Level II: Domestic Politics and Cooperation
This week’s main claim is that the deepest causes sit at Level II: in domestic state–society relations. Moravcsik makes preferences endogenous and prior; Putnam shows that every negotiator plays at two tables at once; Leeds, Moravcsik (2000), and De Vries et al. then trace how regime type, democratic “lock-in,” and the politicization of cooperation determine whether states can credibly commit and actually cooperate.
Reading list
- Moravcsik. 1997. “Taking Preferences Seriously: A Liberal Theory of International Politics.” International Organization.
- Putnam. 1988. “Diplomacy and Domestic Politics: The Logic of Two-Level Games.” International Organization.
- Leeds. 1999. “Domestic Political Institutions, Credibility Commitments, and International Cooperation.” American Journal of Political Science.
- Moravcsik. 2000. “The Origins of Human Rights Regimes: Democratic Delegation in Postwar Europe.” International Organization.
- De Vries, Hobolt, and Walter. 2021. “Politicizing International Cooperation.” International Organization.
- Application: Besch and Youngs. 2025. “Europe Needs a New Way to Cooperate.” Foreign Affairs.
Moravcsik (1997): Taking Preferences Seriously
International Organization · a liberal theory of international politics
Start at home. Societal identities and interests, filtered through domestic institutions, produce state preferences; the fit or clash of those preferences across borders, what he calls policy interdependence, then yields harmony, cooperation, or conflict. His abstract frames liberalism as a positive, non-utopian theory that puts state–society relations at the center: it is the configuration of state preferences, not power (realism), nor information and institutions (institutionalism), that “matters most in world politics.” Importantly, he presents this as a paradigmatic alternative to realism and institutionalism, not a moral stance. He claims liberalism can explain the substantive content of foreign-policy choices, long-term historical change in the system, and the distinctiveness of modern politics among advanced democracies (pluralistic security communities, complex interdependence). (1) Primacy of societal actors: states are embedded in domestic and transnational society; individuals and groups, pursuing welfare and typically risk-averse, formulate demands and induce the state to pursue social purposes, with conflict and cooperation reflecting differences in beliefs, scarcity, and unequal political power. This is strong, abundant evidence shows domestic coalitions shape foreign policy, but preferences are hard to measure, risk-aversion varies, and endogeneity (preferences shaped by prior international politics) must be handled with care. (2) Representation and state preferences: state preferences are the social purposes governments pursue, distinct from strategies, and derived from who gets voice and how intensely; institutions filter and aggregate demands, which is highly plausible, though mapping institutions onto precise preference intensities is hard. (3) Interdependence and the international system: outcomes turn on how each state’s pursuit of its purposes creates externalities for others, compatible purposes give harmony, mixed ones give bargained cooperation, incompatible ones with large negative externalities give conflict. The micro-logic is strong because IR is full of externalities; the difficulty is measuring the interdependence structure and separating it from power and information effects. Ideational liberalism: domestic social identities and values, what social order is legitimate, set preferences over borders, regime type, and socioeconomic regulation. Where national identity and borders align, expect coexistence; where they misalign, expect secession, intervention, or border disputes. Where states share views of legitimate institutions (e.g., both democracies), cooperation is easier; clashing ideals (democracy vs. monarchy/authoritarianism) breed conflict. And contested regulation (immigration, welfare, environment) that imposes transborder externalities shapes contestation (e.g., EU regulatory politics). It overlaps with constructivism but stays distinct by tying identity to domestic preference aggregation. Commercial liberalism: transborder market incentives create domestic winners and losers, and states pursue policies reflecting these distributional preferences. Importantly, it does not predict automatic free trade or peace, Moravcsik flags two disjunctures: concentrated losers lobby for protection even when openness maximizes aggregate welfare (rent-seeking), and imperfectly competitive global markets invite strategic trade (dumping, subsidies) that can raise conflict. Trade is usually a cheaper route to wealth than coercion, but where production structures make coercion profitable (monopolizable resources) force may follow; as economies diversify and network, coercion gets costlier and war incentives fade. Republican liberalism: domestic representation determines whether narrow groups can impose negative externalities (rent-seeking, coercion) on others; broad, accountable institutions internalize externalities and yield the democratic-peace logic. But more representation does not automatically mean more cooperation, elites are sometimes more convergent than mass publics (Concert of Europe), and the bias of representation matters more than its sheer breadth. All three are “liberal” because they share one foundational claim: variation in state preferences, rooted in domestic society and aggregated through institutions, is the key independent variable.
Liberalism endogenizes preferences (domestic society → state preferences), whereas realism treats preferences as fixed and focuses on capabilities. “Functional regime theory” is Moravcsik’s label for neoliberal institutionalism (Keohane, Oye, Martin): it fixes state preferences and studies how rules and information help cooperation under anarchy. He calls it “not liberal” because in his typology liberal means endogenizing preferences via society, while institutionalism takes preferences as exogenous and studies strategic interaction, its assumptions sit closer to realism. He even argues baseline regime predictions can often be derived from liberal theory once preferences are modeled, so ignoring preference formation risks omitted-variable bias. Against constructivism, he charges that it often tells a top-down systemic socialization story and lacks a domestic preference-formation theory; liberal variables are prior and necessary to specify when ideas and socialization should matter, and constructivists increasingly lean on unit-level state–society change (democratic peace, the “distinctiveness of the West”) anyway. Unlike the classical liberalism / idealism of thinkers like Norman Angell, Moravcsik’s is a positive social-scientific theory of preference formation and interdependence, not moral optimism, altruism, or faith in law. His League of Nations vignette is the key contrast: he reads Wilson as a two-stage liberal who wanted republican self-government to converge among the great powers first, with collective-security institutions second, and who was himself skeptical of strong autonomous institutional effects. Wilson set narrow preconditions, near-universal democracy among the powers holding most military force, and when those failed to obtain (by 1936 twelve European states were dictatorships), failure followed, exactly as a liberal analysis predicts: institutions reflect and depend on underlying preference convergence, and design alone cannot save cooperation without it. The two-stage model is also his method: states first define preferences (liberal theories of society and representation), then bargain (where realist and institutionalist mechanisms operate). On coherence and distinctness, yes: the three assumptions hang together and separate liberalism from realism (capabilities) and institutionalism (information/rules) by making preferences endogenous, and they force a clear research design, identify and measure preferences first. On empirical generality, largely yes, the framework travels across security and political economy (democratic peace, trade, integration), though measuring preferences is nontrivial and disentangling endogenous international socialization takes care. But I have four reservations. First, his liberalism, Waltz’s structural realism, and Keohane’s institutionalism all share the rationalist “economic man” assumption and are essentially non-historical, modeling built on the era of the modern nation-state, democracy, and social-contract theory rather than tracing states and institutions anthropologically; they differ over which factor matters most, not over incompatible worlds. Second, the two-stage model cannot really merge with Waltz: Waltz deliberately excludes domestic preferences, treating states as functionally identical black boxes to build an abstract, hard-to-falsify theory of structure, not of specific policy, so you cannot simply “add” a preference layer to it, and Moravcsik’s talk of fusion is more a rhetorical move to claim paradigmatic parity than a logical synthesis. The theories speak to different layers, structure (Waltz), mid-level interaction (institutionalism and liberalism), and discourse/critique (constructivism, feminism, postcolonialism), so at best they complement. Third, Stage 1 is essentially what comparative politics already does (how interest groups form preferences, how institutions aggregate them, how identities shape goals); his real move was to relabel domestic study as IR theory, a redrawing of disciplinary boundaries and a meta-level reorganization of existing findings more than a new mechanism. Fourth, if the core is just domestic preferences, constrained by others’ preferences and the international environment, represented through institutions and producing outcomes in state interaction, why insist on the contested word “liberalism” at all, a term whose meaning has shifted for centuries from aristocratic generosity to constitutional rights, rather than simply renaming it?
Putnam (1988): The Logic of Two-Level Games
International Organization · diplomacy and domestic politics
Putnam sees many international negotiations as a game played simultaneously at two tables. At Level I (the international table) national governments bargain with one another; at Level II (the domestic table) parties, legislators, agencies, interest groups, unions, and public opinion decide whether to ratify what the government negotiates. The key player is the national leader, who appears at both boards at once, flanked domestically by party and interest-group actors and internationally by foreign counterparts and diplomats. For exposition he decomposes the process into two stages: Level I bargaining that yields a tentative deal, then Level II ratification by domestic constituencies. But he warns this sequence is only an analytic convenience, through “expectational effects,” pre-consultations and the anticipation of ratification feed back into Level I from the very start. A country’s win-set is the set of all possible Level I agreements that would win the necessary support at Level II if put to an up-or-down vote. Putnam illustrates with the prewar Falklands/Malvinas talks: several tentative bargains failed because they fell outside one side’s domestic win-set, and once it became clear the win-sets did not overlap, war became likely. The geometry of overlap is everything, if the two win-sets share no common ground, no deal exists that both sides can ratify. There is a trade-off built into “state strength.” Greater autonomy from domestic constraints enlarges your win-set (you can cooperate more easily) but weakens your leverage, because you cannot credibly claim your hands are tied. The relative sizes of the two win-sets decide who captures more of the cooperative gains: the larger your perceived win-set, the more you can be pushed around; a small domestic win-set is a bargaining advantage, “I’d love to do more, but my legislature won’t let me.” This is the classic Schelling logic of credible constraint: useful, but risky, because if your win-set shrinks so far that the overlap disappears you get deadlock, not better terms.
(1) Level II preferences and coalitions: domestic interests, the costs of “no agreement,” issue heterogeneity, and politicization shape who supports ratification. Lower no-deal costs shrink the win-set; heterogeneous issues can split coalitions yet sometimes ease agreement by creating cross-table allies. When an issue is politicized, actors indifferent to “no agreement” get mobilized and the win-set shrinks (Putnam’s example: the U.S. Panama Canal debate). (2) Level II institutions: ratification rules (supermajorities), separation of powers, fast-track procedures, political culture, and party discipline. A two-thirds Senate requirement makes U.S. win-sets smaller than simple-majority systems; the 1974 Trade Act’s fast-track created a straight up-or-down vote to enlarge the win-set and reassure partners; Japan’s consensus style constricts its win-set, while strong party discipline (the 1986 tax reform) enlarges a chamber’s. (3) Level I negotiators’ strategies: targeted side-payments, cultivating generic good will, reinforcing the counterpart’s popularity, issue linkage (especially synergistic linkage), and exploiting uncertainty or a “kinky” win-set, e.g., side-payments to swing votes on the Panama Canal Treaty, or U.S. energy-decontrol signaling to Bonn partners in 1978 that the proposal would pass but slightly tougher terms would not. As chief negotiator I would: ensure overlap before squeezing (use constraints for leverage, but don’t over-shrink the win-set or the deal vanishes); package and link issues synergistically so domestic winners can compensate losers on each side, expanding both win-sets; target offers and retaliation at swing constituencies at home and abroad rather than committed hawks or doves, maximizing ratification probability per concession; cultivate good will and reinforce my counterpart’s standing, since a popular opposite number has a bigger win-set; manage expectations with “kinky win-set” signaling while remembering misrepresentation can backfire in integrative bargaining; and design for deliverability, since involuntary defection (failed ratification) differs from cheating, so timelines and reciprocity must respect domestic ratification calendars. The theory’s main implications can fail when: win-sets simply do not overlap (the prewar Falklands), so constraint yields deadlock or conflict; preferences are not fixed and “reverberation” lets international pressure restructure domestic coalitions, so comparative statics on fixed win-sets need updating; the negotiator’s own preferences diverge from constituents (principal–agent), letting agenda power, timing, and media cut across the constituents’ win-set logic; or there are more than two levels (EC agricultural policy plus member states plus coalition parties), changing the geometry of overlap and the risk of involuntary defection. A promising extension is to empirically test whether governments really target swing domestic groups across borders in trade disputes.
Leeds (1999): Domestic Institutions & Credible Commitment
American Journal of Political Science · regime type and cooperation in the dyad
A dyad is simply a pair of states; Leeds studies bilateral cooperation, how two specific states interact in a given year, so her unit is the dyad-year. Her opening explicitly considers “the influence of domestic political institutions on the ability of leaders to establish bilateral cooperation.” I found the formal model genuinely clarifying because it links agreement formation and agreement fulfillment inside one strategic setting: the decisions to form and to fulfill agreements are “not separate decisions, but inextricably linked,” so if leaders expect they cannot fulfill, they won’t form the agreement in the first place. That cleans up a common mistake, studying compliance while ignoring selection into agreement; samples that include only signed deals are biased, and we should study the choice to cooperate and the success of coordination jointly. Leeds argues democracies face higher domestic accountability and less policy flexibility, so leaders find both breaking promises and being exploited costly; their commitments are therefore more credible and they seek only secure agreements, hence H1: jointly democratic dyads cooperate at comparatively high levels. I would qualify this. Accountability also makes leaders more sensitive to shifts in public opinion and party turnover, which often produces reversals, think of U.S. presidents leaving international agreements after partisan change, whereas one-party states like China can provide continuity without elections. So democracies are both more likely to join and to exit agreements, depending on domestic consensus; Leeds captures one side of the story but misses how polarization and regime longevity shape credibility. In practice secrecy, backroom deals, and the “dirty hands” problem exist in both democracies and autocracies, so what really matters is not regime type alone but leaders’ motivations, domestic preferences, issue salience, and the degree of domestic division. Autocracies have more policy flexibility and lower domestic constraints, so they find it harder to guarantee future behavior but are more willing to accept agreements carrying some risk of abrogation; when both sides share these traits they more often reach an equilibrium compatible with cooperation than a mismatched pair does, H2: jointly autocratic dyads cooperate more than mixed dyads. Mixed dyads cooperate least (H3) because the autocracy has weaker incentives to fulfill while the democracy is less willing to bear the risk of defection, so uncertainty about fulfillment is especially corrosive. I broadly agree on the mechanism of mismatched credibility, but real cases are noisier, shared interest, dependence, or great-power patronage can pull mixed pairs into cooperation regardless of regime fit. To isolate regime type Leeds added controls: shared alliance membership (a Cold War East–West alignment proxy), joint wealth relative to the U.S., and joint political stability (no regime change in five years). What she did not control for is substantial: deeper state capacity (bureaucratic quality, ability to tax and implement), colonial legacies between former colonies and metropoles, economic interdependence (trade, investment, supply chains), foreign-aid dependence (U.S. or Soviet aid steering who cooperated with whom), conflict history and enduring rivalries, and cultural proximity (shared language or religion).
Alternatives that would yield different predictions: an institutionalist/IO-monitoring account where strong institutions raise everyone’s payoff and narrow regime-type gaps, so high-transparency dyads cooperate regardless of regime (Leeds herself acknowledges the classic “no external enforcement” problem); an economic-interdependence account where large mutual dependence and domestic lobbies push even mixed dyads toward cooperation; and a rivalry/history account where longstanding rivalries swamp regime effects, so even democratic rivals cooperate less. Table 1 is titled “Regime Type and International Cooperation, 1953–1978; Unit of Analysis: Dyad-Year” and reports three models, Model 1 (continuous average cooperation), Model 2 (any cooperation, logit), Model 3 (intensity conditional on cooperation > 0), with regime dummies (jointly autocratic, mixed; anocracies as reference) and controls (jointly wealthy, jointly stable, shared alliance); N is large (22,320 dyad-years for M1/M2; 11,815 for M3). The dependent variable is built from COPDAB events (Azar), coded on a 15-point scale, weighted for intensity, keeping only cooperative events and averaging per dyad-year, so higher non-zero values mean greater intensity, not frequency (avoiding media-frequency bias). Regime type comes from POLITY III (jointly democratic if both democracy ≥ 6; jointly autocratic if both autocracy ≥ 6; mixed otherwise); the sample is every politically relevant dyad (PRIE: contiguous states plus major/regional powers) each year, 1953–1978. The takeaways: in Model 1 both jointly democratic and jointly autocratic are positive and highly significant versus anocracies, with mixed positive but lower; in Model 2 predicted probabilities run 58.4% (joint democracies), 53.5% (joint autocracies), 51% (mixed), 43.3% (at least one anocracy); in Model 3 joint democracies retain significantly higher intensity, and wealth, stability, and alliance are all positive and significant. The design is careful and the joint formation–fulfillment logic is a real contribution, but the regime measure is the weak link. POLITY III scores each country-year on how leaders are selected, whether opposition is allowed, and whether executive power is constrained, and the ≥ 6 thresholds give a clean binary that is convenient for large-N work, but it is far from perfect. Authoritarian states are diverse (military, party-based, personalist), and Western democracies emerged gradually through state-building, while many developing countries inherited institutions shaped by colonial powers and aid incentives; post-colonial states often adopted constitutions that look democratic on paper but lack social legitimacy and bureaucratic capacity, so “democracy scores” can capture formal rules rather than real practice. The distinction has analytic value, but it risks attributing outcomes to regime type when they are really driven by state capacity, colonial legacies, informal power structures, or the social acceptance of institutions, exactly the confounders I would most want to see addressed.
Moravcsik (2000): The Origins of Human Rights Regimes
International Organization · democratic delegation in postwar Europe
Honestly I don’t find it deeply puzzling, whatever happens, there is some reason for it, but Moravcsik frames it as a puzzle: why would sovereign states bind themselves to international courts that can overrule them on rights? He resolves it with a republican-liberal “lock-in” argument: democratic governments sometimes delegate to international courts to tie the hands of future governments and stabilize a vulnerable domestic democracy. Governments “turn to international enforcement when an international commitment effectively enforces the policy preferences of a particular government at a particular point in time against future domestic political alternatives,” especially when the benefits of reducing future political uncertainty outweigh the sovereignty costs. Placing rights interpretation “in the hands of independent authorities managed in part by foreign governments” is a two-level, credibility-enhancing commitment that locks in democracy against backsliding. Realism (power/coercion) expects great powers to impose or induce acceptance of human-rights rules, so binding regimes should track concentrations of power and hegemonic interest, strong commitments where a hegemon can push reluctant states. Ideational theory (altruism/persuasion) expects altruistic governments in established democracies and transnational networks to spread universal norms through socialization and the logic of appropriateness; less-democratic states are persuaded or socialized. Both share a hunch that established democracies will lead. I don’t fully buy his framing. He flattens the alternatives into “hard promotion” versus “soft promotion,” but realism is not the same as coercion: it starts from anarchy and the security imperative, so whether great powers back human-rights regimes depends on whether such regimes serve their security interests, and if binding commitments in other states reduce instability or make the system more predictable, great powers might support them, even promoting them through the language of values, not just force. So Table 1’s columns contrast two modes of diffusion, powerful imposition versus moral persuasion, more than they faithfully render either tradition; the simplification conveniently sets up a neat contrast with his own republican-liberal theory by reducing rivals to stylized extremes. Republican liberalism stresses how domestic representation and institutions shape foreign policy, and how international institutions can strengthen the state at home by making current policies more credible, locking in rules against reversal. He sets three conditions: fear of future domestic uncertainty; supportive consensus among foreign governments; and international cooperation that induces domestic support for maintaining current policies. I’m not fully convinced. First, his test omits other motives, governments might back binding rules to signal commitment to European integration, to attract investment, to burnish their image, or to use external courts as leverage for domestic reform, none systematically examined. Second, binding commitments raise the cost of backsliding but don’t guarantee stability: enforcement is slow and compliance often partial, so “lock-in” is better seen as one tool that raises the probability of stability, not a decisive safeguard. Third, if domestic actors can build stability through strong constitutional courts, independent oversight, or reform, why incur costly international obligations? He answers that international commitment is a “second lock” harder for future governments to undo, but that doesn’t explain why some states rely on domestic solutions while others embrace delegation. Finally, treating states as unitary and measuring preferences only through support for two institutional clauses cannot capture the variety of domestic debates, party interests, and strategic calculations; we’d need case-by-case studies tracing domestic politics and international bargaining together.
Sovereignty costs are the loss of national discretion when delegating to international judges, higher than domestic delegation because a common international list of rights will diverge from national traditions and impose centralized, harmonizing jurisprudence. Governments accept such costs when reducing domestic uncertainty is worth more than discretion, so the strongest support should come from recently established, potentially unstable democracies, not from long-established democracies or dictatorships. This generates his signature inverse-U: new democracies, with the most to fear from backsliding and the least entrenched traditions to protect, are the ones most willing to bind themselves. The European Convention on Human Rights (ECHR), under the Council of Europe in Strasbourg, is widely seen as the most advanced and effective enforceable human-rights regime, with a Commission and Court whose decisions bind governments; optional clauses (Articles 25 and 46) enable individual petition and compulsory jurisdiction, and many states have incorporated it domestically. Moravcsik calls his study the first systematic empirical test of competing theories, using the 1949–50 negotiations, and it is a hard test precisely because the ECHR is substantively effective, so the design choices reveal states’ real willingness to incur sovereignty costs. The takeaway: new democracies championed binding enforcement (compulsory jurisdiction plus individual petition), while established democracies and dictatorships opposed reciprocal mandatory commitments, the inverse-U of his Table 2. The bargaining process shows no coercion by great powers and little durable persuasion; instead a classical lowest-common-denominator bargain with abundant optional clauses, consistent with fixed national preferences. In the important UK case, opposition centered on fears of domestic legal nullification, again the sovereignty-cost logic. It is a good test because he triangulates three strategies, cross-national variation in positions, bargaining-process evidence, and confidential domestic deliberations (unusually rich UK archives), that converge on the lock-in claim, though it is region- and time-bound (Western Europe, 1949–53) and hard to fully net out variables like post-WWII moral shock. Beyond the ECHR he argues the logic explains other rights institutions (the ICC, even “anomalous” U.S. behavior), mirror-image authoritarian self-binding (the Holy Alliance, the Warsaw Pact/Comintern shoring up autocracy), and economic lock-in (NAFTA as a credibility device for Mexican liberalization; European integration as executive centralization to entrench commitments). I’d add counterexamples he himself brackets: transitional regimes under military sway that don’t lock in, and norm cascades producing broad declaratory uptake without reciprocal adjudication, which he dismisses as largely rhetorical absent enforcement. The locking-in logic plausibly generalizes to trade and monetary coordination and even non-liberal aims, provided his three conditions hold.
De Vries, Hobolt & Walter (2021): Politicizing International Cooperation
International Organization · when does cooperation become contested?
Voters in Brazil, Britain, the Philippines, and the United States elected politicians who promised to leave or challenge international institutions (the Paris Agreement, the ICC, the EU), alongside a wider turn toward isolationism, nationalism, and protectionism, with once “virtually impossible” events, a member state leaving the EU, WTO dispute settlement unraveling, actually happening. Using the ISSP across 21 countries (1995/2003/2013), they show declining support for IOs’ right to enforce solutions and rising agreement that IOs take too much power from national governments, plus ambivalence: many endorse international problem-solving while rejecting automatic compliance with IO decisions. The Chapel Hill Expert Survey shows challenger parties are more Euroskeptic and mobilize opposition (often via immigration/national-identity frames) while mainstream parties stay pro-EU on average. European Parliament elections, second-order, with permissive rules, act as an incubator for cooperation-skeptic entrepreneurs (such parties topped the polls in Britain, France, and Italy in 2019). And referendums on cooperation have grown more common and more often go against governments: of 84 worldwide between 1970 and 2019, “No” votes have increased, and since 2010 voters have rejected cooperation in every other such referendum. Since 2021 I would add Trump as a further example. They group existing explanations into three families. Economic: distributional losses and inequality from open economies, the “left behind” thesis. Cultural: value and identity change, the salience of migration, and status-threat narratives. Institutional: how rising international authority and domestic channels (elections, referendums) structure discontent and translate it into challenges. Popular news narratives are the unsystematic cousins of these: jobs and the “China shock” (trade hollowing out manufacturing); sovereignty and control (IOs “bossing Washington around,” courts and trade bodies limiting policy freedom); immigration and identity (cultural change linked to open borders); elite technocracy and “globalists” (a democratic deficit, decisions made far from voters); and misinformation and polarization (social media amplifying distrust and framing international rules as threats). The article’s contribution is to systematize these heuristics into the economic/cultural/institutional families and then move past a static list of grievances to a theory of when grievance becomes politics. Politicization is “the process of making an issue political, that is, debating it in the public sphere as an issue of public contestation.” It is not inherently bad. Politicization can complicate creating and sustaining institutions, but it can also increase legitimacy through public buy-in, depending on the contestation and the opportunity structures. So the normative valence is conditional: the same process that can paralyze cooperation can also democratically legitimize it.
The core claim has three scope conditions: an international issue becomes politicized when (i) public discontent exists and is mobilized by (ii) political entrepreneurs, and it turns into a systemic challenge only when (iii) political opportunity structures are permissive. Discontent is necessary but not sufficient, entrepreneurs must activate it, and opportunity structures let it become an effective challenge. A political entrepreneur ignites dormant opinion or mobilizes a slice of discontent for electoral advantage, often through wedge-issue strategies amplified by media; challenger parties are central because they sit at a disadvantage on the main left–right axis and use issue entrepreneurship on cooperation (especially the EU), but there are also cosmopolitan entrepreneurs (Trudeau, Macron, Ardern) pushing pro-cooperation frames. Table 1 crosses public contestation (high when entrepreneurs mobilize opinion; low when the public is content or unmobilized) against political opportunity structures (the institutional channels, elections, referendums, coalition rules, that translate domestic contestation into international challenges), yielding four cells: low × low is a “permissive consensus” of elite-driven, low-salience cooperation; low × high is “democratically legitimized” cooperation (open institutions, no mobilized backlash, legitimacy gains); high × low is “public discontent” without systemic effect (Austria, strong mobilized Euroskepticism but no referendums and a constraining coalition); high × high is “challenged cooperation” (the Dutch 2005 EU Constitution referendum with Wilders; elections like Trump or Bolsonaro running against institutions). My main worry is endogeneity: the framework treats whether a referendum is held or what electoral system a country has as external, fixed “channels,” but these are often the product of political choices that are themselves shaped by discontent and elite strategy. That breeds three problems. Reverse causality: it looks like “holding a referendum causes rejection of cooperation,” but high discontent may push elites to call the referendum, so discontent causes both. Omitted-variable bias: hidden factors, media polarization, party splits, declining trust, may raise discontent and make referendums more likely, making the referendum look causal. Strategic selection: governments or oppositions allow referendums only when they expect to win or to shift blame, so the very cases where a referendum happens are already high-conflict, exaggerating the referendum–failure link. The article leans on Europe because many measures are EU-specific, CHES party positions on integration, EP elections as second-order incubators, and EU-centric referendums, even though global referendums exist (Brazil, Taiwan, the UK). The mechanism (discontent → entrepreneurs → opportunities) should travel to other democracies with referendums or outsized executive authority (the U.S. via elections, Canada/Australia via party entrepreneurs, Switzerland via referendums), and the authors explicitly discuss U.S. and Brazilian cases. Across non-democracies it travels poorly: where mass contestation can’t be mobilized (repression, media control) or opportunities are absent (no free elections or referendums), the levers weaken, and elite-driven exits like junta withdrawals from ECOWAS or the ICC produce international change without public politicization, a different causal path. Useful extensions: study when cosmopolitan entrepreneurs de-politicize versus re-legitimize cooperation; distinguish mandatory (more exogenous) from optional (strategic) referendums; test how frame–issue match (economic vs. identity) conditions mobilization across trade, courts, climate, and migration; and study beyond Europe.
Application · Theory in current events
Besch and Youngs. 2025. “Europe Needs a New Way to Cooperate.” Foreign Affairs.
Besch and Youngs start from the premise that Europe’s main alliances, the EU and NATO, are often paralyzed, and inside the article the premise holds up: repeated failed EU reform efforts; reliance on U.S. leadership in NATO (the United States as “first among equals”); value and strategy splits (e.g., Hungary); and the failure of ad hoc minilateralism to scale without losing effectiveness. EU members “do not fully trust” Brussels or one another; a long string of crises (the eurozone crisis, Brexit, COVID-19, Russia’s full-scale invasion) depleted the sense of community, so without the “disciplining power” of U.S. leadership Europeans must agree among themselves, something they struggle to do, with reform promises on decision-making, budgeting, and participation never materializing amid sovereignty reticence, bureaucratic inertia, and legal complexity. Their proposal is “treaty-based clusters”: multiple policy clusters anchored by an institutional core, each with its own governance led by participating governments plus oversight (an intergovernmental secretariat or parliamentary body), with voluntary entry and overlapping memberships. Keep what the EU already does well (tech, digital, trade) but use clusters to break deadlock on contentious areas (defense, climate); security clusters require a firm commitment to liberal-democratic values while others can be broader; membership is not static, since governments can be suspended or expelled for norm violations, and decisions bind only members. Illustrative use-cases: a UK–EU political and security alliance bridging Brexit rifts; an institutional path to bring Ukraine into European security despite NATO and EU accession constraints; and a Nordic–Mediterranean climate cluster of complementary energy profiles, all reusing NATO command-and-control and EU financing as the spine rather than duplicating them.
My assessment is critical but constructive. The clusters tackle the speed–legitimacy–capacity trilemma better than ad hoc clubs: they create rule-bound minilateralism and open doors to non-EU partners (the UK, Ukraine) without requiring EU unanimity. The risks are fragmentation and institutional overlap (who actually leads strategy, NATO, the EU, or the clusters?), budgetary legitimacy if clusters raise and spend money outside EU processes, and the political difficulty of value-gatekeeping (suspending a backsliding member is hard), all solvable but requiring precise legal templates, financing rules, and interface agreements that the essay sketches without fully specifying. Read through the De Vries et al. framework, new cooperation gains support when three kinds of conditions align: economic (visible distributional gains for defined constituencies, defense jobs from joint procurement, cheaper energy via a complementary climate cluster, lowering backlash and building coalitions of winners); cultural/identity (clusters framed as defending the liberal order and European security can mobilize cosmopolitan constituencies, especially after Russian aggression and perceived U.S. retrenchment heightened threat perceptions); and institutional/opportunity (salient shocks like war and pandemics that raise attention to cross-border externalities; openings for pro-cooperation entrepreneurs to frame clusters as problem-solving rather than “Brussels overreach”; and procedural venues, treaties ratified by parliaments, that legitimize and bind governments while avoiding EU-wide unanimity). On likelihood I’d call it moderately plausible: likely champions are liberal democracies that share threat perceptions and want speed without full unanimity, a UK–EU security rapprochement, Ukraine’s near-term inclusion via clusters (since NATO is “out of reach” and EU accession “too slow”), and issue-aligned subgroups like the Nordic–Mediterranean climate cluster, while likely obstacles are illiberal or obstructive members (Hungary, possibly Slovakia) whom the authors would move away from vital cooperation. The choke points are legal design, money, and value enforcement, all designable but politically heavy. Finally, on whether cooperation is easier by reforming existing institutions or replacing them, Besch and Youngs answer “both/and”: keep EU instruments that work while adding parallel treaty-based clusters to bypass unanimity in deadlocked areas. I agree, reform leverages path dependence and legitimacy but stalls under unanimity and value splits, while replacement/parallelism buys speed and fit at the risk of fragmentation unless interfaces and finance are carefully designed (the regime-complex literature would call clusters a modular complement, not a clean replacement); practically, do targeted parallelism plus incremental EU/NATO reform. My closing intuition is that Europe’s deeper problem may be insufficient will to cooperate rather than a flaw in the cooperation mechanism itself.
← Back to the genealogy tree · ← Prev: Week 5, Constructivism · Next: Week 7, Regime Type & Conflict →