Two Concepts of Belonging: Legal Citizenship and Social Recognition

JEL Classification: D72, K37, Z13

Keywords: constitutional theory, membership, citizenship, social recognition, boundary making, civic ideals, Indigenous sovereignty, comparative constitutional analysis

Abstract

Contemporary American political conflict often reflects an implicit dispute about who counts as a member of the polity. This paper develops a framework for analysing this dispute as a structural source of constitutional stress. It distinguishes two ideal-type membership criteria — legal-constitutional citizenship and social recognition — and characterises, through an illustrative analytical apparatus, three qualitative implications of their coexistence: the existence of a polarisation threshold above which ordinary policy bargaining gives way to a contest over boundary rules; the possibility of self-sustaining boundary politics through elite activation; and the persistence of perceived legitimacy losses under full legal compliance. A scope test of the framework is conducted through comparative analysis of Indigenous membership in the United States, Canada and Australia, showing that the binary structure of legal citizenship requires disaggregation wherever layered sovereignty is present, and that constitutional entrenchment of layered sovereignty is a design dimension orthogonal to the two-criterion axis. The framework bridges a disciplinary gap between constitutional political economy, which has developed a sophisticated theory of rules but an underdeveloped theory of membership, and the sociology of ethnic boundaries, which has developed a sophisticated theory of boundary dynamics but limited engagement with formal constitutional analysis. The implications reframe recurring debates about immigration, citizenship and “identity politics” as conflicts between rival membership regimes, with consequences for constitutional coordination and perceived legitimacy.

1. Introduction

A constitution presupposes a people: some set of persons jointly bound by a system of rules and accepting those rules, to some degree, as legitimately authoritative. In constitutional political economy this presupposition is not a sociological footnote; it is a central variable. If political conflict includes conflict over who the members are, then standard accounts of constitutional agreement, political exchange and rule-following become unstable. Yet much of the constitutional political economy tradition — from Buchanan and Tullock (1962) through Brennan and Buchanan (1985), and continuing in the work of Voigt (2011, 2020), Congleton (2011) and Hardin (1999) — has treated the composition of the membership as largely fixed. This paper argues that the instability of membership criteria is itself a primary source of constitutional stress, and offers a framework for analysing it that draws together resources from constitutional political economy and the sociology of ethnic boundaries.

In American public discourse, the label “American” functions as a boundary judgement: who belongs, who may rule, whose interests count. Contemporary conflict reflects not merely disagreement over immigration levels or cultural change, but disagreement over the criterion that determines membership in the polity. Two ideal types are distinguished: (1) legal-constitutional membership, defined by citizenship and the rights and duties the constitutional order attaches to that status; and (2) social-recognition membership, defined by culturally mediated recognition of belonging. These criteria are salient, systematically different and, together, help explain recurring patterns in political disagreement that otherwise appear as incoherent “identity politics.” The paper takes the normative question of which criterion should be authoritative as analytically secondary to the positive question of how heterogeneous criteria interact. Where normative framing is relevant, we follow what may be called the Buchananite presumption that procedurally legitimate rules deserve normative priority (Buchanan 1987; Brennan and Buchanan 1985); but this presumption is not load-bearing for the positive results developed below.

The framework engages two bodies of work that have remained largely unconnected. The first is Rogers Smith’s Civic Ideals (1997), the most direct predecessor of the analysis developed here. Smith demonstrates that American citizenship has never been governed by a single ideology; three traditions have coexisted in uneasy tension: liberalism (rights-bearing individuals equal before the law), republicanism (civic virtue and participatory self-governance) and ascriptive Americanism (hierarchies of race, ethnicity and religion used to define who “truly” belongs). The present paper’s distinction between Definition L (legal) and Definition C (social recognition) maps onto — but is not identical to — Smith’s liberal versus ascriptive traditions. Definition C is broader: it includes cultural and linguistic proximity that need not be racialised, and it accommodates the implicit, heuristic forms of in-group judgement that Brubaker (1992) and Joppke (2010) have analysed in comparative perspective. The framework treats the coexistence of multiple membership traditions as a structural feature of constitutional politics, parameterised here as α, the effective political weight of social-recognition-oriented voters. This weight includes population share, but may also reflect turnout, issue salience, agenda-setting power and elite amplification.

The second body of work is Andreas Wimmer’s framework of ethnic boundary making (2008, 2013). Wimmer argues that ethnic boundaries are not primordially fixed but are strategically activated by political entrepreneurs when the expected returns from activation exceed the costs — activation is more likely when political competition intensifies, when institutions distribute benefits along ethnic lines and when elites can credibly signal that boundary politics serves in-group interests. This has a direct implication for our framework: α is not merely demographic but strategic, because effective political weight can be altered by mobilisation, salience and elite amplification. Even a polity in which most voters apply legal criteria can be pushed toward boundary conflict through activation strategies. Wimmer, Cederman and Min (2009) provide cross-national evidence that ethnic boundary activation is associated with armed conflict, underscoring the constitutional stakes.

The constitutional-political-economy stakes are clear. If citizenship is a focal point that coordinates expectations (Schelling 1960; Brennan and Buchanan 1985; Hardin 1999), then systematic divergence between legal membership and social recognition generates a legitimacy gap: outcomes that are lawful yet experienced by some citizens as politically illegitimate, and therefore less capable of coordinating compliance and cooperation. In Hart’s (1961) terms, the internal acceptance of the rules may erode — not because the rules are unknown, but because the relevant population disagrees about who the rules are for. The illustrative framework developed in Section 4 captures this mechanism and characterises three propositions: a polarisation threshold α^∗ above which politics collapses from ordinary policy bargaining to a contest over boundary rules; a condition under which elite activation can shift the polity from a legal-focal-point equilibrium into a boundary-contest regime; and an asymmetry that prevents legal compliance from restoring perceived legitimacy once that shift has occurred.

The paper makes three focused contributions. First, it formulates the coexistence of multiple membership criteria as a problem of constitutional coordination, with heterogeneous citizens whose evaluation of institutions depends on different criteria of political legitimacy. The aim is not to provide a general theory of identity politics or populist mobilisation, but to isolate one specific mechanism — membership-criterion divergence — through which conflict over rules becomes conflict over the persons to whom rules apply. Second, the framework characterises a qualitative discontinuity: the shift from ordinary policy bargaining to boundary-rule contestation is discrete at α^∗ rather than continuous in the effective political weight of social-recognition-oriented citizens. This claim is presented as an analytical implication of the framework rather than as an empirical regularity, and the framework is best read as making explicit a structural feature that other accounts have approached descriptively. Third, the framework is subjected to an explicit scope test through comparative analysis of Indigenous membership in three settler-colonial constitutional orders. The scope test is not an illustrative excursus: it shows where the binary criterion fails (wherever layered sovereignty obtains), and it reveals a constitutional design dimension — degree of entrenchment of layered sovereignty — that is orthogonal to the two-criterion axis. By being explicit about what the framework does not cover, the scope test also clarifies what it does cover and on what grounds.

The illustrative status of the formal apparatus deserves emphasis at the outset. The model in Section 4 is intended to clarify the mechanism described in Sections 2 and 3, not to generate sharply identified empirical predictions beyond the qualitative propositions stated above. Its primary purpose is communicative: to make precise what is otherwise an informal claim about how heterogeneous membership criteria interact under majoritarian institutions. Readers concerned with empirical identification will find suggestions for testable extensions in Section 5.

Recent work in this journal has examined neighbouring questions from complementary disciplinary angles: civic and ethnic conceptions of nationalism, discontinuous polarisation dynamics, institutional buffers against populist mobilisation and the erosion of institutional trust (Dar 2022; Ehret, Witt and Banisch 2024; Tabatabaei 2024; Krange, Kaltenborn and Hultman 2021). The present paper contributes to these conversations by specifying the constitutional mechanism that links boundary-making dynamics to the focal-point function of legal citizenship. Its claim is not simply that nationalism, polarisation or populism matter for constitutional order, but that constitutional coordination becomes fragile when citizens disagree about the criterion by which persons count as members of the polity.

The remainder of the paper proceeds as follows. Section 2 distinguishes the two membership criteria, characterises their interaction and derives a structural implication for constitutional stability. Section 3 develops a comparative constitutional analysis of Indigenous membership in the United States, Canada and Australia as a third category irreducible to either criterion. Section 4 presents the illustrative formal framework. Section 5 discusses institutional implications, comparative extensions and limitations. Section 6 concludes.

2. Two Membership Criteria and Their Interaction

Legal systems do not merely regulate behaviour; they also classify persons. Under the legal-constitutional criterion (Definition L), an American is a person who holds U.S. citizenship acquired by birth on U.S. soil under the Fourteenth Amendment’s jus soli provision or by naturalisation through legal procedure. The defining features are: (i) binary and administrable — one either is or is not a citizen; (ii) publicly verifiable — status is recorded and legally cognisable; and (iii) procedurally legitimate — membership is granted through recognised legal acts, not ethnic or cultural tests. This criterion aligns with the Buchananite requirement that constitutional rules bind members equally through general, non-discriminatory procedures (Buchanan 1987; Voigt 2011). Its institutional advantage is administrability and susceptibility to reform through publicly legible procedures, though its stability depends on sustained enforcement and public acceptance.

Under the social-recognition criterion (Definition C), “American” denotes a person perceived as belonging through cultural identification, shared symbols, language, norms and in-group acceptance. The defining features are: (i) continuous and contestable — belonging comes in degrees; (ii) context-dependent — recognition varies across regions, groups and time; and (iii) non-administrable — there is no authoritative procedure for determining it. Importantly, Definition C is not monolithic. It aggregates ethnic-racial identification, linguistic proximity, religious commonality and regional cultural fit, and these dimensions can diverge from one another: a Latino Catholic from Texas may score high on religious and regional recognition but low on ethnic recognition for some in-group members. Definition C does not require explicit racial animus; it can be applied implicitly through heuristics about “fit” or “shared values” (Hainmueller and Hopkins 2015; Enos 2017). This treatment of the ethnic–civic distinction as a spectrum rather than a dichotomy aligns with the recent critical reappraisal of the civic–ethnic dichotomy in nationalism studies, where Dar (2022) and the contributors he reviews argue that the two ideal types are best understood as endpoints of a continuum along which actual national self-understandings can be located. Because it is non-administrable, it is systematically vulnerable to strategic manipulation: as Wimmer (2008) shows, boundary activation depends on entrepreneurs who benefit from mobilising in-group identity, and Definition C provides the raw material for such activation precisely because it is informal and contested.

The two criteria can coincide (a native-born citizen widely recognised as belonging) or diverge in three patterns: (i) legal member, socially contested — a naturalised citizen whose belonging is disputed; (ii) socially accepted, not a legal member — a culturally proximate long-term resident treated as “one of us” in everyday interaction; and (iii) ambiguous layered status — populations governed by distinctive legal arrangements, including Indigenous peoples (Section 3). The paper’s central analytical claim is not that one criterion is metaphysically correct, but that political conflict changes character when citizens disagree over which criterion is authoritative for allocating political standing. When that disagreement is sharp, politics shifts from bargaining over ordinary policy toward contesting the boundary rules themselves.

One structural implication follows immediately from the coexistence of both criteria. Buchanan’s constitutional political economy emphasises agreement on rules that structure ordinary politics (Buchanan and Tullock 1962; Buchanan 1975; Brennan and Buchanan 1985), and recent work has extended this concern to the conditions under which constitutional rules become self-enforcing focal points (Weingast 1997; Hardin 1999; Ordeshook 1992). That agreement presupposes a relatively stable set of members bound by the constitution. If membership is contested, constitutional bargaining resembles bargaining in the absence of a settled “we,” making constitutional constraints harder to sustain as focal points. Hayek’s (1973) account of emergent orders of rules that structure expectations outside formal legislation is also relevant: the social-recognition criterion operates as exactly such an informal order, powerful precisely because it is uncodified. The conflict between a formal legal order of membership and an informal social order of recognition is therefore not incidental to constitutional politics but constitutive of it.

A further structural implication concerns the asymmetry between the two criteria in their susceptibility to legal challenge and political mobilisation. Definition L, as a formal legal status, is in principle contestable through courts, administrative procedures and legislative reform. Its exclusionary episodes — the Chinese Exclusion Act, racially discriminatory naturalisation requirements, the internment of Japanese Americans — were eventually challenged and in most cases overturned or formally repudiated, precisely because they operated within a publicly legible legal framework. Definition C, by contrast, is non-administrable and therefore largely immune to legal challenge. One cannot sue an electorate for failing to recognise a candidate as “truly American.” This asymmetry has an important implication for constitutional stability: as the political salience of Definition C increases relative to Definition L, the institutional tools available to enforce equal citizenship weaken, because those tools are designed to address formal legal violations rather than informal social judgements. The framework in Section 4 captures this asymmetry through a parameter r — the robustness of equal-citizenship enforcement — which type-C voters prefer to reduce even when it is set at its maximum value, because the mismatch events that drive their utility loss are generated by informal recognition failures that legal enforcement cannot eliminate.

The two criteria also differ in their dynamics. Definition L is relatively stable over short time horizons: citizenship, once acquired, is difficult to revoke, and naturalisation rates are governed by administrative processes that change slowly. Definition C is volatile: it is subject to rapid shifts in salience driven by political events, media framing and elite activation. This volatility is at the heart of the activation result in Section 4. It means that a polity can transition rapidly from a regime in which legal-constitutional criteria dominate political competition to one in which social-recognition criteria do, without any change in the formal legal structure of citizenship. Such transitions appear, from the outside, as sudden eruptions of “identity politics” or “cultural backlash,” but they are the predictable outcome of elite activation strategies operating on an underlying distribution of voter preferences that was always heterogeneous, awaiting the cost-benefit configuration that renders boundary-rule politics electorally attractive (Inglehart and Norris 2019; Sides, Tesler and Vavreck 2018).

3. Indigenous Peoples and Layered Sovereignty: A Comparative Analysis

This section conducts an explicit scope test of the two-criterion framework. A scope test asks where a framework breaks down, with the aim of identifying the boundary conditions within which its claims hold. The two-criterion framework treats membership as a contest between two ideal types: legal citizenship and social recognition. If that contest exhausts the relevant alternatives, the framework should apply uniformly across constitutional orders; if there exist populations whose political standing is grounded in a third source — neither legal citizenship of the dominant order nor social recognition by the dominant in-group — then the framework requires either extension or a clear specification of its boundaries. Indigenous peoples in settler-colonial constitutional orders provide precisely such a test population: their political rights derive from a prior sovereign status that predates and is partially independent of the dominant constitutional framework. The aim of this section is therefore not to add a comparative excursus but to determine whether and how the two-criterion framework needs to be qualified in the presence of layered sovereignty. The analysis proceeds through three jurisdictions — the United States, Canada and Australia — each of which has evolved a different doctrinal response to the same underlying problem of layered sovereignty (Tully 1995; Williams 1990). The variation in doctrinal response is informative: it shows which features of the two-criterion framework remain operative across constitutional orders and which require local adaptation, and it identifies a constitutional-design dimension — the degree of entrenchment of layered sovereignty — that the framework cannot capture without disaggregation. Section 3.4 makes the implications of the scope test explicit.

3.1 United States: domestic dependent nations and the trust relationship

The foundational American cases — Cherokee Nation v. Georgia (1831) and Worcester v. Georgia (1832) — established that federally recognised tribes are neither foreign nations nor simple constituents of state or federal government. Chief Justice Marshall’s opinion described tribes as “domestic dependent nations,” a phrase that has proven both indispensable and doctrinally unstable. Tribes exercise inherent sovereign authority over their members and territories, subject since Lone Wolf v. Hitchcock (1903) to a plenary power of Congress that the Supreme Court has alternately expanded and constrained. The Indian Citizenship Act of 1924 conferred U.S. citizenship on all Native Americans without extinguishing tribal membership or its associated sovereign rights. The result is a formal layering: a citizen of the Cherokee Nation holds tribal citizenship under tribal law, U.S. citizenship under federal constitutional law and, in some cases, state citizenship — three simultaneously operative membership regimes. Where federal statutes are ambiguous, the canon of construction requires courts to interpret them in favour of tribal interests; Congress, however, retains plenary power to abrogate tribal rights, and has exercised it repeatedly.

The doctrinal complexity has ongoing practical significance. The Indian Child Welfare Act of 1978 creates a distinct federal framework governing child custody proceedings involving tribal members, asserting tribal sovereign interests in ways that can override state family law. McGirt v. Oklahoma (2020) reaffirmed that the Creek Nation’s reservation had never been disestablished, with immediate consequences for criminal jurisdiction over a substantial portion of the state’s territory. More recently, Haaland v. Brackeen (2023) upheld the constitutionality of the Indian Child Welfare Act against Article I and Tenth Amendment challenges, reaffirming Congress’s authority to legislate in the domain of Indian affairs and leaving the layered structure of federal-tribal sovereignty intact. These decisions illustrate that tribal sovereignty is not a historical relic but an active legal structure that continues to generate jurisdictional conflicts. Any account of American membership that treats citizenship as binary must grapple with the fact that for millions of tribal members, the relevant membership question is not “are you a citizen?” but “which sovereign’s law governs this relationship?” — a question that the binary L(x) of the formal model cannot represent without disaggregation.

3.2 Canada: section 35 and the constitutionalisation of Aboriginal rights

Canada has developed a distinct constitutional architecture for Indigenous membership through section 35 of the Constitution Act, 1982, which recognises and affirms the “existing aboriginal and treaty rights” of the Aboriginal peoples of Canada. The doctrinal trajectory differs from the American case in three important respects. First, the recognition is constitutional rather than statutory: Aboriginal rights are entrenched in the supreme law of Canada, and federal or provincial legislation that infringes them is subject to a justification test developed in R. v. Sparrow (1990). Second, the Royal Proclamation of 1763 — a foundational document for Crown-Indigenous relations — remains legally operative, and Canadian courts have interpreted it as recognising prior Indigenous sovereignty. Third, the doctrine of Aboriginal title developed in Delgamuukw v. British Columbia (1997) and Tsilhqot’in Nation v. British Columbia (2014) recognises a sui generis form of collective land title grounded in pre-contact occupation rather than Crown grant.

The Canadian doctrinal pattern reveals a different solution to the same underlying problem. Whereas U.S. doctrine assigns plenary power to Congress and treats tribal sovereignty as a residual category preserved at federal sufferance, Canadian doctrine constitutionalises Aboriginal rights and requires the Crown to justify infringements through a structured proportionality analysis. In framework terms, the Canadian solution attempts to reduce mismatch costs by binding the institutional centre through constitutional rules rather than relying on social or political recognition. The duty to consult and accommodate, developed in Haida Nation v. British Columbia (2004), creates procedural obligations that operate prior to any specific governmental decision, embedding Indigenous claims in routine administrative practice. From the perspective of the present framework, the Canadian doctrine demonstrates that the asymmetry between Definition L and Definition C can be partially mitigated by constitutionalising the recognition of layered sovereignty itself — by moving questions of Indigenous membership out of the domain of ordinary politics and into the domain of constitutional rules.

3.3 Australia: Mabo, native title and the absence of constitutional recognition

Australia provides a third pattern, distinct from both the American and Canadian models. The foundational case is Mabo v. Queensland (No. 2) (1992), in which the High Court of Australia overturned the doctrine of terra nullius — the legal fiction that Australia had been unoccupied at the time of British settlement — and recognised the existence of native title rights surviving the assertion of British sovereignty. The Native Title Act 1993 established a statutory regime for the recognition and protection of native title, but unlike Canadian section 35, Australian Indigenous rights are not constitutionally entrenched. The Australian Constitution contains no recognition of Aboriginal and Torres Strait Islander peoples, and the 2023 referendum on constitutional recognition through an Aboriginal and Torres Strait Islander Voice to Parliament was decisively defeated.

The Australian case reveals a third structural pattern in our framework. Where U.S. doctrine constitutionalises tribal sovereignty through judicially developed federal law and Canadian doctrine constitutionalises it through textual amendment, Australian doctrine recognises Indigenous claims primarily through common-law evolution under continuing parliamentary supremacy. This makes native title vulnerable to legislative override in a way that Canadian section 35 rights are not. The defeat of the 2023 Voice referendum illustrates a feature directly predicted by our framework: when constitutional recognition of layered sovereignty is itself submitted to majoritarian vote in a polity where Definition C has significant electoral weight, the recognition proposal becomes subject to boundary-rule contestation. Voters applying Definition C may interpret constitutional recognition not as an institutional protection of a prior sovereign claim but as a redefinition of the membership criterion itself — a redefinition they have political incentives to oppose. The Australian referendum thus offers a natural illustration of the activation mechanism described in Section 4: a proposal that, in framework terms, would have shifted the institutional treatment of Indigenous membership toward a Canadian-style constitutional rule was defeated by mobilisation that activated boundary-rule reasoning.

3.4 Implications for the L–C framework

The three-jurisdiction comparison yields four implications for the framework. First, the binary structure of Definition L is inadequate wherever layered sovereignty is present. The appropriate generalisation is a vector-valued membership function L(x) = (L_federal(x), L_Indigenous(x), L_subnational(x), …), where different components are governed by different legal orders. The specific components and their relations vary across jurisdictions — the U.S. tribal regime, Canadian section 35 rights, Australian native title — but the general structural feature is the same: a person’s political standing is determined by their position within a layered set of juridical relationships, not by a single binary status.

Second, Indigenous status illustrates that the two criteria are not jointly exhaustive: a complete account must include a third category of prior sovereign membership, in which political rights derive from a legal status that predates and is partially independent of the dominant constitutional order. This third category cannot be accommodated by expanding either Definition L or Definition C; it requires structurally distinct treatment. The comparison demonstrates that this is a general constitutional category, not an artefact of American doctrinal history.

Third, when social-recognition voters apply Definition C to Indigenous peoples — asking whether they “fit” into mainstream culture — they misclassify a legal category as a cultural one, generating a legitimacy conflict in which the politically activated claim is not just normatively wrong but legally erroneous. The Australian Voice referendum illustrates the political costs of this misclassification: the institutional reform that would have most directly addressed layered sovereignty was rendered unattainable by mobilisation that treated Indigenous claims as cultural preferences subject to majoritarian validation.

Fourth, the comparison suggests a comparative-statics conjecture for the framework: the institutional vulnerability of Indigenous claims to boundary-rule contestation varies inversely with the degree of constitutional entrenchment. The Canadian section 35 model places the largest barrier between Indigenous rights and ordinary majoritarian politics; the Australian common-law model places the smallest; the U.S. model with congressional plenary power lies between them. Where the framework predicts greatest constitutional stress is precisely where Indigenous membership remains subject to majoritarian determination — a prediction consistent with the empirical pattern of recurring legitimation crises in all three jurisdictions, but with different doctrinal forms.

Table 1. Layered sovereignty and constitutional entrenchment

4. An Illustrative Framework for Membership-Criterion Conflict

This section develops a formal framework for the mechanism described in Sections 2 and 3. The framework is illustrative: its purpose is to make precise what would otherwise be an informal claim about how heterogeneous membership criteria interact under majoritarian institutions. The propositions below characterise qualitative features of the equilibrium structure — the existence of a polarisation threshold, the possibility of multiple equilibria under elite activation, the persistence of legitimacy losses under legal compliance — rather than offering point-identified empirical predictions. Readers interested in the substantive constitutional argument may skim this section; those interested in formal reconstruction will find the structure deliberately minimal, with each modelling choice motivated by the substantive claims in the preceding sections.

4.1 Sets, criteria and the structure of K(x)

Let the polity contain individuals x ∈ Ω. Legal citizenship status is L(x) ∈ {0, 1}, inducing the legal membership set M_L = {x : L(x) = 1}. Social recognition is represented by K_i(x) ∈ [0, 1], the degree to which voter i recognises x as belonging. Each voter applies a threshold τ_i, inducing a recognised set M_{C, i}(τ_i) = {x : K_i(x) ≥ τ_i}. Unlike legal citizenship, K_i is observer-dependent: different voters can disagree about both the score and the threshold.

For tractability, K_i(x) may be read as a weighted index of perceived cultural proximity across dimensions such as language, ethnicity, religion and regional fit. This is not meant to provide a complete measurement model. It simply formalises the feature that matters for the argument: social recognition is continuous, multidimensional and non-administrable. In empirical work, the components of K could be proxied through survey items on national identity, cultural proximity and civic-behavioural signals; the present paper uses the notation only to clarify the mechanism.

4.2 Policy space and voter utility

Policy p includes three instruments: (m) migration and admission intensity; (n) naturalisation rule strictness; and (r) the robustness and visibility of equal-citizenship enforcement. We assume r and S(p) are jointly constrained such that increases in r reduce S(p): strong equal-citizenship enforcement limits the institutional relevance of social-recognition signals, since a policy that fully privileges legal standing leaves less room for social-recognition criteria to govern access to offices and benefits.

Two ideal-type voter types t ∈ {L, C} are distinguished. Let ρ denote the raw population share of type-C citizens. The political contest below is written in terms of α, the effective political weight of type-C citizens after turnout, issue salience, agenda-setting and elite amplification are taken into account; the corresponding type-L weight is 1 − α. Under pure head-count majority rule with equal turnout and equal salience, α = ρ. In more realistic settings, α may exceed ρ when type-C citizens are more intensely mobilised or when elites make boundary politics unusually salient. Utility has three additively separable components: a material payoff B(p), a criterion-salience payoff and a mismatch cost D(p) capturing conflict and perceived illegitimacy when institutional boundary rules diverge from a voter’s operative criterion:

U_L(p) = B(p) + λ_L · r − γ_L · D(p) (1)

U_C(p) = B(p) + λ_C · S(p) − γ_C · D(p) (2)

Here S(p) captures the extent to which policy makes social recognition decisive in practice. D(p) is specified as the salience-weighted expected frequency of mismatch events:

D(p) = Pr(salient beneficiary ∈ M_L  M_{C, i}) + Pr(salient beneficiary ∈ M_{C, i}  M_L) (3)

That is: legal members who fail social recognition receiving salient benefits, and vice versa. The mismatch cost is intentionally symmetric across the two directions, though empirically the salience weights may differ.

4.3 Three propositions

Policy is determined in a single-dimensional contest between two regimes: a high-r, low-S legal-focal-point regime and a low-r, high-S boundary-rule regime. The regime that receives greater effective political support wins. This is not a claim that legal institutions formally count preference intensity; it is a reduced-form way to capture the political fact that turnout, salience, campaign intensity and elite amplification affect which preferences become decisive. Pure numerical majority rule is the limiting case in which α equals the electorate share and salience weights are symmetric.

Proposition 1 (Polarisation threshold). Let λ_C > λ_L > 0 and γ_L = γ_C = γ. There exists a threshold α^∗ ∈ (0, 1) in the effective political weight of type-C voters such that (i) if α < α^∗, the legal-focal-point regime remains dominant and ordinary policy bargaining continues; and (ii) if α > α^∗, the boundary-rule regime becomes politically dominant, r is pushed toward its minimum and S(p) is maximised. Above the threshold, politics collapses to a contest over the boundary rules themselves.

Sketch. Treat the decisive choice as binary between a legal-focal-point regime (high r, low S) and a boundary-rule regime (low r, high S). Effective support for the legal-focal-point regime is proportional to (1 − α)λ_L, while effective support for the boundary-rule regime is proportional to αλ_C. The boundary-rule regime becomes dominant when αλ_C > (1 − α)λ_L, yielding α^∗ = λ_L / (λ_L + λ_C). Because λ_C > λ_L, this threshold lies below 1/2. Thus type-C voters need not be a numerical majority to dominate boundary-rule politics if their effective political weight is amplified by salience, turnout or elite mobilisation. Under pure head-count majority rule with symmetric salience, the threshold collapses to the familiar 1/2 case. □

Proposition 2 (Multiplicity under elite activation). Suppose α is endogenous: α(e) = min(α₀ + βe, 1), with β > 0. Elite payoff is π(e) = θS(p^∗(α(e)))) − c(e), with c′(e) > 0 and c″(e) > 0. Assume (a) π′(0) < 0: small activation is not privately worthwhile at the margin; and (b) there exists ē such that α(ē) > α^∗ and π(ē) > 0. Then there exist two locally stable equilibria: a low-activation equilibrium at e = 0 with α₀ < α^∗ and interior policy, and a high-activation equilibrium at e^∗ > 0 with α(e^∗) > α^∗ and boundary-rule politics.

Sketch. If activation does not cross the threshold, it is costly and politically unproductive, so e = 0 is locally stable. If activation raises α above α^∗, the policy regime shifts discontinuously and increases the payoff to further activation. With convex costs, this creates a second locally stable equilibrium at positive activation. □

Proposition 3 (Legitimacy loss under legal compliance). Fix r = 1. Suppose that with positive probability a legal member in M_L  M_{C, i} enters a salient office or receives a prominent benefit. If α > α^∗, then U_C(p) < U_C(p′) for some p′ with r′ < 1: type-C voters prefer institutional changes that reduce recognition-mismatch salience to fully lawful outcomes that trigger “legal but not legitimate” reactions.

Sketch. At r = 1, a salient legal member in M_L  M_{C, i} raises D(p) for type-C voters. If the mismatch cost exceeds the gain from legal enforcement, type-C voters prefer policies that reduce the salience or institutional consequences of the mismatch. The result captures “legal but not legitimate” rhetoric as a coherent response to recognition mismatches rather than as ignorance of the law. □

4.4 Comparative statics and the calibration of α∗

Four comparative statics follow. First, increases in λ_C relative to λ_L lower α^∗ and expand the conditions under which boundary-rule politics is electorally attractive. Second, increases in β (the sensitivity of α to elite effort) reduce the activation-cost threshold required for the high-boundary equilibrium. Third, increases in γ (the divergence-cost weight) raise the cost of the high-activation equilibrium without eliminating the multiplicity of Proposition 2. Fourth, constitutional design that increases the salience of legal membership — through anti-discrimination enforcement or high-visibility naturalisation ceremonies — effectively increases λ_L relative to λ_C, raising α^∗ and reducing the range of conditions under which boundary-rule politics obtains.

Empirical orientation. Existing surveys measure raw attitudes rather than effective political weight. Pew, ANES and comparative survey instruments suggest that culturally oriented or ethnonational conceptions of membership are held by a substantial minority of voters, but the measures are not identical and should not be aggregated into a precise estimate (Pew Research Center 2021; Sides et al. 2018; Inglehart and Norris 2019). At most, they suggest a plausible range for ρ, the raw population share disposed toward social-recognition criteria. The model’s α may be higher or lower than ρ, depending on turnout, salience and elite activation. The calibration is therefore illustrative: a constituency that is not a numerical majority may still cross the effective threshold if boundary politics becomes unusually salient.

4.5 Historical cases of elite activation

Proposition 2 establishes conditions under which elite investment can shift a polity from a legal-focal-point equilibrium to boundary-rule politics. Three episodes illustrate how the framework’s parameters map onto observable phenomena. These cases are illustrative rather than causal tests.

The Know-Nothing movement of the 1850s emerged from large-scale Irish Catholic immigration and economic dislocation. Its entrepreneurs — nativist editors, fraternal lodge organisers and political brokers — framed Catholic immigrants as agents of papal authority incompatible with republican self-governance, targeting a population legally eligible for citizenship but scoring low on K along religious and ethno-cultural dimensions. In framework terms: θ was high (patronage made boundary politics lucrative), β was high (fraternal networks made activation effective), and c(e) was low (partisan press and lodge infrastructure was already in place). Rapid electoral success between 1854 and 1856 was followed by collapse when the sectional crisis over slavery made the legal boundary question more salient than the cultural one — the activation payoff fell and the high-activation regime dissolved.

Patrick Buchanan’s campaigns of 1992–1996 represent a calculated activation effort in an era when immigration had not yet become a primary electoral cleavage. Buchanan’s rhetoric reframed legal immigration as a cultural threat, instantiating Definition C as politically salient. The activation was only partially successful nationally: Buchanan came a strong second in the 1992 New Hampshire primary (winning 37 per cent against an incumbent president), then won the 1996 New Hampshire primary against Dole, but did not secure the nomination in either cycle. The campaigns nonetheless shifted the Republican primary electorate’s effective α measurably, as evidenced by the subsequent incorporation of immigration restriction into mainstream Republican platforms after 1994 — an instance of partial activation raising α₀ for subsequent entrepreneurs.

The post-2015 period in the United States and Europe provides the most data-rich illustration. High-volume social media amplification (high β) combined with low marginal costs of digital communication (low c(e)). Sides, Tesler and Vavreck (2018) document that ethnonationalism became the strongest predictor of vote switching between 2012 and 2016, displacing economic anxiety. Inglehart and Norris (2019) show that cultural backlash predicts support for populist parties across Western democracies in the same period. In framework terms, this represents successful crossing of α^∗ in pivotal electoral units — Rust Belt states, Brexit-deciding constituencies — shifting politics from interior policy bargaining to boundary-rule contestation. The durability of this shift, with immigration remaining a primary cleavage as of 2026, is consistent with Proposition 2’s prediction that once boundary politics becomes electorally decisive, continuation payoffs for elite activists remain high and reverting to the low-activation equilibrium requires coordinated withdrawal that is individually irrational.

5. Constitutional Implications, Comparative Extensions and Limitations

The three propositions jointly generate a coherent constitutional diagnosis. Proposition 1 shows that a legal-constitutional focal point is sustainable when the effective political weight of social-recognition-oriented voters is below the polarisation threshold α^∗. Proposition 2 shows that even when the baseline distribution of social-recognition-oriented citizens would not generate boundary-rule politics, elite activation can raise effective political weight and shift the polity into a self-sustaining boundary-contest regime. Proposition 3 shows that once boundary-rule politics obtains, legal compliance alone cannot restore perceived legitimacy for type-C voters, because their utility loss is tied to salient recognition mismatches rather than procedural violations — explaining why formal legal equality can coexist with persistent legitimacy deficits.

The core diagnosis is therefore not “immigration causes conflict,” but: heterogeneous membership criteria generate constitutional stress by destabilising the set of persons for whom the rules are taken as authoritative, and this stress can be strategically amplified by political entrepreneurs whose payoffs increase in boundary salience. This reframes several recurring policy debates.

Assimilation is a contested attempt to reduce D(p) by increasing overlap between legal and social membership — either by changing persons (integration), by changing social recognition (norms and narratives) or by changing legal boundaries (restrictions). The framework shows that each instrument has different consequences. Tightening naturalisation rules can reduce some mismatch events by shrinking M_L but creates new legitimacy deficits by excluding long-term residents from legal membership. Conversely, strong equal-citizenship enforcement reduces the institutional relevance of social recognition but may not reduce perceived mismatch when high-salience events activate boundary judgements regardless. Cultural integration programmes — attempts to shift the weights w^j_i in the K(x) decomposition — operate on Definition C directly, potentially reducing D(p) but at the political cost of appearing to demand cultural conformity as a condition of belonging.

The framework also has implications for constitutional design beyond immigration and citizenship policy. Proposition 2 implies that institutional measures that raise the relative cost or reduce the relative payoff of elite activation — for example, by requiring supermajority support for changes to citizenship law, by making the legal-constitutional criterion highly salient through symbolic and ceremonial means, by investing in civics education that strengthens the focal-point function of legal membership, and by transparent campaign-finance and broadcasting rules that limit asymmetric amplification of boundary-activating messages — can help sustain the low-activation equilibrium. These are positive institutional measures rather than restrictions on political speech, and they operate by changing the cost-benefit structure of activation rather than by penalising particular content. The calibration in Section 4.4 suggests that the United States currently operates near the polarisation threshold in pivotal electoral environments. Constitutional design that raises the cost of the high-activation equilibrium — rather than simply enforcing legal citizenship ex post — is therefore an important and underexplored tool for maintaining constitutional stability (cf. Brennan and Hamlin 2000; Ostrom 1990, on institutional design for self-enforcing rules).

On Indigenous membership: the analysis of Section 3 demonstrates that the divergence cost D(p) for Indigenous communities is structurally different from that for immigrants or naturalised citizens. For immigrants, mismatch arises from the gap between legal and social membership. For Indigenous peoples, mismatch arises from a conflict between two legal regimes — the dominant constitutional law and a prior sovereign law — that the binary L–C structure does not capture without disaggregation. The three-jurisdiction comparison developed in Section 3 reveals a constitutional-design dimension orthogonal to the L–C axis: the degree to which layered sovereignty is itself constitutionalised. Canadian section 35 represents the strongest entrenchment; Australian common-law native title under continuing parliamentary supremacy represents the weakest; the U.S. trust doctrine lies between them. The framework predicts greatest constitutional stress where Indigenous membership remains subject to majoritarian determination — a prediction borne out by the Australian 2023 referendum outcome.

Comparative extensions. Although the framework was developed for the American case, its mechanisms generalise. In European parliamentary systems, the analogue of Definition C is often operationalised through explicit ethnonational rather than cultural criteria, and citizenship regimes vary from the strongly civic (France) to the historically more ethnic (Germany, until the 1999 nationality law reform; cf. Brubaker 1992; Joppke 2010). The European cases suggest that the polarisation threshold α^∗ depends not only on voter preferences but on institutional features that channel boundary politics into different forms: proportional representation systems with multiple parties tend to absorb boundary-rule politics into specialised parties rather than transforming mainstream politics into boundary contestation, while two-party majoritarian systems are structurally more vulnerable to the regime shift identified in Proposition 2. This is a testable prediction: holding voter preferences constant, boundary-rule politics should be more durable in majoritarian than in proportional systems. Cross-national data on the salience of immigration as a primary electoral cleavage — available from the Manifesto Project and from European Social Survey waves — could be used to test this systematically. The Canadian case examined by Tabatabaei (2024) is instructive in this respect: he attributes the relative weakness of populist boundary politics in Canada to four institutional buffers (a non-ascriptive national narrative, consolidated multiculturalism, the eclipse of class politics, and the constraints of federal-provincial competition), which can be read in framework terms as factors that simultaneously raise α^∗ and raise the activation cost c(e), preventing the boundary-contest equilibrium from becoming self-sustaining even when domestic and international shocks generate temporary mobilisation. A parallel mechanism is documented by Krange, Kaltenborn and Hultman (2021), who find that right-wing populist anti-elitism corrodes institutional trust in agencies advocating scientific positions perceived as elite-aligned — an institutional-trust mechanism that, by framework logic, raises the salience of recognition mismatch by making official compliance with legal rules read as elite signalling rather than as legitimate authority. Together these comparative observations suggest that the constitutional design implications of the framework are not exhausted by the formal-institutional variables of Section 4: informal institutions of trust and narrative also enter the cost structure of activation, with corresponding implications for stability.

An empirical research agenda. The framework points toward a specific empirical programme. First, α^∗ is not directly observable, but its comparative statics are testable: if α^∗ = λ_L / (λ_L + λ_C), then polities or jurisdictions in which social-recognition voters exhibit higher λ_C should exhibit lower effective thresholds and therefore more frequent transitions to boundary-rule contestation. Cross-sectional variation in immigration-issue salience across U.S. states and municipalities with different demographic compositions provides a natural testing ground. Second, the Wimmer-type prediction of boundary activation is testable: if elite activation is the primary driver of α shifts, the timing of transitions to boundary-rule politics should be predicted by identifiable elite-level events — candidacies, major speeches, media campaigns — rather than by gradual demographic change. Distinguishing these mechanisms requires time-series data at high frequency, which is increasingly available from social media and news archives.

Limitations. Several limitations bound the analysis. The framework is static; a dynamic extension in which α_t evolves in response to policy outcomes, elite strategies and demographic change — with possible hysteresis effects following crossings of α^∗ — is a natural next step, particularly important for understanding whether the high-activation equilibrium identified in Proposition 2 is permanent or subject to eventual reversion. The two-type structure abstracts from within-type heterogeneity, from strategic interaction among multiple competing elite entrepreneurs and from the role of media as an independent amplifier of boundary salience with its own incentive structure. The empirical calibration of α^∗ relies on proxy measures rather than direct estimates; a priority for future empirical work is the development of survey instruments that more cleanly separate legal-constitutional from social-recognition orientations toward membership. The historical cases are illustrative rather than causal; identification of the effect of elite activation on α would require natural experiments or other credible identification strategies. Finally, the comparative extension to Indigenous membership has been developed across three jurisdictions; further work should examine other settler-colonial constitutional orders (New Zealand, post-apartheid South Africa) and post-imperial constitutional orders with substantial layered-sovereignty features.

6. Conclusion

American political conflict frequently reflects a disagreement about what constitutes membership in the polity and which membership criterion should anchor political standing. This paper distinguished two ideal-type membership criteria — legal citizenship (Definition L) and social recognition (Definition C) — and characterised, through an illustrative framework, the conditions under which their coexistence generates constitutional stress. A substantive constitutional analysis of Indigenous membership in the United States, Canada and Australia demonstrated that the two-criterion structure requires extension whenever layered sovereignty is present, and that the degree of constitutional entrenchment of layered sovereignty is itself a constitutional design dimension orthogonal to the L–C axis.

The legal criterion can coordinate expectations because it is administrable and verifiable. Proposition 1 characterised when it functions as the effective focal point of ordinary politics: when the effective political weight of social-recognition-oriented voters is below α^∗, ordinary policy trade-offs dominate. Propositions 2 and 3 showed how the legal focal point can be destabilised even without procedural violations — through elite activation that shifts effective α above threshold, and through salient recognition mismatches that generate legitimacy losses even under full legal compliance. The empirical orientation in Section 4.4 suggested that the United States may currently operate near the polarisation threshold in pivotal electoral environments, making the stability of legal-constitutional coordination an empirical as well as a theoretical question.

Three implications follow for the research agenda. First, membership conflict should be treated as an endogenous source of constitutional instability, not an exogenous shock. Constitutional design should therefore attend not only to voting rules, supermajority requirements and rights protections, but to the institutional infrastructure that sustains the legal-constitutional criterion as the operative rule for allocating political standing — robust citizenship enforcement, transparent naturalisation procedures and institutional practices, such as high-visibility naturalisation ceremonies and civics education, that make legal membership salient as a criterion of belonging. Second, the framework points toward a specific empirical agenda: cross-sectional variation in immigration-issue salience, time-series patterns of elite activation, and comparative variation in constitutional entrenchment of layered sovereignty all yield testable predictions. Third, the analysis has implications for how legitimacy claims in contemporary democratic politics should be evaluated. Proposition 3 shows that “legal but not legitimate” rhetoric is not necessarily confused: it is a coherent response by voters whose operative membership criterion differs from the legal one. This does not vindicate the rhetoric — the Buchananite presumption assigns normative priority to procedurally legitimate rules — but it does mean that dismissing such rhetoric as bigotry or false consciousness misses its structural source. If the goal is to restore perceived legitimacy, the relevant intervention is not merely legal enforcement but a shift in the salience of λ_L relative to λ_C, making the legal-constitutional criterion more central to how voters evaluate political outcomes. This is a long-run institutional project rather than a short-run legal fix.

The paper has bridged a disciplinary gap: constitutional political economy has developed a sophisticated theory of rules but has paid limited systematic attention to the membership question that those rules presuppose; the sociology of ethnicity has developed a sophisticated theory of boundary dynamics but has remained largely separate from formal constitutional analysis. Bringing these traditions into closer contact reframes recurring debates about immigration, citizenship and identity politics as conflicts between rival membership regimes. The broader implication is simple but consequential: a constitution requires not only stable rules, but a politically stable criterion for deciding who those rules are for. Where that criterion fractures, constitutional conflict becomes conflict over the boundaries of the people itself.

References

American National Election Studies (2021) ANES 2020 Time Series Study. University of Michigan and Stanford University.

Brennan G, Buchanan JM (1985) The Reason of Rules: Constitutional Political Economy. Cambridge University Press, Cambridge

Brennan G, Hamlin A (2000) Democratic Devices and Desires. Cambridge University Press, Cambridge

Brubaker R (1992) Citizenship and Nationhood in France and Germany. Harvard University Press, Cambridge MA

Buchanan JM (1975) The Limits of Liberty: Between Anarchy and Leviathan. University of Chicago Press, Chicago

Buchanan JM (1987) The constitution of economic policy. Am Econ Rev 77(3):243–250

Buchanan JM, Tullock G (1962) The Calculus of Consent: Logical Foundations of Constitutional Democracy. University of Michigan Press, Ann Arbor

Cherokee Nation v. Georgia, 30 U.S. (5 Pet.) 1 (1831)

Congleton RD (2011) Perfecting Parliament: Constitutional Reform, Liberalism, and the Rise of Western Democracy. Cambridge University Press, Cambridge

Constitution Act, 1982, being Schedule B to the Canada Act 1982 (UK), 1982, c. 11

Dar AM (2022) Revisiting key debates in the study of nationalism. Humanit Soc Sci Commun 9:411.

Delgamuukw v. British Columbia, [1997] 3 S.C.R. 1010

Ehret S, Witt A, Banisch S (2024) Convolution of individual and group identity: self-reliance increases polarisation in basic opinion model. Humanit Soc Sci Commun 11:829.

Enos RD (2017) The Space Between Us: Social Geography and Politics. Cambridge University Press, Cambridge

Haaland v. Brackeen, 599 U.S. ___ (2023)

Haida Nation v. British Columbia (Minister of Forests), [2004] 3 S.C.R. 511

Hainmueller J, Hopkins DJ (2015) The hidden American immigration consensus: a conjoint analysis of attitudes toward immigrants. Am J Polit Sci 59(3):529–548

Hardin R (1999) Liberalism, Constitutionalism, and Democracy. Oxford University Press, Oxford

Hart HLA (1961) The Concept of Law. Clarendon Press, Oxford

Hayek FA (1973) Law, Legislation and Liberty, Vol. 1: Rules and Order. University of Chicago Press, Chicago

Indian Child Welfare Act of 1978, 25 U.S.C. §§ 1901–1963

Indian Citizenship Act of 1924, Pub. L. No. 68-175, 43 Stat. 253

Inglehart R, Norris P (2019) Cultural Backlash: Trump, Brexit, and Authoritarian Populism. Cambridge University Press, Cambridge

Joppke C (2010) Citizenship and Immigration. Polity Press, Cambridge

Krange O, Kaltenborn BP, Hultman M (2021) “Don’t confuse me with facts” — how right wing populism affects trust in agencies advocating anthropogenic climate change as a reality. Humanit Soc Sci Commun 8:255.

Lone Wolf v. Hitchcock, 187 U.S. 553 (1903)

Mabo v. Queensland (No. 2), (1992) 175 CLR 1

McGirt v. Oklahoma, 591 U.S. ___ (2020)

Native Title Act 1993 (Cth) (Australia)

NORC at the University of Chicago (2022) General Social Survey, 1972–2021 cumulative data file.

Ordeshook PC (1992) Constitutional stability. Const Polit Econ 3(2):137–175

Ostrom E (1990) Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge University Press, Cambridge

Pew Research Center (2021) What it means to be an American: a survey of public views on national identity. Pew Research Center, Washington DC

R. v. Sparrow, [1990] 1 S.C.R. 1075

Royal Proclamation of 1763, R.S.C. 1985, App. II, No. 1

Schelling TC (1960) The Strategy of Conflict. Harvard University Press, Cambridge MA

Sides J, Tesler M, Vavreck L (2018) Identity Crisis: The 2016 Presidential Campaign and the Battle for the Meaning of America. Princeton University Press, Princeton

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

Tabatabaei H (2024) Populism and the quest for political power: the pitfalls to populist electoral success in Canada. Humanit Soc Sci Commun 11:290.

Tsilhqot’in Nation v. British Columbia, [2014] 2 S.C.R. 257

Tully J (1995) Strange Multiplicity: Constitutionalism in an Age of Diversity. Cambridge University Press, Cambridge

Voigt S (2011) Positive constitutional economics II — a survey of recent developments. Public Choice 146(1):205–256

Voigt S (2020) Constitutional Economics: A Primer. Cambridge University Press, Cambridge

Weingast BR (1997) The political foundations of democracy and the rule of law. Am Polit Sci Rev 91(2):245–263

Williams RA Jr (1990) The American Indian in Western Legal Thought: The Discourses of Conquest. Oxford University Press, Oxford

Wimmer A (2008) The making and unmaking of ethnic boundaries: a multilevel process theory. Am J Sociol 113(4):970–1022

Wimmer A (2013) Ethnic Boundary Making: Institutions, Power, Networks. Oxford University Press, Oxford

Wimmer A, Cederman L-E, Min B (2009) Ethnic politics and armed conflict: a configurational analysis of a new global data set. Am Sociol Rev 74(2):316–337

Worcester v. Georgia, 31 U.S. (6 Pet.) 515 (1832)