Democracy as a Contemporary Global Concern

Democracy entered international politics as a domestic form of government and left the twentieth century as a global standard of legitimacy — the thing states are graded on, aided for, sanctioned over and lie about. The concern today is not whether democracy is desirable, a question on which world opinion is remarkably settled, but whether it is reproducing itself. For two decades the count of democracies has fallen, the erosion has come from elected leaders rather than colonels, and the states that once financed democracy abroad have withdrawn from the business.

Why democracy is a global concern and not merely a domestic one

• Democracy became an international question the moment legitimacy became internationally judged. A state’s internal constitution now determines its access to aid, alliances, trade preferences and recognition.
  • Membership conditionality made the link concrete — the European Union’s Copenhagen criteria, the Council of Europe, the Organization of American States, and the African Union’s bar on unconstitutional government.
  • Recognition politics followed: after a coup, the question of who sits in a state’s UN seat and who signs its loans becomes an international decision.
  • Democratic peace theory supplied the security argument — if democracies do not fight one another, the internal regime type of other states is a legitimate object of one’s own foreign policy.
• The externalities of misgovernment cross borders, which is what converts a constitutional question into a global one.
  • Refugee flows, capital flight and transnational crime follow state failure, and state failure follows the collapse of accountable government.
  • Election interference and disinformation are now cross-border operations, so the integrity of one state’s ballot is affected by another state’s agencies.
• Democracy is the only global norm with no global treaty. There is no democracy convention, no supervisory body, no compulsory jurisdiction.
  • What exists instead is a patchwork of soft law, regional conditionality, funding networks and rankings — an architecture strong on judgement and weak on obligation.
  • The consequence is that enforcement is political rather than legal, and therefore selective, which is the single most powerful argument used against democracy promotion.

Democracy is the only major global norm with no treaty behind it — judged everywhere, obligatory nowhere.

Procedural and substantive democracy: what is actually being counted

• The minimalist or procedural definition treats democracy as a method of selecting rulers, not a set of outcomes.
  • Przeworski’s version is deliberately austere: a democracy is a system in which incumbents lose elections and leave — a regime where alternation is possible, and observed.
  • The virtue of minimalism is measurability; its cost is that it counts as democratic a state where the vote is real and every other freedom is hollow.
  • Schumpeter gave the position its canonical form, reducing democracy to a competitive struggle for the popular vote and denying that it entails any particular common good.

“That institutional arrangement for arriving at political decisions in which individuals acquire the power to decide by means of a competitive struggle for the people’s vote.” — Schumpeter

• The substantive or liberal definition adds the constraints without which the vote is decorative.
  • Rule of law, an independent judiciary, media pluralism, minority protection and civil liberties are treated as constitutive, not consequential.
  • Dahl’s polyarchy sits between the two, listing the institutional guarantees — free elections, inclusive suffrage, freedom of expression, alternative information, associational autonomy — that make contestation real.
• The gap between the two is where the last thirty years of politics happened. Nearly every contemporary autocrat clears the procedural bar and fails the substantive one.
• A third, developmental definition is pressed largely from the Global South: political freedom without material capability is formal.
  • Sen’s capability argument reverses the sequencing case — famines do not occur in functioning democracies with a free press, so democracy is instrumentally protective, not a luxury of the rich.
  • Indian leaders have consistently argued that political and socio-economic rights must advance together, which is a substantive rather than a minimalist commitment.

Conception	Democracy consists in	Principal exponent	What it misses
Procedural / minimalist	Competitive elections and the possibility of alternation	Schumpeter, Przeworski	Elected governments that dismantle liberties
Liberal / substantive	Elections plus rights, rule of law and pluralism	Dahl, Zakaria	Risks defining democracy as whatever the West has
Participatory	Continuous citizen involvement, not periodic consent	Pateman, Barber	Impracticable at the scale of the nation-state
Developmental / capability	Political freedom plus the capability to use it	Sen	Blurs the line between democracy and development
Cosmopolitan	Democratic control over transnational decisions	Held	No demos and no enforcement above the state

• Held’s cosmopolitan democracy extends the concern outward: as decisions migrate to bodies beyond the state, democracy inside states becomes insufficient.
  • The claim is that democratic government does not stop at borders and that transnational institutions require their own accountability.
  • The standing objection is the absence of a global demos and of any enforcement above the state, which leaves cosmopolitan democracy a regulative ideal.
• Representative democracy remains the operative form worldwide — limited, because participation is brief and infrequent, and indirect, because the public selects rulers rather than ruling.
  • Its defenders call it the only practicable form at scale and credit it with stability, since distance from daily decision encourages compromise.
  • Radical democrats reply that it reduces participation to a ritual vote every few years, producing apathy and alienation rather than self-government.

Huntington’s waves and the reverse waves

• Huntington organised democratisation into three long waves, each followed or interrupted by a reverse wave in which some new democracies fell back.
  • A wave is a group of transitions from non-democratic to democratic regimes occurring within a specified period and significantly outnumbering transitions the other way.
  • The framework’s real contribution is the reverse wave — the insistence that democratisation is not a ratchet, which is exactly the claim the present decade has vindicated.

Wave	Period	Drivers	Reverse wave
First	1828–1926	Industrialisation, franchise extension, national self-determination	1922–42 — fascism in Italy, Germany, Spain, Japan
Second	1943–62	Allied victory and post-war occupation, decolonisation	1958–75 — military rule in Latin America, Asia, Africa
Third	from 1974	Legitimacy crises of authoritarian rule, Catholic Church’s shift, EC/US policy, snowballing demonstration effects	Contested: the democratic recession from roughly 2006

• The third wave began in Portugal’s Carnation Revolution and spread by demonstration effect — Huntington’s “snowballing”, where one transition lowers the perceived cost of the next.
  • Southern Europe went first: Portugal, Greece, then Spain after Franco.
  • Latin America followed through the 1980s as military juntas withdrew after economic failure and the Falklands defeat.
  • East and Southeast Asia came next — the Philippines in 1986, South Korea in 1987, Taiwan’s lifting of martial law.
  • Eastern Europe transformed the wave into a global claim after 1989, followed by the post-Soviet states and, less durably, parts of Africa after 1990.
• Huntington’s own explanation was multi-causal, and it is the combination that matters rather than any single factor.
  • Performance legitimacy is fragile: authoritarian regimes that justify themselves by growth lose their case when growth stops.
  • Economic development created urban, educated middle classes with the resources to organise.
  • External actors mattered — the shift of the Catholic Church after the Second Vatican Council, the European Community’s membership incentive, and a change in United States policy.
• Huntington’s earlier and darker argument is the one the last decade has revived. In Political Order in Changing Societies he argued that political decay follows when participation outruns institutionalisation.

“The most important political distinction among countries concerns not their form of government but their degree of government.” — Huntington

• Rapid mobilisation without parties, courts and bureaucracies capable of absorbing it produces instability, not democracy.
• The prescription — order and institutionalisation before mass participation — is the intellectual ancestor of every contemporary sequencing argument.

The recession and the vocabulary built to describe it

Diamond and the democratic recession

• Diamond named the democratic recession, dating its onset to around 2006 and defining it by three simultaneous movements.
  • A rising rate of democratic breakdown among third-wave democracies, concentrated in low-income and conflict-affected states.
  • Erosion of quality inside surviving democracies — narrowing space for media, opposition and civil society without formal breakdown.
  • A deepening of authoritarianism in the states that never democratised, together with a collapse of confidence among the established democracies themselves.
• His later formulation shifted from count to confidence. The number of electoral democracies fell only modestly; what fell sharply was the belief among democratic publics and governments that the model was winning.

Zakaria and illiberal democracy

• Zakaria separated the two halves of “liberal democracy” and argued they were coming apart.
  • Illiberal democracy is a regime that holds real elections while eroding civil liberties, rule of law and the separation of powers — democratically legitimate, constitutionally hollow.
  • He noted that a very large share of the world’s democracies fit the pattern, and that the share was rising.
  • The corollary is constitutional liberalism without democracy — Hong Kong under British rule, or Singapore — which he treated as a real and sometimes preferable category.
• The policy conclusion was a sequencing claim: rule of law and state capacity should precede unrestricted mass competition, because elections held into a legal vacuum entrench the strongest faction.
  • Critics answer that the sequencing argument is infinitely postponable — every autocrat claims the country is not ready — and that no reliable mechanism makes an unaccountable ruler build the rule of law first.

Levitsky and Way on competitive authoritarianism

• Levitsky and Way identified a distinct regime type rather than a way-station on a transition.
  • Competitive authoritarianism is a civilian regime where formal democratic institutions are the principal route to power but incumbents so abuse the state that the contest is unfair.
  • It is not a diminished democracy but a diminished authoritarianism — competition is real and incumbents can lose, but the playing field is skewed by control of media, finance, the courts and the security services.
  • Their key finding is that linkage to the West — trade, travel, migration, media — predicted democratisation far better than leverage or aid, which is a direct challenge to democracy-promotion theory.

Bermeo’s typology of backsliding

• Bermeo’s intervention was to show that the forms of democratic breakdown changed even as the fact of it persisted.
  • Classic coups d’état and executive coups — the old, visible modes — became rare after the Cold War.
  • Promissory coups replaced them: takeovers justified as a temporary restoration of democracy, with elections promised at an unspecified date.
  • Executive aggrandisement is the dominant contemporary mode — elected leaders weakening checks through legal channels, one institution at a time.
  • Strategic election manipulation replaces ballot-box fraud with earlier, subtler interference: media access, voter-roll management, disqualification of candidates, campaign finance.
• The typology explains why international response is always late. Aggrandisement produces no single moment of rupture that a treaty body or a regional organisation can point to.

Levitsky and Ziblatt on the unwritten rules

• Levitsky and Ziblatt located democratic survival in norms rather than in constitutional text.
  • Mutual toleration — accepting rivals as legitimate rather than as enemies of the nation — and institutional forbearance — declining to use lawful powers to their destructive maximum.
  • Their argument is that constitutions cannot enumerate their own preconditions; court-packing, rule by decree and endless investigation are usually legal.
  • They add a gatekeeping claim: parties, not voters, are the filter that historically kept demagogues off the ballot, and party gatekeeping has collapsed.

Ginsburg and Huq on constitutional retrogression

• Ginsburg and Huq distinguish authoritarian collapse from constitutional retrogression — the slow, lawful degradation of competitive elections, rights of speech and association, and the rule of law.
  • Retrogression proceeds through amendment, appointment and administrative practice, each step individually defensible and cumulatively fatal.
  • Their design argument is that constitutional engineering can raise the cost of capture — dispersing appointment power, entrenching electoral administration, using multiple veto points — but cannot eliminate it.
• Scheppele named the technique autocratic legalism — the use of law, and especially of constitutional amendment, to disable constitutionalism.
  • The autocratic legalist does not suspend the constitution; he acquires the supermajority to rewrite it and then governs strictly by the text he wrote.
  • This is why legal formalism is no defence: the ruler complies with every rule while removing every constraint.

The contemporary autocrat does not suspend the constitution. He wins it, amends it, and then invokes it against his critics.

Deconsolidation and its critics

• Mounk and Foa argued from survey data that support for democracy was deconsolidating, especially among the young in rich democracies.
  • Declining shares called it “essential” to live in a democracy; rising shares expressed openness to army rule or a strong leader unconstrained by parliaments.
  • Mounk’s framing pairs undemocratic liberalism — rights protected but decisions removed from popular control by courts, central banks and treaties — with illiberal democracy, and treats populism as the collision of the two.
• The critics dispute the inference, not the data. Norris and others argue the effect is largely a cohort and question-wording artefact, that the young have always been less deferential, and that support for democracy in principle remains overwhelming.
  • The exchange matters because it decides whether the crisis is one of demand — publics no longer wanting democracy — or of supply — elites no longer providing it.
  • The weight of evidence points to a supply problem: aspiration stays high while institutions weaken, which is precisely the shape of the backsliding–backlash paradox.

Other lines of attack

• Runciman argues that mature democracies fail differently from young ones — not by collapse but by drift, gridlock and a middle-aged inability to act, with catastrophe arriving through technology and epistemic breakdown rather than through tanks.
• Applebaum focuses on the defection of elites, asking why intellectuals and professionals in established democracies choose the authoritarian side, and traces a transnational network of illiberal politics.
• Mudde treats populism as a thin-centred ideology dividing society into a pure people and a corrupt elite, hostile not to democracy but to liberal democracy’s constraints on the majority.
• Diamond’s later work adds the transnational dimension — sharp power, the projection of authoritarian influence through media, universities and party financing rather than through attraction.

Concept	Core claim	Distinctive feature	Author
Illiberal democracy	Elections without liberties	Locates failure in the liberal, not democratic, half	Zakaria
Competitive authoritarianism	Real but unfair competition	A regime type, not a transition stage	Levitsky and Way
Executive aggrandisement	Elected leaders dismantle checks legally	No rupture point for observers to name	Bermeo
Constitutional retrogression	Lawful degradation of three core goods	Emphasises constitutional design as remedy	Ginsburg and Huq
Autocratic legalism	Law used to disable constitutionalism	The amendment power as the weapon	Scheppele
Deconsolidation	Falling public commitment to democracy	Locates the problem in demand, contested	Mounk and Foa

The end of history, and what replaced it

• Fukuyama’s claim was never that events would stop, but that no rival ideology retained universal legitimacy — liberal democracy was the end point of ideological evolution.
  • The strongest evidence for the thesis remains negative: no competitor makes a universal claim. China exports a model of governance, not a doctrine other peoples are invited to adopt as their own.
  • Fukuyama himself has qualified it substantially, arguing that developed democracies are subject to political decay — institutional rigidity, capture by interest groups, and what he calls vetocracy — and conceding that democratic recession and the rise of identity politics were not anticipated by the original argument.
  • His mature position holds that the sequence matters: a capable state, the rule of law and accountability must all be present, and states that get democratic accountability before state capacity struggle.
• Ikenberry’s defence of the liberal international order is the institutionalist successor to the thesis.
  • The order is easy to join and hard to overturn because it is rule-based, open and embedded in institutions from which even rising powers draw benefit.
  • He argues that democracy and the rule of law remain the global standard of legitimate governance, and that no competing political ideal is even on the sidelines.
  • The counter-argument is that an order can survive as a set of economic arrangements while its political content is hollowed out — trade and finance continuing while democracy retreats.

“While democracy is not yet universally practised, nor indeed universally accepted, in the general climate of world opinion democratic governance has achieved the status of being taken to be generally right.” — Sen

Does development produce democracy?

• Lipset’s modernisation thesis held that economic development produces the social conditions — literacy, urbanisation, a large middle class — under which democracy emerges and survives.
  • Aristotle’s version long preceded it: the society best suited to democracy is one with a large middle class, without an overbearing rich or a desperate poor.
• Przeworski’s reformulation is the most important qualification in the literature and is routinely misread.
  • His finding separates emergence from survival: democracies do not appear more often as countries get richer, but once they exist, they almost never die above a certain income threshold.
  • This is exogenous democratisation — transitions occur for political reasons at all income levels, and wealth explains durability, not birth.
  • The implication for policy is uncomfortable: growth consolidates democracies but does not manufacture them, so aid aimed at producing transitions through development is aimed at the wrong mechanism.
• Acemoglu and Robinson locate democracy in the distribution of power rather than in income.
  • Elites concede the franchise when the threat of revolution makes concession cheaper than repression, and democratic institutions are the credible commitment that makes the bargain hold.
  • Their institutional argument distinguishes inclusive institutions, which generate broad-based growth, from extractive ones, which can produce growth for a period but not sustain it.
  • Their later work adds the narrow corridor: liberty survives only where state capacity and society’s capacity to constrain the state grow together, which is neither a strong state nor a weak one but a balance.
• The China case keeps this debate live. Rapid growth under one-party rule falsified the strong reading of modernisation theory and supplied a rival development narrative.
  • The comparison drawn most often is China’s growth trajectory against India’s slower path, framed as a question about whether democratic procedure imposes a development penalty.
  • The counter-case is that authoritarian variance is enormous — the same regime type produced both the Chinese boom and the Venezuelan collapse — so the average authoritarian record is worse, not better, even if the best case is spectacular.

Backsliding and backlash: the twin process

• The two terms describe different actors at different levels, and the paradox dissolves once they are defined apart.
  • Backsliding is elite-driven — the erosion of democratic constraints by those already holding office.
  • Backlash is popular mobilisation — and it runs in two opposite directions, which is what makes the word treacherous.

The three features of contemporary backsliding

• It is executive-led. Elected civilians, not generals, do the dismantling, which denies the international system its clearest trigger for a response.
• It is legal in form. Emergency powers, court appointments and retirement ages, media licensing and regulation, electoral commission composition, and NGO funding rules are each individually defensible.
• It is incremental. There is no moment at which an observer can say democracy ended, so condemnation, suspension and sanction always arrive after the fact.
• The combined effect is a detection problem, not merely a will problem. By the time the pattern is unambiguous, the institutions that would reverse it have already been staffed.

Backlash in its two senses

• Populist backlash runs against liberal democracy — against courts, media, expertise, supranational constraint and the political establishment.
  • Norris and Inglehart’s cultural backlash thesis explains it as a reaction by socially conservative cohorts to a long value shift toward post-materialism, secularism and cosmopolitanism.
  • Their argument is that the driver is cultural more than economic — the sense of a silent revolution having been won without their consent — although economic insecurity sharpens it.
• Pro-democracy backlash runs against backsliding incumbents — mass protest in defence of institutions.
  • The last decade has supplied an unusually large number of instances, from Sudan and Belarus to Bangladesh, Nepal, Georgia, Serbia and Madagascar.
  • The protests share a repertoire — leaderless, digitally coordinated, youth-led, triggered by a specific grievance and escalating to demands for the resignation of the government.

Resolving the paradox

• They operate at different levels. Backsliding works through institutions; backlash works in the street. High protest is entirely compatible with institutional erosion — and may be caused by it, since erosion removes the institutional channels that would otherwise absorb grievance.
• They share roots. Economic insecurity and cultural anxiety generate support for strongman leaders in some constituencies and protest against them in others — the same shock, opposite responses, in a polarised society.
• Technology serves both equally. The same platforms that make leaderless mobilisation cheap also make surveillance, targeted disinformation and network mapping cheap for the state.
• Diffusion runs in both directions. Autocratic legal techniques cross borders through advisers, model statutes and imitation; protest repertoires cross borders through social media, which is why unconnected countries display both processes simultaneously.
• The correct description is polarisation, not uniform decline. Democratic institutions are weakening while democratic aspiration remains high, and the two facts are not in tension.

Backsliding works through institutions and backlash works in the street, which is why the same year can produce record erosion and record protest.

The youth protest wave of 2024–26

• A distinct generational wave has run across four continents, triggered by corruption, unemployment and specific policy insults rather than by ideology.
  • Bangladesh in 2024 began over public-service job quotas and ended a fifteen-year government; Kenya forced the withdrawal of a finance bill.
  • Nepal in 2025 was triggered by a social media ban layered on anger at elite children’s displays of wealth, and brought the government down within days.
  • Indonesia erupted over a parliamentary housing allowance many times the minimum wage; Morocco over health spending against stadium spending; Madagascar over water and power failures; Serbia over a fatal station-canopy collapse read as a symptom of construction corruption.
• The wave’s political yield has been uneven, and that is the more instructive half. Where it toppled governments, what followed was an interim administration, a military intermediary, or an election the young did not control.

Measuring democracy: the index industry and its critics

• The recession is known through four measurement projects, and their disagreements are as informative as their agreements.

Project	What it measures	Method	Output
V-Dem	Five high-level principles — electoral, liberal, participatory, deliberative, egalitarian — over hundreds of indicators	Country experts, multiple per indicator, aggregated by a Bayesian item-response model that estimates coder reliability	Continuous indices plus a four-fold regime classification
Freedom House	Political rights and civil liberties in 25 checklist questions	In-house and external analysts, scored 0–4 per question, reviewed by regional panels	Score out of 100; status Free / Partly Free / Not Free
Economist Intelligence Unit	Five categories — electoral process and pluralism, functioning of government, political participation, political culture, civil liberties	Expert assessment plus public opinion survey data	Score out of 10; full / flawed democracy, hybrid / authoritarian regime
Polity	Institutionalised authority patterns — executive recruitment, constraints on the executive, political competition	Historical coding of institutional rules, not rights	Score from −10 to +10, autocracy to democracy
International IDEA	Four attributes — Representation, Rights, Rule of Law, Participation	Composite of many existing indicators, explicitly not a single ranking	Performance bands, no league table

• They disagree because they measure different objects. Polity is institutional and ignores civil liberties almost entirely; Freedom House is rights-centred; V-Dem is disaggregated and explicitly multi-conceptual; the EIU alone incorporates mass attitudes through its political culture category.
  • A state can therefore be stable on Polity while falling on Freedom House — a constitution unchanged while the press is captured.
  • V-Dem’s electoral component moves slowly while its liberal component moves fast, which is why V-Dem detects executive aggrandisement earlier than rights-based measures.
• The current readings agree on direction and differ on level.
  • V-Dem’s 2026 report finds the global level of democracy back at 1978 levels, with 44 countries autocratising against 18 democratising, and roughly three-quarters of the world’s population living in autocracies. Its four-fold count is approximately 31 liberal democracies, 56 electoral democracies, 57 electoral autocracies and 35 closed autocracies.
  • Freedom House’s 2026 edition records a twentieth consecutive year of decline, with 54 countries deteriorating against 35 improving, and only about a fifth of humanity living in Free countries, down from close to half two decades ago.
  • The Economist Intelligence Unit’s 2025 index breaks the pattern, recording the first rise after eight years of decline to a global average of 5.19, with 26 full democracies and 61 authoritarian regimes, and asking whether this is an inflection point or an interlude.
  • International IDEA’s assessment finds more than half of assessed countries declining on at least one attribute over five years, with press freedom falling in more countries than at any point since 1975 and Representation at its lowest level since 2001.
• The disagreement between V-Dem and the EIU in the same season is the most instructive fact in the measurement layer. One finds accelerating collapse; the other finds stabilisation. The gap is a coding gap, not a factual one — attitudinal and institutional measures are moving in opposite directions.

The critique of the indices

• Expert coding imports the coders’ priors. V-Dem’s model corrects for systematic coder bias but cannot correct for a shared conception of what democracy is.
• Composite scores conceal their weights. A single number hides which component moved, so a fall driven by press freedom looks identical to one driven by electoral fraud.
• Perception lags and leads reality asymmetrically. Salient international coverage moves expert perceptions faster than quiet institutional decay does, which biases scores toward loud cases.
• Ordinal rankings invite meaningless comparison. A three-place fall is presented as news when it falls within the measurement error of the underlying index.
• The methodology is rarely reproducible from outside. Underlying coder responses are typically not published in a form that allows an affected government to audit its own score.
• What counts as democratic is itself contested. Indices weight liberal-constitutional features heavily, so states pursuing majoritarian or developmental models score badly by construction rather than by finding.
• The Indian case is the sharpest illustration. V-Dem’s classification of India as an electoral autocracy, Freedom House’s downgrade to Partly Free — where it remains, at 62 out of 100 in the 2026 edition — and the EIU’s persistent flawed democracy label have all been rejected by the Indian government.
  • The official objection is threefold: opaque and unverifiable methodology, small unrepresentative expert panels, and a normative framework that treats Western constitutional practice as the measure.
  • There have been calls within India for an indigenous democracy index and, more concretely, a formal machinery to engage global rankings and contest their inputs.
  • The counter-position, argued by Indian critics themselves, is that a downgrade is evidence to be answered, not a conspiracy to be denounced, and that dismissing a measurement does not repair what it measures.

An index the affected state cannot audit stops being a measurement and becomes an argument about who is entitled to define democracy.

The international architecture around democracy

The United Nations layer

• The UN endorses democracy indirectly, through rights rather than through a democracy mandate. The Charter is silent on regime type and rests on sovereign equality.
• Article 21 of the Universal Declaration of Human Rights is the foundational text: everyone has the right to take part in government, directly or through freely chosen representatives, and the will of the people expressed in periodic and genuine elections is the basis of the authority of government.
  • Article 25 of the International Covenant on Civil and Political Rights converts this into a binding treaty obligation for its parties, with the Human Rights Committee’s General Comment 25 elaborating what genuine elections require.
• The UN Democracy Fund, established in 2005, is the organisation’s only body dedicated to democratisation.
  • It was created by the Secretary-General with India and the United States as its founding and principal donors, and its advisory board seats the largest donors alongside civil society representatives.
  • The large majority of its grants go to local civil society organisations, not to governments — the design decision that gives it independence and limits its scale, with individual grants in the low hundreds of thousands of dollars.
  • Its funding position has been transformed by the withdrawal of United States support, which leaves the fund dependent on a smaller donor group in which India’s role is proportionately larger.
• The Human Rights Council links civil and political rights to democratic governance through the Universal Periodic Review, special procedures and thematic resolutions, though it has no power to declare a government illegitimate.
• The UN’s electoral assistance machinery — technical support to election commissions on request — is the most operationally significant thing the organisation actually does, and it is deliberately framed as assistance to a sovereign, not supervision of one.

Intergovernmental coalitions of democracies

• The Community of Democracies, founded in 2000, is a cross-regional grouping committed to democratic norms, with a permanent secretariat and a governing council.
  • India is a founding member, alongside the United States, the United Kingdom, Japan, Canada, Italy, Poland, South Korea, Chile and others.
  • Its weakness is definitional: membership criteria have never been enforced, and states that backslide are rarely removed, which drains the signal from belonging.
• The Summit for Democracy was the most ambitious recent attempt to convene democracies as a bloc.
  • The first summit was hosted virtually by the United States in December 2021; the second in March 2023 was co-hosted with Costa Rica, the Netherlands, South Korea and Zambia; the third was hosted by South Korea in Seoul in March 2024.
  • India participated at head-of-government level and used the platform to advance its own framing of its democratic tradition rather than to endorse conditionality.
  • The initiative has not been convened since Seoul and has no successor host, ending as an exercise whose principal difficulty was always the guest list — inviting backsliding states drained its credibility, excluding them drained its reach.
• The Open Government Partnership takes the opposite approach — transparency commitments with peer review, no regime-type test — and has proved more durable for exactly that reason.

Regional conditionality

• The European Union operates the strongest democracy conditionality in the world, because it controls something states want.
  • The Copenhagen criteria require stable institutions guaranteeing democracy, the rule of law, human rights and minority protection before accession.
  • Article 2 of the Treaty on European Union states the founding values and Article 7 provides for suspension of membership rights, though the unanimity requirement made it unusable while backsliding governments protected one another.
  • The rule-of-law conditionality regulation attached compliance to budget disbursement, which proved far more effective than Article 7 because it required a qualified majority rather than unanimity.
  • The lesson generalises: conditionality works where the incentive is large and proximate, and conditional on continuing compliance, and fails once membership is secured.
• The Organization of American States has the oldest hard instrument in the Inter-American Democratic Charter of 2001, which permits suspension for an unconstitutional interruption of the democratic order — invoked over Honduras in 2009, and the subject of a bitter and inconclusive dispute over Venezuela.
• The African Union built the most explicit anti-coup regime, treated separately below.
• The Commonwealth operates through the Harare Declaration and its Ministerial Action Group, which has suspended members after coups, though its practice has been criticised as inconsistent.
• ASEAN’s non-interference principle makes regional conditionality nearly impossible; its Five-Point Consensus on Myanmar, and the decision to exclude the junta’s political representatives from summits, is the outer limit of what it has been willing to do.

The non-governmental layer

• Freedom House produces the longest continuous rights-based dataset and is simultaneously the most cited and most contested actor in the field.
• International IDEA is intergovernmental but technical — comparative constitutional and electoral knowledge, turnout data and the Global State of Democracy indices, deliberately avoiding a league table.
• The National Endowment for Democracy is the archetype of the quasi-governmental grant-maker: congressionally funded, independently governed, working through four core institutes tied to the two American parties, business and labour, with the World Movement for Democracy as its network arm.
• Election observation is supplied by the Carter Center, EU and Commonwealth missions and domestic monitoring networks, whose verdicts now compete with those of authoritarian observer missions.
• The structural vulnerability of this layer is that it is externally funded, which is what makes foreign-agent legislation such an efficient weapon against it.

Democracy promotion as statecraft

The instruments and their theory of change

• Aid to political society — party strengthening, parliamentary support, election commission capacity, judicial training — assumes that democracy fails for want of institutional competence.
• Aid to civil society — NGOs, independent media, human rights monitors, election observers — assumes that democracy fails for want of countervailing power, and is the model the UN Democracy Fund and the endowment networks chose.
• Diplomatic and economic conditionality — accession criteria, aid tied to governance benchmarks, targeted sanctions on individuals — assumes that elites respond to incentives.
  • The Millennium Challenge Corporation is the most disciplined version: large compacts awarded only to countries passing published third-party indicators on ruling justly, investing in people and economic freedom, with selection rather than conditions as the mechanism, so that no aid ever has to be withdrawn from a backsliding partner.
• Coercive democratisation — sanctions, no-fly zones, regime change — assumes that the obstacle is a single ruler, and is where the enterprise acquired its reputation.
• Demonstration and attraction — being a functioning democracy that others wish to emulate — is the least direct instrument and, on the evidence of the third wave’s snowballing, among the most powerful.

Instrument	Assumed cause of failure	Strongest case	Characteristic failure
Institutional aid	Lack of capacity	Post-1989 Central Europe	Builds technique, not political will
Civil society aid	Lack of countervailing power	Long-run pressure in closed societies	Invites foreign-agent retaliation
Conditionality	Elite incentives	EU accession; budget conditionality	Loses force once the reward is banked
Selection-based aid	Rewarding good performers	Millennium Challenge compacts	Excludes the hardest cases by design
Military intervention	A single ruler	Very few, contested	Iraq, Libya — state collapse
Demonstration effect	Absence of a credible model	Third-wave snowballing	Slow, indirect, uncontrollable

The record, and the two cases that define it

• The strongest evidence for democracy promotion is the European enlargement. Membership conditionality converted a set of post-communist states into consolidated democracies within fifteen years, because the incentive was enormous and compliance was verified continuously.
• The strongest evidence against it is the coercive branch. Iraq produced sectarian conflict, state collapse and a durable insurgency; Libya, where a civilian-protection mandate was widely perceived to have been used for regime change, produced a decade of factional war.
  • The Libya episode did lasting damage beyond Libya: it hardened Russian and Chinese resistance in the Security Council, which is a substantial part of the explanation for the Council’s paralysis over Syria.
• Levitsky and Way’s finding cuts across the whole enterprise: dense linkage to democratic states predicted transitions far better than external leverage did — meaning the mechanism that actually works is integration, not pressure.

Carothers and the backlash against democracy promotion

• Carothers documented a systematic, coordinated backlash against democracy promotion running for two decades, and it is not confined to the states usually called authoritarian.
  • The instruments are registration requirements, funding restrictions, criminal liability for undeclared foreign support, and expulsion of international NGOs.
  • The rhetoric is sovereigntist rather than anti-democratic — the objection is framed as one to foreign interference, not to democracy, which is what makes it internationally saleable.
  • He also criticised the field’s own transition paradigm: the assumption that any country leaving authoritarianism is in transit to democracy, that transitions follow a fixed sequence, and that elections are the decisive step. Most such countries entered a grey zone and stayed.
• Foreign-agent legislation has diffused globally and is the single most effective anti-promotion technique.
  • Russia’s law, progressively widened, made the label career-ending and the compliance burden fatal.
  • Georgia’s foreign influence legislation, passed over mass protest, precipitated the effective halt of its EU accession and sanctions on its officials.
  • Comparable statutes and registration regimes now operate across Central Asia, parts of Africa, the Middle East and Southeast Asia, and closing civic space is the most consistently reported trend in the field.
• The Global South critique is broader than the sovereigntist one, and is not answered by better technique.
  • Selectivity is the core charge — democracy is demanded of the weak and waived for the strategically useful, and the record of tolerating friendly autocracies is long.
  • Neo-imperialism is the second — that promotion has repeatedly served as the vocabulary of intervention, from Iraq to Libya.
  • Sequencing is the third — that liberalisation imposed on states without capacity produces instability, and that the promoter bears none of the cost of the resulting collapse.

The rupture of 2025 and after

• The infrastructure of Western democracy assistance was dismantled within a single year, and the field has no precedent for a shock of this kind.
  • The United States Agency for International Development ceased to exist as an independent agency in July 2025, with surviving programmes absorbed into the State Department and the great majority terminated.
  • The National Endowment for Democracy’s congressionally appropriated funds were withheld, forcing programme suspensions worldwide and litigation to compel disbursement before partial restoration.
  • United States support for the UN Democracy Fund was ended, removing its largest donor since inception.
• The consequences run in three directions.
  • Grantees lost their operating base: independent media outlets, election monitors and human rights documentation projects across dozens of countries had been sustained by these flows.
  • The vocabulary lost its sponsor: with the Summit for Democracy dormant and the aid architecture gone, democracy promotion has ceased to be an organising principle of American foreign policy.
  • The sovereigntist argument was vindicated in an unexpected way — the abrupt withdrawal demonstrated exactly the dependency that recipients had been warned about.
• The vacuum is not being filled symmetrically. European instruments remain, but they are conditionality-based and therefore concentrated on the EU’s own neighbourhood; no other donor operates a global civil-society network at comparable scale.

The supply side: autocracy promotion and authoritarian diffusion

• The recession has a supply side that the literature was slow to study. Autocracies do not merely resist democracy promotion; several now actively export governance models, techniques and protection.

China

• China’s export is a model of governance, not an ideology, which is precisely what makes it attractive and what limits it.
  • The claim is performance legitimacy: that a disciplined single-party state delivers growth, infrastructure and poverty reduction faster than a contested one.
  • Party-to-party training programmes transmit techniques of cadre management, propaganda work and internet governance to ruling parties across Africa, Southeast Asia and Central Asia.
  • Digital infrastructure exports — surveillance systems, safe-city projects, telecommunications backbones — supply capabilities that are regime-neutral in principle and regime-strengthening in practice.
  • Belt and Road financing carries no governance conditionality at all, which is a positive selling point for recipients tired of benchmarks.
• Beijing’s normative project has become more explicit. Alongside its earlier initiatives on development, security and civilisation, China has advanced a Global Governance Initiative built on sovereign equality, non-interference, multilateralism grounded in the UN Charter, and an explicit rejection of any single country’s model as universal.
  • The framing is a defence of pluralism among political systems — a claim that development paths are for each people to choose — which reframes non-democracy as diversity rather than deficiency.
  • Its practical limitation is that it makes no universal claim on individuals. It offers other governments a defence, not other peoples a vision, which is why it does not function as an ideological competitor in Fukuyama’s sense.

Russia

• Russia’s toolkit is disruptive, not constructive — the aim is to lower the prestige of democracy, not to raise that of any alternative.
  • Electoral interference and disinformation target the credibility of the process itself, so that the message is not “our system is better” but “yours is no different”.
  • Support for incumbents under pressure — political technologists, security assistance, energy pricing, and in the Sahel and elsewhere deployable armed contractors — has repeatedly kept threatened governments in place.
  • Foreign-agent legislation was pioneered as a model statute and copied widely.
  • Regional institutions and the vocabulary of “sovereign democracy” supply the doctrinal cover.

Authoritarian learning and diffusion

• Autocratic techniques travel through imitation, advice and institutional networks, in the same way protest repertoires do.
  • Legal templates diffuse fastest — foreign-agent statutes, NGO registration regimes, “fake news” and cybercrime laws, and the technique of amending judicial retirement ages to force turnover.
  • Regional organisations of authoritarian states offer election observation missions that certify votes Western observers condemn, breaking the monopoly on legitimation.
  • Mutual insurance is the deepest form: regimes support each other against domestic challenge, so that no single incumbent’s fall becomes a demonstration effect for the others.
• The two enterprises are not mirror images, and the comparison is instructive.

Dimension	Democracy promotion	Autocracy promotion
Target	Civil society and institutions, often against the government	Incumbent governments, always with them
Claim	Universal — this is right for all peoples	Pluralist — each people chooses its own path
Instrument	Grants, conditionality, observation, occasionally force	Unconditional finance, technology, security assistance, certification
Vulnerability	Framed as interference; dependent on donor politics	Dependent on the patron’s own performance
Effect on norms	Raises the cost of visible repression	Lowers the cost by removing the audience

The African anti-coup norm under stress

• The African Union built the most explicit anti-coup regime in the world, and it is the one under the greatest strain.
  • The Lomé Declaration of 2000 first defined unconstitutional change of government and provided for suspension, but was non-binding, narrowly defined and weakly enforced.
  • The African Charter on Democracy, Elections and Governance, adopted in 2007 and in force since 2012, made the commitments binding for its parties, extended the definition to cover constitutional coups — amendments to extend terms or entrench incumbents — and barred coup perpetrators from contesting the elections that follow.
  • Its central weakness is coverage: it binds only the two-thirds of member states that have ratified it, and the Peace and Security Council’s practice has been to suspend after the fact rather than to prevent.
• The coup belt of 2020–23 tested the norm to destruction.
  • Mali in 2020 and again in 2021, Guinea in 2021, Sudan in 2021, Burkina Faso twice in 2022, Niger in 2023 and Gabon in 2023 produced the densest run of successful coups since the 1970s.
  • The AU suspended each, and the suspensions changed nothing, because the cost of exclusion was lower than the cost of leaving power and alternative partners were available.
  • The common preconditions were insurgency the civilian government could not contain, contested elections, and constitutional manipulation by incumbents — which is why the AU’s inclusion of constitutional coups matters analytically even though it is unenforced.
• The rupture with ECOWAS is the most consequential institutional break in a generation.
  • ECOWAS’s earlier record was the strongest in Africa — mediation, sanctions and, in The Gambia in 2017, the credible threat of force to remove a defeated incumbent.
  • Its sanctions and threatened intervention against Niger in 2023 failed, and the three sanctioned states responded by forming the Alliance of Sahel States, first as a security pact and then as a confederation.
  • Mali, Burkina Faso and Niger completed their withdrawal from ECOWAS in January 2025, taking with them a large share of the bloc’s territory, and have proceeded to a common passport and joint force — the first successful secession from a major African regional organisation.
  • The lesson is that regional conditionality collapses when exit is available, and the availability of alternative external partners is what made exit survivable.
• Two recent episodes show the norm’s remaining edge and its limits.
  • Madagascar in October 2025 ran through mass youth protest over water and power failures, an elite army unit defecting, the president fleeing, and a colonel installed by the constitutional court.
    • The AU determined that an unconstitutional change of government had occurred and suspended the country; the regional southern bloc, reading the transfer as constitutional, did not.
    • Analysts named the hybrid form a “coupvolution” — popular revolution consummated by military seizure — and it is the hardest case for a norm written against generals rather than crowds.
  • Guinea-Bissau in November 2025 saw a military takeover days after a disputed election, condemned and sanctioned by immediate AU suspension — a real cost only where the junta wants readmission.
• The unresolved question is whether the norm should bend for popularly supported takeovers. Bending it invites every ambitious officer to manufacture a crowd; refusing to bend it puts the AU on the side of the incumbent the crowd rejected.

Case studies

Third-wave transitions that consolidated

• Spain after Franco is the model of pacted transition — negotiated between reformers inside the regime and moderates in the opposition, with the Moncloa Pacts binding parties and unions to economic terms and an amnesty burying the civil war’s accounts.
  • Consolidation was sealed by the failed coup attempt of 1981, after which the monarchy’s public defence of the constitution removed the army from politics, and by accession to the European Community.
• South Korea in 1987 transitioned under mass pressure rather than by elite pact, with the June Democratic Uprising forcing direct presidential elections, followed by the prosecution of former presidents and a durable pattern of alternation.
• Eastern Europe after 1989 consolidated fastest where the European Union and NATO membership incentive was credible — Poland, the Czech lands, Hungary, the Baltic states — and slowest or not at all where it was absent.
  • Russia’s trajectory is the counter-case: elections without accountable institutions, a constitution with an overwhelming presidency, and a reversion completed within a decade.
  • The divergence within the same starting cohort is the strongest single piece of evidence for conditionality as the effective external instrument.

The Arab Spring and what followed

• Tunisia was the sole transition to produce a democratic constitution, competitive elections and alternation, and it has since been reversed from within.
  • The president’s suspension of parliament in July 2021, rule by decree, and a 2022 constitution establishing a hyper-presidential system dismantled the settlement; opposition leaders have since been convicted in mass trials.
  • By 2026 the country had passed five years of emergency rule, with sustained protest returning to the streets over economic conditions and detentions — a textbook autocratic-legalist reversal of the region’s only success.
• Egypt elected a government in 2012 and reverted to military-backed rule in 2013, illustrating both the fragility of a transition without a pact and the ease with which a majoritarian election can end a transition rather than complete it.
• Libya and Syria demonstrated the cost of transition attempts where state institutions are weak: state collapse, factional war and external intervention.
  • Syria’s revolution outlasted the region’s: the Assad regime fell in December 2024, thirteen years after the first protests, and a transitional administration under a constitutional declaration has followed.
  • Its first post-Assad legislative process, run through indirect electoral colleges with a substantial bloc of appointed seats, is a managed opening rather than a democratic transition, and has been criticised for the thin representation of women and minority communities.
  • The comparative point is that the fall of a dictator is the beginning of the problem, not its solution — the classic finding of the transitions literature, restated.

Backsliding from within

• Turkey is the paradigm of executive aggrandisement completed. Media ownership was consolidated, the judiciary purged after the 2016 coup attempt, and the 2017 referendum replaced the parliamentary system with an executive presidency.
  • The pattern continues through the arrest of the Istanbul mayor and principal opposition figure in March 2025, the largest protests in over a decade, and prosecutions reaching into the main opposition party’s leadership — competition that remains real, and a field that is not level.
• Hungary was the laboratory of autocratic legalism: a supermajority used to write a new Fundamental Law, pack and bypass the constitutional court, redraw constituencies, capture public and then private media, and subject universities and NGOs to hostile legislation.
  • Its ending is the most important single fact of the current cycle. At the April 2026 election, on a turnout near eighty per cent, the opposition Tisza party defeated Fidesz and Viktor Orbán conceded after sixteen years in power.
  • The case now cuts both ways: it shows how far a democracy can be degraded and that a sufficiently large opposition coalition can still win on a tilted field, which is a direct rebuttal to the claim that backsliding is irreversible.
• Venezuela shows erosion running past the point of return — a constituent assembly used to displace the elected legislature, courts and electoral authority captured, opposition candidates disqualified, and the 2024 presidential result declared without publishing precinct tallies while the opposition published its own.
  • The regime’s end came not through domestic institutions but through external force — the capture of Nicolás Maduro by a United States operation in January 2026, followed by the installation of his vice-president as interim head of state.
  • That outcome is analytically awkward for everyone. It rescued nothing constitutionally: a regime removed by foreign action is not a democratic transition, and it revives the sovereignty objection to coercive democratisation in its sharpest form.
• Poland’s reversal is the cleaner counter-example. Judicial capture through the constitutional tribunal and the disciplinary chamber was contested by the EU through infringement proceedings and budget conditionality, and reversed at the ballot box in 2023 on record turnout — with the subsequent difficulty of unwinding captured institutions demonstrating that restoration is slower than capture.

Pakistan: democracy under a permanent tutelary power

• Pakistan’s problem is not the absence of elections but the presence of a supervising institution. Four periods of direct military rule are punctuated by civilian governments that the army has removed, constrained or replaced.
  • No elected prime minister has completed a full term, and the mechanism of removal has shifted from coup to judicial disqualification and engineered defections.
  • The doctrine of necessity supplied judicial validation for takeovers, tying the superior judiciary into the tutelary structure from the beginning.
• The 2024 general election was contested under systematic constraint — the largest party’s leader imprisoned, its electoral symbol withdrawn so its candidates ran as independents, and a coalition formed against the plurality of seats those independents won.
• The constitutional consolidation since is the substantive development. The 26th Amendment in 2024 moved the appointment of senior judges to a parliamentary committee and created a separate constitutional bench, curtailing the Supreme Court’s suo motu reach.
  • The 27th Amendment in November 2025 went further, restructuring the military’s command under a Chief of Defence Forces, granting constitutional immunity to certain office-holders, and further limiting the courts.
  • The significance is that the army’s dominance is now written into the constitution rather than exercised against it — Pakistan has moved from recurrent coups to autocratic legalism, which is a change of technique, not of substance.
• Pakistan registered one of the steepest declines in the most recent global index — a hybrid regime by every classification in use.

Nepal: participation outrunning institutions

• Nepal compressed a monarchy, an insurgency, a peace process and a federal republic into two decades, and the institutions never caught up with the mobilisation.
  • The People’s Movement of 2006 ended royal rule; the Comprehensive Peace Agreement brought the Maoists into politics; the monarchy was abolished in 2008; the constitution of 2015 established a secular federal republic amid protest in the Madhes.
  • What followed was chronic instability — governments lasting months, coalitions rearranged among the same three party leaderships, and two attempted dissolutions of parliament reversed by the Supreme Court.
• The disillusion produced a royalist revival. Pro-monarchy protests in Kathmandu in March 2025 demanded restoration of the monarchy and of Hindu statehood, turning violent and fatal — nostalgia for order in a democracy that had delivered turnover without governance.
• The rupture came in September 2025. A government ban on twenty-six social media platforms, imposed onto existing anger about corruption and the visible wealth of politicians’ children, triggered protests that burned the parliament building and killed dozens.
  • The prime minister resigned within a day; Sushila Karki, a former chief justice, was appointed interim prime minister — Nepal’s first woman to hold the office — parliament was dissolved, and elections were called.
  • At the election of 5 March 2026 the Rastriya Swatantra Party won an outright majority, the Nepali Congress and the CPN (UML) suffered the worst results in their histories, both major former prime ministers lost their seats, and the pro-monarchy party won only a handful of seats.
  • The corrective to the March 2025 reading is precise: the royalist mood was a symptom of anti-incumbency, not a demand for restoration. Given a credible alternative, voters chose a new party, not an old crown.
• Huntington’s framework explains the sequence better than any other. Rapid political mobilisation ran far ahead of institutionalisation; the result was decay, then rupture, and the open question is whether a landslide for a new party builds institutions or merely replaces the personnel occupying weak ones.

Myanmar: the reversal completed and legalised

• The decade of quasi-democracy from 2011 was always constitutionally bounded. The 2008 constitution reserved a quarter of legislative seats for the military, placed the interior, defence and border ministries under army control, and set an amendment threshold above 75 per cent — an unamendable veto.
• The coup of February 2021 followed the National League for Democracy’s landslide, and the military’s fraud claim was rejected by observers.
  • It provoked a nationwide armed resistance that, unlike earlier crackdowns, combined ethnic armed organisations with a new generation of urban fighters and cost the army control of substantial territory.
  • ASEAN’s Five-Point Consensus went unimplemented, and the regional response settled into excluding the junta’s political representatives from summits.
• The junta’s answer was an election designed to legalise the coup. Polls were held in phases from late December 2025 into early 2026, with the NLD dissolved for failing to re-register, its leadership imprisoned, and voting cancelled in a large number of constituencies for insecurity.
  • The military-aligned party took an overwhelming share of a heavily curated contest, and the junta chief left military command to become eligible for the presidency of a nominally civilian government.
  • Observers across the spectrum described the exercise as lacking credibility; it is a textbook illustration of the difference between holding elections and being a democracy.

Small-state resilience and democratic breakthroughs

• Ghana and Senegal have repeatedly transferred power between parties peacefully, including from incumbents who lost — Senegal’s 2024 election producing an opposition victory weeks after an attempt to postpone the vote was struck down by its constitutional council.
• Taiwan completed the transition from single-party rule to multiple peaceful alternations while under external military pressure, and South Korea demonstrated in December 2024 that its institutions could reverse an attempted martial-law decree within hours, followed by impeachment and a fresh election.
• Costa Rica abolished its army in 1948 and invested the savings in education and health, remaining a stable democracy through a period in which most of its neighbours were not.
• Breakthroughs continue and are systematically under-reported. Malaysia’s 2018 election ended six decades of single-coalition rule by peaceful transfer; Zambia in 2021 and Czechia saw incumbents concede; Brazil’s institutions absorbed an insurrectionary attempt on its capital and continued.
  • Sri Lanka in 2024 produced the most complete rejection of an entrenched political class in South Asia, electing a president and legislature from outside the established parties — among the largest improvements the 2026 indices record, alongside the finding that reform has since slowed.
  • Bolivia’s 2025 election ended two decades of single-party dominance, and Gabon’s return to civilian-form government produced the single largest score improvement in the most recent index — while remaining a former coup leader’s government.
• The pattern is consistent: what distinguishes these states is not wealth, culture or geography but institutions that raise the cost of capture — election commissions that cannot be staffed by the incumbent, courts with dispersed appointment, and parties that treat defeat as survivable.

The challenges democracies now face

Information and the epistemic base

• Misinformation and foreign electoral interference erode the informed consent on which the vote depends, because the problem is not persuasion but the destruction of a shared factual baseline.
• Generative synthetic media has industrialised the cost side. Fabricated audio and video of candidates, produced cheaply and distributed at scale in the final days of a campaign, arrive faster than any verification process can respond.
• Platform incentives reward outrage over accuracy, so polarising content is amplified structurally rather than by anyone’s decision.
• The response itself is a hazard. Laws against “fake news” and online harms are the most commonly abused instrument in the backsliding toolkit, so the cure and the disease share an instrument.

Institutions and capacity

• In young democracies, judiciaries, election commissions and audit bodies are weak against patronage and corruption, and repeated failure produces democratic fatigue and open nostalgia for strong rulers.
• The sequencing dispute is unresolved. Whether rule of law and state capacity should precede mass competition is a genuine North–South argument, with Zakaria’s warning on one side and the objection that the sequence is a licence for indefinite postponement on the other.
• Capture proceeds through appointments, not through statutes. Where a single actor controls appointments to the courts, the election commission and the audit office, no constitutional text prevents the outcome.

Majority rule against minority rights

• Majoritarianism is insensitivity to minorities expressed through democratic procedure, and it is the failure mode that indices detect last, because nothing formally illegal occurs.
  • Tocqueville’s tyranny of the majority and Madison’s warning about factions are the classical statements; Dahl noted the empirical caution that minorities have not demonstrably fared better under non-democratic alternatives.
  • Sri Lanka’s long post-independence trajectory and Hungary’s treatment of migrants and civil society are the cases usually cited, and both were produced by elected governments acting within their mandates.

Structural and external pressures

• Political inequality follows economic inequality, which is why the concentration of campaign finance and of control over information is a first-order democratic problem rather than a regulatory detail.

“If citizens are unequal in economic resources… they are likely to be unequal in political resources; and political equality will be impossible to achieve.” — Dahl

• External promotion is criticised as sovereignty-violating and selectively applied, and after 2025 it is also criticised as unreliable, which is a new and more damaging charge.
• Authoritarian powers have grown bolder in promoting their model, so backsliding incumbents now have alternative finance, alternative technology and alternative election observers.
• The standard criticisms of democracy itself remain in the field — policy gridlock in divided societies, short electoral horizons against long problems like climate, and susceptibility to demagoguery.
  • The reply that has held up best is that the remedy for the defects of democracy is more democracy — transparency, civic education, inclusive deliberation and institutional repair — because every alternative on offer removes the mechanism by which errors are corrected without violence.

“Democracy is the worst form of Government except for all those other forms that have been tried from time to time.” — Churchill

India’s position

India occupies an unusual place in this concern: it is simultaneously the most-cited evidence that democracy can work in a poor and plural society, a state that refuses on principle to promote democracy abroad, and a state whose own ratings have fallen. Those three facts are connected, and the connection is what makes the Indian position coherent rather than contradictory.

The record as the argument

• India is the central case in the debate about whether democracy suits poor and diverse societies, because it adopted universal adult franchise at independence, at very low income, with mass illiteracy and unresolved linguistic, religious and caste divisions.
  • Every modernisation-theory prediction said this should have failed, and the survival of competitive elections, alternation and civilian control of the military is the single largest anomaly the literature has to accommodate.
  • The record is unbroken except for the Emergency of 1975–77, and the significant fact about the Emergency is that it was ended by an election that the government lost and accepted.
• India’s development bargain differs from the East Asian one in kind, not degree. It attempted redistribution, industrialisation and social transformation through consent rather than compulsion, which is slower and is the substance of the “China versus India” comparison.
• The institutional achievements most relevant internationally are the ones India’s own diplomacy foregrounds: an independent Election Commission conducting elections at unmatched scale, a Supreme Court with an expansive rights jurisdiction, and constitutionalised local self-government through the 73rd and 74th Amendments with reserved seats for women and marginalised groups.

Non-interference as doctrine

• India has never operated a democracy-promotion policy, and the abstention is doctrinal rather than incidental.
  • The foundation is the Panchsheel formulation of 1954 — mutual respect for sovereignty and territorial integrity, non-aggression, non-interference in internal affairs, equality and mutual benefit, and peaceful coexistence.
  • Nehru’s anti-colonial reading held that the right of a people to determine their own political form is precisely what colonialism denied, so imposing a form of government reproduces the wrong even when the form is democratic.
  • Non-alignment required India to deal with states across the ideological spectrum, and any regime-type test would have destroyed that position.
• India does not sanction states for being undemocratic and does not fund opposition movements abroad. It works with the government that exists.
  • India’s engagement with the Taliban administration is the sharpest recent illustration — a technical mission upgraded, an acting foreign minister received in Delhi in October 2025, and the Kabul embassy restored, without any grant of formal recognition.
  • The rationale offered is that influence requires presence, and that abstention cedes the field to others — the classic realist answer to a normative objection.
• The critique is that non-interference is indistinguishable from indifference, and that a state claiming democratic exemplarity while treating regime type as irrelevant abroad is holding two positions at once.
  • The Indian answer is that the two are consistent because the offer is demonstration and capacity, not conditionality: India will help any state run an election and will not tell it what result to reach.

What India does support

• India was a founding contributor to the UN Democracy Fund and has been its second-largest donor, with cumulative contributions above thirty million dollars — a fact India cites precisely because the fund’s money goes to civil society and not to governments, which is as close to democracy assistance as the doctrine permits.
  • With United States support to the fund now ended, India’s proportionate weight among the remaining donors has risen sharply, which is a real if unsought opening.
• India is a founding member of the Community of Democracies and sits on its governing council.
• India’s operational contribution is technical and bilateral. The Election Commission has trained election officials from a long list of countries through its international training institute and has supplied advice and materials to neighbours conducting difficult polls.
• India is a member of International IDEA, and its development assistance — lines of credit, capacity building, the technical and economic cooperation programme — is deliberately unconditional, which is presented as a contrast with Western practice rather than as an absence of values.

The “Mother of Democracy” framing

• The claim is that consultative and elective practices in the subcontinent long predate the West, and it was elaborated as a formal state project around India’s G20 presidency, with an exhibition, a multilingual publication and a curated set of historical examples.
  • The evidence advanced includes the sabha and samiti of the Vedic corpus, the ganas and sanghas of the sixth century BCE, the deliberative procedure of Buddhist monastic assemblies, and the Uthiramerur inscriptions recording the Chola-era kudavolai ballot for village assembly membership.
  • The political function is to detach democracy from a colonial genealogy: if popular consultation is indigenous, then democracy is not a Western import and its Indian practice is not a derivative.
• The scholarly objections are specific and worth stating rather than waving at.
  • These were assemblies of restricted and hereditary membership, not systems of equal citizenship; the Uthiramerur qualifications were bounded by property, learning and lineage.
  • The line of descent to the modern republic is discontinuous. The Constituent Assembly worked from comparative constitutional material and from the colonial legislative record, not from the sabha.
  • The strongest version of India’s case does not need the ancient claim at all: the decision to grant universal adult franchise at independence, against every expert prediction, is the more remarkable achievement and rests on nothing but its own audacity.

India and the democracy summits and indices

• India participated in all three Summits for Democracy at leader level, using the platform to present its own record and the “mother of democracy” framing rather than to endorse any conditionality agenda — participation as a democracy, not enlistment in a bloc.
  • India’s discomfort with the format was structural: a summit that sorts states into democracies and non-democracies is a bloc-building exercise, and bloc-building against China and Russia is not something India will join on normative grounds.
• India’s response to the ratings agencies has been consistent rejection on methodological grounds — small expert panels, unpublished underlying data, and a framework India argues does not fit a society of its scale and diversity.
  • The government’s position is that its democracy is robust and self-correcting, and that its internal questions will be resolved through its own constitutional processes.
  • The most recent readings are nonetheless the material India must answer: Partly Free at 62 on Freedom House’s 2026 scale, an electoral-autocracy classification from V-Dem, and one of the six steepest single-year declines in the most recent EIU index, attributed to civil liberties and functioning of government.
  • Indian critics of the government make the internalist argument — that the correct response to a downgrade is to fix what it points at, since press freedom, institutional autonomy and the treatment of dissent are Indian constitutional questions before they are international ones.

The neighbourhood record

• India’s practice in South Asia has been to support processes rather than particular regimes, and the record is more interventionist in effect than the doctrine suggests.

Case	India’s role	Doctrinal reading
Nepal 2006–08	Facilitated the 12-point understanding between parties and Maoists that ended royal rule	Process facilitation, not regime choice
Maldives 2018	Backed the electoral outcome and constitutional order after an emergency and a judicial crisis	Defence of a result, not of a party
Sri Lanka	Engaged every government including the 2024 outsider victory; pressed the 13th Amendment on devolution	Minority accommodation, framed as bilateral
Bangladesh 2024–26	Adjusted to the fall of a long-standing partner; engaged the interim administration and then the elected government of 2026	Continuity of state relations over party ties
Myanmar	Kept working relations with the junta while funding development in border states	Presence over principle, contested
Afghanistan	Restored the embassy and received the acting foreign minister without recognition	Engagement without endorsement

• Bangladesh is the hardest test and the clearest illustration. After the fall of a government with which India had built a deep security relationship, India dealt with an interim administration it had not chosen, and then with the party it had historically been most distant from, following the February 2026 election.
  • The subsequent friction — over the trial of the exiled former prime minister, over minority security, and over the pace of restoring high-level visits — is the price of having been identified with an incumbent rather than with a process.
  • The lesson India’s own commentators draw is that relationships built with parties rather than institutions do not survive alternation, which is an argument for process neutrality made on realist grounds.

The record as the instrument

• India’s implicit theory of democracy promotion is demonstration: that a poor, plural, post-colonial state sustaining competitive elections at continental scale is the most powerful argument available, and that the argument is weakened by every domestic erosion and strengthened by every domestic repair.
  • This makes India’s international position unusually dependent on its own performance, in a way that a conditionality-based promoter’s is not.
  • It also gives India a distinctive standing in the current moment: as Western democracy assistance withdraws and the field is left to sovereigntist objection and unconditional authoritarian finance, a large non-Western democracy that neither preaches nor sanctions is the only actor positioned to make the case without triggering the backlash.

India’s answer to democracy promotion has always been to be the demonstration rather than to fund it — which makes its own record its foreign policy.

The state of play in 2026

• The direction of travel is still negative, but the last two years broke the assumption that backsliding is one-way.
  • Reversals happened: Orbán’s defeat after sixteen years, Poland’s continuing restoration, Sri Lanka’s rejection of an entrenched class, Bolivia’s alternation, and South Korea defeating a martial-law decree within hours.
  • Erosion continued: Turkey’s prosecution of its leading opposition figure, Tunisia’s fifth year of emergency rule, Georgia’s halted accession, and the first entry of established Western democracies onto the autocratising lists.
• The most consequential change is on the promotion side, not the recession side — the dismantling of United States democracy assistance, the dormancy of the Summit for Democracy and the end of American funding to the UN Democracy Fund removed the field’s dominant patron within a single year.
• Elections have closed transitions as often as they have opened them: Myanmar’s phased poll legalised a coup, Bangladesh’s held a real contest with the former ruling party banned, Syria’s produced a partly appointed assembly.
• The regional norms are diverging — the European Union found an instrument that works in budget conditionality, while the African Union’s anti-coup norm survived on paper and lost three ECOWAS members to a rival bloc.
• The measurement projects now disagree about the trend, one recording collapse to late-1970s levels and another the first improvement in nine years, which is itself the most honest summary of the moment.

Conclusion

The concern has inverted within a generation. The literature of the 1990s was about how transitions happen and how democracies consolidate; the literature of the 2020s is about how democracies are hollowed out by the people they elect. What the evidence does not show is a collapse of demand: publics keep filling streets at real cost to defend or demand accountable government, and no rival ideology makes a universal claim on individuals.

• The paradox of the twin process is only apparent. Institutions weaken while aspiration holds, because the two are produced by different actors operating at different levels.
• The practical implication follows directly. The resilience of democracy now depends far less on its popularity, which is not in doubt, than on whether institutions can be made expensive to capture — dispersed appointments, insulated electoral administration, and parties that treat defeat as survivable.
• For India the implication is closer to home. Its influence on this global concern is exercised through its own constitutional practice, which means that the domestic and the international questions are, unusually, the same question.