique africaine, Journal of African Elections, EISA reports) [TBD-VERIFY: representative titles].
Comparative literature on the sugar-colony plural societies: the Guyana (PPP/PNC), Fiji (1987 and 2000 coups), Trinidad, and Sri Lanka cases as treated in the comparative-politics literature on ethnic conflict and power-sharing [TBD-VERIFY: the principal comparative chapters and articles pairing Mauritius with Guyana/Fiji].
Alliance du Changement, 2024 general-election manifesto and post-election Government Programme 2025–2029 — the electoral-reform and constitutional-reform commitments [TBD-VERIFY: precise commitments on the BLS and declaration requirement].
L'Express (Port Louis) and Le Mauricien archives, 1968–2026 — the continuous press record of slate composition, cabinet balance, communal controversy, and the reform debates.

Related Documents:

MU-A-01: Independence and the Founding Era (1968–1982) — the 1968 riots, the Ramgoolam coalitions, and the emergency period in which the settlement was forged
MU-B-01: The Anerood Jugnauth Era (1982–1995; 2000–2003) — the boom that funded the bargain and the alliance-rotation practice that institutionalised it
MU-C-01: Democratic Alternation and the 2024 Elections — The Ramgoolam Restoration — the alternation whose two readings frame the model's 2030s test (Section 7.2)
MU-G-05: The Mauritian Welfare State — Free Education, Free Health, and the Universal Pension (1948–2026) — the full account of universalism as peace technology (Section 2.3)
MU-H-PM-01: Sir Seewoosagur Ramgoolam — A Biography — the consociational instinct's author (Section 3.2)
MU-H-PM-03: Paul Bérenger — A Biography — the model's challenger, exception, and absorbed critic
MU-I-01: The Mauritian Electoral Architecture — Best-Loser System and Constituency Design (1968–2026) — the constitutionalised machinery summarised in Sections 2.2, 5.1, and 7.4
MU-I-02: The Mauritian Judiciary and the Privy Council (1968–2026) — the legal forum of the declaration litigation and the settlement's rule-of-law substrate
MU-N-01: Mauritius in International Perceptions — The African Exception and the Offshore Question (1968–2026) — the external career of the model's success narrative and its blind spots
MU-O-03: Mauritius Megatrends — The 2030s Questions — the demographic, economic, and generational trend lines on which Section 7 draws
ZA-M-01: Nonracialism, Rainbow Nation, and the Contested Ideas of South African Nationhood
RW-M-01: Ndi Umunyarwanda and the Post-Ethnic Citizenship Project
NG-M-01: Zoning, Federal Character, and the Ideas of Nigerian Balance
MU-K-03: The 1983 MMM Split and the Jugnauth Realignment Decision