Political Parties and Party Systems in Advanced Industrial and Developing Societies

Political parties are the one institution that no modern political system, democratic or authoritarian, has managed to do without. Yet the comparative study of parties has arrived at an awkward conclusion: in the societies that invented the mass party, parties are losing members, loyalty and legitimacy, while in the societies that imported the form, parties are frequently personal vehicles that hold elections without organising anything. Parties are simultaneously indispensable and failing, and the two halves of that sentence describe different halves of the world.

Why parties exist at all: the problem they solve

PSIR Optional Notes {PDF}

Complete Notes Package (Paper I & II) | 1500+ Pages | Instant Download

Discounted Price | Buy Now @ ₹1000 →

• Political participation means little unless dispersed individual preferences can be converted into collective political force, and that conversion needs an organisation.
  • Gabriel Almond and Bingham Powell located this precisely: the political system’s input functions include interest articulation and interest aggregation, and the party is the characteristic structure of aggregation.
  • Interest articulation is the raising of a demand; interest aggregation is the bundling of many demands into a small number of governing alternatives a voter can actually choose between.
  • Without aggregation an electorate faces an unmanageable menu; parties reduce a society’s infinite disagreements to a handful of packages.
• The academic study of parties has its own name — stasiology, a coinage associated with Duverger — and it became a specialised subfield precisely because parties refused to behave the way constitutional theory expected.
• Parties are extra-constitutional in origin: almost no founding constitution mentions them, yet within a generation of any franchise extension they become the operating system of the state.

The anti-party tradition

• Hostility to party is older than the defence of it, and it has never fully died.
  • Rousseau held that partial associations distort the general will, because a citizen voting inside a faction votes the faction’s will rather than the community’s.
  • Madison, in Federalist 10, treated faction as a disease of popular government, to be controlled by the size and structure of the republic rather than cured.
  • George Washington’s farewell address warned against “the baneful effects of the spirit of party.”
  • In India the tradition runs through Gandhi, who proposed disbanding the Congress as a party after independence, M. N. Roy‘s radical humanism with its party-less organised democracy, and Jayaprakash Narayan‘s partyless democracy and Reconstruction of Indian Polity.
• The persistence of this tradition matters analytically: every contemporary complaint about parties — that they are oligarchic, self-serving, and interposed between citizen and state — is a restatement of an eighteenth-century objection that party defenders never fully answered.

The four theoretical lenses on the political party

The normative approach: Burke

• Edmund Burke produced the first serious defence of party, and did so by distinguishing it sharply from faction.

“Party is a body of men united, for promoting by their joint endeavours the national interest, upon some particular principle in which they are all agreed.” — Edmund Burke

• Three elements are doing work in that definition, and each is a claim about legitimacy.
  • A shared principle — not a shared interest, which is what makes it a party rather than a faction. A faction pursues private advantage; a party pursues a public conception of the good.
  • Joint endeavour — collective and sustained action, not an occasional alignment of convenience.
  • The national interest — the object is the whole, not the part, however partial the route to it.
• Burke’s defence connects directly to his theory of representation, given in the Speech to the Electors of Bristol: the representative owes the constituency his judgement, not obedience, and betrays it by sacrificing judgement to opinion.
  • By the same logic a party should shape public opinion through reason and constitutional responsibility, not merely register it.
• The lasting contribution is the legitimation of organised disagreement. Before Burke, opposition was suspected of disloyalty; after him, an organised opposition around a rival principle is part of constitutional government rather than a threat to it.

The behavioural approach: parties as powerhouses

• Behavioural scholars dropped the normative question entirely and asked what parties actually do, concluding that they are instruments for acquiring and exercising power.
  • Joseph Schumpeter made the position blunt: a party is “a group whose members propose to act in concert in the competitive struggle for political power”, and democracy is the institutional arrangement by which such groups compete for the people’s vote.
  • Anthony Downs, in An Economic Theory of Democracy, formalised it: parties formulate policies in order to win elections, rather than winning elections in order to formulate policies. The median voter theorem follows — under single-peaked preferences and two competitors, both converge on the centre.
  • Sartori supplied the minimal working definition that avoids both moralising and cynicism: a party is any political group that presents at elections, and is capable of placing through elections, candidates for public office.
• The distance between Burke and Downs is the whole argument in miniature: is a party a vehicle of principle that must win, or a vehicle of winning that must have a principle?

The Marxist–Leninist approach

• Marx himself never developed a theory of party, and was suspicious of the form — a permanent organisation implies hierarchy, leaders and led, which sits badly with the classless society it claims to be building.
• Lenin supplied what Marx had not, in What Is To Be Done?, and every later communist organisation descends from it.
  • Workers cannot generate revolutionary consciousness on their own. Left to themselves they achieve only trade-union consciousness — better wages and shorter hours within capitalism.
  • The party is therefore the vanguard — a body of professional revolutionaries who carry theory into the class from outside it.
  • Lenin’s distinction between the communist party and the trade union follows: unions are structurally inside the bourgeois system and can win concessions; only the party can bring workers to power.
  • Under Tsarist repression the party had to be small, disciplined and clandestine, organised in cells and pyramidal in structure — a design forced by conditions but retained after them.
• Democratic centralism is the organising principle, and it is routinely misdescribed.
  • Freedom of discussion before a decision; unity of action after it. Demands travel upward and are debated; once the leadership decides, the decision is binding at every level and dissent ends.
  • In practice the centralist half consumed the democratic half — the ban on factions at the Tenth Party Congress in 1921 removed the discussion phase, leaving only the discipline.
• Gramsci later softened the model for societies with dense civil societies: the party as the “Modern Prince”, whose task in the West is not the insurrectionary war of manoeuvre but the long war of position for hegemony.

The organisational–sociological approach: Michels and Ostrogorski

• Moisei Ostrogorski, in Democracy and the Organization of Political Parties, made the first systematic charge: the permanent party machine — the caucus — turns the citizen into a passive follower and the party into an end in itself.
• Robert Michels turned that observation into a law, in Political Parties (1911).
  • His method was a hard-case test: he studied the German Social Democratic Party, the most internally democratic, ideologically egalitarian mass party in Europe — precisely because if oligarchy appeared there, it would appear anywhere.
  • It appeared. Power sat with a permanent leadership that controlled information, finance and communication; the mass membership ratified but never decided.
• “Who says organisation, says oligarchy.” The mechanism has three legs:
  • Technical necessity — any large organisation needs specialisation, delegation and speed, none of which mass assemblies deliver.
  • Psychological disposition — the mass is apathetic, grateful and deferential, and prefers to be led.
  • Leadership self-interest — leaders acquire skills, networks and status that make them irreplaceable, and act to stay so.
• The comparative payoff is the one Michels intended: ideology does not predict internal structure. A revolutionary socialist party and a conservative party are organised alike, which is why revolutionary parties in power reproduce hierarchy rather than dissolving it.
• The standard criticism is that “iron law” overstates it. Oligarchy is a strong tendency that internal competition, factional pluralism, primaries and judicial regulation can restrain — Lipset, Trow and Coleman‘s study of the International Typographical Union, with its genuine two-party internal life, is the exception that establishes the conditions.

Michels’s finding is uncomfortable precisely because it is comparative: it predicts oligarchy inside parties regardless of what those parties believe.

Types of party: an evolutionary sequence

Each party type is a solution to the problem the previous type could not handle, and each solution creates the problem the next type addresses. Reading the typology as a sequence rather than a list is what makes it analytically useful.

• Cadre or elite parties — the original form, belonging to the restricted franchise era.
  • Loose caucuses of notables, held together by personal standing; finance from members’ own wealth; quality of membership over quantity.
  • Duverger‘s point is that they are creations of parliament — parliamentary groups that reached outward to constituency committees.
  • Their failure is structural: once the franchise widens, notables cannot mobilise millions.
• Mass parties — the socialist answer, and Duverger’s central type.
  • Formal individual membership, dues, branches, a programme, an ideology, and an ancillary network of unions, cooperatives, newspapers, youth wings and clubs.
  • Created outside parliament by social groups excluded from it, which is why they treat the parliamentary group as an agent of the party rather than its head.
  • Duverger’s contagion from the left: mass organisation worked so well that conservative parties were forced to imitate it.
  • Their limit is ideological rigidity — a party anchored in one class cannot reach beyond it as that class shrinks.
• Catch-all parties — Otto Kirchheimer‘s concept, and the pivot of the whole sequence.
  • The party undertakes a “drastic reduction of its ideological baggage”, abandons the attempt to morally and intellectually encadre the masses, and reaches for the electorate at large.
  • Five changes travel together: ideological dilution; strengthening of top leadership; downgrading of the individual member; de-emphasis of a specific class or denominational clientele in favour of the population at large; and securing access to a wide variety of interest groups for finance and votes.
  • The causes are social, not merely strategic — the postwar welfare settlement blunted class antagonism, affluence and education loosened inherited loyalties, and broadcast media let leaders reach voters without the branch.
  • The gains are real: catch-all parties integrate diverse groups, reduce ideological polarisation, and respond faster to shifting opinion.
  • The costs are equally real: programmatic difference narrows, so voters lose a reason to identify; representation of specific class interests thins; and the party becomes dependent on leaders, media managers and donors rather than on members.
• Electoral-professional parties — Angelo Panebianco‘s refinement of Kirchheimer.
  • The bureaucrat inside the party is displaced by the professional outside it: pollsters, media consultants, fundraisers, and now data analysts.
  • Vertical organisational ties are replaced by direct appeals, and finance shifts from dues to interest groups and public funding.
• Cartel parties — Richard Katz and Peter Mair‘s argument, and the most disturbing of the sequence.
  • Parties have moved from civil society toward the state: public funding, state-regulated media access, and patronage in state and para-state bodies now sustain them.
  • Having become agents of the state, competing parties develop a shared interest in collusion — stabilising rules, sharing state resources, and treating politics as a profession rather than a vocation. Electoral competition becomes contained: outcomes change personnel, not policy direction.
  • The consequence Katz and Mair themselves stress is the opening of space for anti-establishment and populist challengers, whose central charge — that the mainstream parties are one cartel wearing different colours — is a description of the model.
  • Criticisms: the thesis fits small consensual European democracies better than majoritarian ones; Ruud Koole argued it conflates state subsidy with cartelisation; and cartelisation cannot explain why some systems produced strong challengers and others did not.
• Business-firm parties — Hopkin and Paolucci‘s type, where a party is built as an extension of a commercial organisation, with Forza Italia as the model: no membership base, borrowed corporate infrastructure, and a product-launch approach to elections.
• Movement parties — Herbert Kitschelt‘s type: minimal organisation, low professionalisation, an activist base carried over from a social movement, and a hostility to the very idea of a party. Podemos, the Five Star Movement, Syriza, the Aam Aadmi Party and Nepal’s Rastriya Swatantra Party all belong here.
• Digital parties — Paolo Gerbaudo‘s addition: platform-based membership, online plebiscites, and a hyperleader whose direct relationship with a dispersed membership bypasses intermediate structures entirely, reproducing Michels’s oligarchy in software.

Duverger’s structural analysis

• Beyond cadre and mass, Duverger classified parties by their basic organisational unit, and the classification still travels.
  • Branch — a permanent, geographically defined local unit, open membership, characteristic of socialist and liberal parties.
  • Cell — small, tightly-knit, organised at the workplace rather than the neighbourhood, permitting secrecy and constant political work; characteristic of communist parties.
  • Caucus — a small, closed group of notables and influentials, activated at election time; characteristic of cadre and conservative parties.
  • Militia — a private army organised on military lines, with uniforms, ranks and the readiness to use force; characteristic of fascist parties, and echoed in any party with an armed or paramilitary wing.
• The typology matters comparatively because the unit predicts the party’s relationship to the state: branch parties compete for it, cell parties infiltrate it, caucus parties inhabit it, militia parties intend to seize it.

Party systems: from counting to classifying

• A party system is not a party, and not a collection of parties, but the pattern of interaction between them — the stable, patterned competition that shapes government formation, opposition and policy.
• The oldest classification is numerical — one-party, two-party, multi-party — and its inadequacy is the starting point of modern work.
  • Counting cannot distinguish Weimar Germany from the Netherlands, though both had many parties, because it ignores what the parties believed and whether any of them wanted the system to survive.

Sartori’s classification

• Giovanni Sartori made two moves that turned counting into analysis.
  • Count only relevant parties. A party is relevant if it has coalition potential — it has been or could be needed for a governing majority — or blackmail potential — it is large enough that its existence alters the tactics of the parties that will govern, even though nobody will govern with it.
  • Add ideological distance to the number, producing a two-dimensional classification: fragmentation and polarisation.
• The resulting seven-fold scheme:
  • One-party — a single legal party; no competition permitted. Historic USSR, contemporary China, North Korea.
  • Hegemonic party — other parties are legally permitted but exist as satellites; alternation is not on offer. Mexico under the PRI for most of the twentieth century.
  • Predominant party — genuinely free competition in which one party keeps winning absolute majorities over a long run. India to 1967, Japan’s LDP, Sweden’s Social Democrats, South Africa’s ANC until 2024.
  • Two-party — two parties with a realistic prospect of single-party government, alternating in office. Britain, the United States.
  • Moderate pluralism — three to five relevant parties, small ideological distance, centripetal competition, coalition government, bilateral opposition willing to govern. Germany, the Netherlands.
  • Polarised pluralism — more than five relevant parties; anti-system parties present; bilateral oppositions on both flanks that cannot combine; a centre occupied rather than competed for; centrifugal competition and irresponsible opposition free to outbid. Weimar Germany, Italy of the First Republic, Chile before 1973.
  • Atomised — no party of significance; a residual category for systems with no structure at all.
• The category with the most contemporary purchase is polarised pluralism, because its symptoms — anti-system challengers, centrifugal drives, outbidding by parties that never expect to govern — describe several advanced democracies today far better than any two-party model does.

Other classifications worth having

• Jean Blondel classified by number and relative size: two-party, two-and-a-half party (two large plus a persistent small partner, as in Germany with the FDP for decades), multiparty with a dominant party, and multiparty without a dominant party.
• Myron Weiner and Joseph LaPalombara classified by competitiveness first.
  • Non-competitive systems split into one-party authoritarian, one-party pluralistic and one-party totalitarian.
  • Competitive systems split by their ideological character into ideological (Britain historically) and pragmatic — the American case, where the two parties have been described as the same wine in different bottles with different labels — and by structure into hegemonic, turnover and dominant patterns.
• Rokkan and later Lijphart added the crucial distinction between majoritarian systems, which concentrate power in a single-party executive, and consensus systems, which disperse it across coalitions, chambers, and levels — the difference that decides whether a divided society’s party system is a container for conflict or an amplifier of it.

What shapes a party system

Electoral rules: Duverger’s law

• Duverger’s law: the single-member plurality system tends to produce a two-party system. Duverger’s hypothesis, stated more weakly: proportional representation and the two-ballot majority system tend to produce multipartism.
• The law works through two distinct effects, and separating them is what makes it more than a correlation.
  • The mechanical effect — plurality rules systematically under-reward small parties in seats relative to votes, however loyally people vote for them.
  • The psychological effect — voters anticipate the mechanical effect and desert hopeless candidates for viable ones, and donors and candidates do the same, so the third party is squeezed from both ends.
• Refinements matter as much as the law:
  • William Riker restated it as a probabilistic tendency, not a determinism, and noted its failure where a third party is regionally concentrated.
  • Gary Cox‘s M+1 rule generalises it: in a district of magnitude M, the number of viable candidates tends toward M+1 — plurality (M=1) yields two, and PR with large districts yields many.
• India is the standing exception, and the explanation is the interesting part.
  • The two-party logic does operate, but within constituencies and within states rather than nationally.
  • Chhibber and Kollman supply the mechanism: whether local two-party competition aggregates into national two-party competition depends on the centralisation of authority and resources. Where states control what matters, party systems stay regionalised; where the centre does, they nationalise. India’s fiscal and political decentralisation from the late 1960s is what de-aggregated its party system.
  • The additional Indian factors are social heterogeneity and the territorial concentration of linguistic, caste and regional cleavages, which give regional parties a plurality-proof base.
• The lesson for developing countries specifically: importing an electoral system does not import its party system, because the law describes how rules interact with an existing cleavage structure, not how they create one.

Social cleavages: Lipset and Rokkan

• Seymour Martin Lipset and Stein Rokkan argued that European party systems were the frozen residue of historic conflicts, produced at two critical junctures.
  • The National Revolution generated two cleavages: centre versus periphery (the nation-building core against subject cultures) and state versus church (the secular state against the Church’s claims over education and morality).
  • The Industrial Revolution generated two more: land versus industry (primary versus secondary economic sectors) and owner versus worker (the class cleavage).
• The freezing hypothesis is the striking claim: the party systems of the 1960s reflected, with few exceptions, the cleavage structures of the 1920s. Parties, once formed, froze the alignments that formed them.
• Whether a cleavage becomes a party depends on four thresholds: legitimation (is dissent tolerated?), incorporation (can the group vote?), representation (can it win seats on its own?), and majority power (can it govern?).
• Cleavages in developing societies differ in kind, not merely in content.
  • The salient cleavages are frequently ascriptive — caste, ethnicity, religion, language, region — rather than the functional cleavages of class and sector.
  • Ascriptive cleavages are harder to compromise: an economic demand is divisible, an identity claim often is not, which is why Horowitz treats ethnic party systems in divided societies as a mechanism for hardening conflict rather than aggregating it.
  • Cross-cutting cleavages moderate conflict, since today’s opponent is tomorrow’s ally on another axis; reinforcing cleavages, where religion, language, region and class all divide the same line, are the dangerous configuration — the argument common to Lijphart and Horowitz, who differ about the cure and agree about the disease.
  • In much of the developing world the cleavage structure was not settled before parties formed. Parties therefore did not reflect a social structure so much as help construct one, which is why party competition can create or politicise identities as readily as it represents them.

In Europe parties froze cleavages that already existed; in much of the post-colonial world parties arrived first and helped decide which cleavages would matter.

Political parties in advanced industrial societies

The evidence for dealignment

• The trend is consistent across advanced democracies and is documented rather than asserted.
  • Party membership has fallen steeply across Western Europe since the 1960s, both absolutely and as a share of the electorate, in almost every established democracy.
  • Party identification has weakened — fewer voters report feeling close to any party, and those who do report weaker attachment.
  • Electoral volatility has risen sharply: more voters switch between elections, and new parties take a far larger share of the vote than in the post-war decades.
  • Turnout has declined in most advanced democracies, with global turnout falling by roughly a tenth over the last fifteen years.
  • Trust in parties is consistently the lowest of any political institution measured.
• Dealignment is not the same as realignment. Realignment means voters shift to a new stable alignment; dealignment means the anchoring itself weakens and no new anchor forms — which produces volatility without direction.

Why it happened

• Cognitive mobilisation — Russell Dalton‘s explanation. Better-educated voters with abundant political information no longer need the party as an informational shortcut, so they engage issue-by-issue rather than through a standing loyalty.
• Cleavage decline — the class and religious cleavages Lipset and Rokkan found frozen have thawed: manual working classes shrank, religious observance fell, and the social groups that anchored the mass parties dissolved.
• Post-materialism — Ronald Inglehart‘s silent revolution. Generations raised in physical and economic security shift priorities from material security to autonomy, self-expression, participation and quality of life, cutting across the old economic axis and producing new parties (Greens) and new demands the old parties could not process.
• Media change — leaders address voters directly; the party’s role as intermediary is bypassed.
• Cartelisation — where the governing alternatives converge on the same economic policy, the perceived stake in choosing between them falls.
• The honest qualification: participation has not simply declined, it has changed form. Party membership fell while petitioning, boycotting, demonstrating, single-issue campaigning and online mobilisation rose. Treating dealignment as the decline of democracy misses the substitution.

The decline-of-parties thesis, and its limits

• Lawson and Merkl‘s When Parties Fail argued that parties fail when they stop performing linkage — connecting citizens to the state — and that alternative organisations step into the gap: interest groups, movements, NGOs, and increasingly, direct digital mobilisation.
• Peter Mair‘s Ruling the Void is the sharpest statement, and its force lies in a reversal.
  • His claim is not that parties have grown weak but that they have withdrawn. Citizens withdrew from parties; parties withdrew from civil society into the state.
  • The representative link is therefore hollowed from both ends, leaving government that is technically competent and popularly unanchored.
  • The void is what populist challengers occupy, which is why Mair’s diagnosis and the populism literature are two descriptions of one process.
• Colin Crouch‘s post-democracy describes the same condition institutionally: elections continue, but effective decision-making migrates to executives, technocratic bodies and corporate interests.
• The counter-case is strong and should be made.
  • Parties still monopolise office. Movements pressure; only parties govern.
  • Parties adapt: cartelisation, professionalisation, primaries, digital membership and platform organisation are all adaptations, not obituaries.
  • New parties are still parties — the Greens, the populist right, Podemos and Five Star all built organisations, contested elections and entered coalitions.
  • The realistic conclusion is transformation, not disappearance: parties are weaker as membership organisations and as social communities, and no weaker as governing organisations.

The ascendancy of the centre-right

• The pattern is real and requires several explanations weighed together rather than one asserted.
  • Economic insecurity. Deindustrialisation, wage stagnation, and the regionally concentrated losses from trade and automation produced places with no alternative employment and a durable grievance. Rodrik‘s work on the political consequences of hyper-globalisation and the literature on the China shock supply the mechanism.
  • Cultural backlash. Pippa Norris and Ronald Inglehart argue the primary driver is not economics but a silent counter-revolution — older, less-educated and more rural voters reacting against the post-material values that became culturally dominant, with immigration as the focal issue.
  • Migration and security salience, raised by refugee movements and by terrorism, on issues where centre-left parties were internally divided between their working-class and university-educated wings.
  • The centre-left’s own convergence. Third-way acceptance of market economics removed economic differentiation, leaving the cultural axis as the main line of competition — an axis on which the right holds the advantage.
  • Cartelisation and distrust, which made “throw the establishment out” a viable positioning, and social media incentives, which reward polarising content.
  • Post-pandemic inflation and cost-of-living anger, which have punished incumbents of every stripe — a mechanism that looks like a rightward tilt only because more incumbents were of the centre and centre-left.
• Normalisation is the process that converted vote share into power: Cas Mudde‘s populist radical right — nativism, authoritarianism and populism — has been mainstreamed as centre-right parties adopted its issue positions and abandoned the cordon sanitaire, so that governing coalitions now depend on radical-right support in several European states.
• The counter-evidence must be stated, because it disciplines the claim.
  • The ascendancy is uneven and reversible: Poland’s 2023 alternation, and Australian and Canadian results in 2025, ran the other way; a liberal party led coalition formation in the Netherlands after the 2025 election even as the radical right retained strength.
  • Several centre-left incumbents survived; several radical-right parties lost presidential contests they were expected to win.
  • The accurate description is heightened volatility with a rightward tilt on the cultural axis, not a uniform global realignment.

Political parties in developing societies

The characteristic pattern

• Parties in much of the developing world display a paradox that Fred Riggs‘s prismatic society captures: they are modern and bureaucratic in form while remaining traditional in function, because the imported structure sits on an unreformed social base.
  • Formalism — the gap between prescribed rules and actual practice — is Riggs’s key term, and party constitutions with elaborate internal-democracy provisions that nobody follows are the textbook illustration.
• The recurring features:
  • Personalistic leadership — the party is an extension of a leader rather than an institution the leader occupies, so it does not reliably survive the leader.
  • Dynastic control — leadership circulates within families; candidate selection is centralised at the top.
  • Absent internal democracy — no organisational elections, no transparent candidate selection, no institutional route for dissent.
  • Identity-based mobilisation — appeals along caste, religion, ethnicity, language and region rather than programme.
  • Clientelism — the exchange of targeted, contingent benefits for votes, monitored through brokers. Kanchan Chandra‘s patronage democracy explains why ethnic identity is the efficient signal in such a system: where the state controls scarce goods and discretion is high, voters read ethnicity as a prediction of who will be favoured.
  • Weak financial transparency, with opaque funding, and a rising cost of entry that narrows candidacy to the wealthy.
  • Factionalism and high rates of defection, since a party built on patronage has no ideological glue.
• These are ideal-typical tendencies, not descriptions of every case. Botswana, Uruguay, Costa Rica, Chile, Taiwan and South Korea all have institutionalised party competition; and dynasty, opaque funding and personalism are conspicuously present in advanced democracies too.

Institutionalisation: the variable that decides

• Samuel Huntington supplied the framework in Political Order in Changing Societies: political decay occurs when participation outruns institutionalisation, and the strong party is the principal institution capable of absorbing new participants.
  • His four measures of institutionalisation: adaptability (does it survive change?), complexity (are there many differentiated units?), autonomy (is it independent of any single social force or leader?), and coherence (is there agreement on boundaries and procedures?).
  • His conclusion — that the organisational strength of the party matters more than its ideology or the constitution’s design — is the single most useful proposition in this literature.
• Mainwaring and Scully operationalised party system institutionalisation on four dimensions, and the measure travels well.
  • Stability of inter-party competition — low volatility, the same parties from election to election.
  • Roots in society — voters have durable attachments; parties have organisational presence between elections.
  • Legitimacy — parties and elections are accepted by voters and elites as the only route to power.
  • Organisational independence — the party has resources, structure and a life independent of its current leader.
  • Their finding is the load-bearing one: institutionalised party systems correlate strongly with democratic survival, and inchoate ones with breakdown, personalist rule and outsider presidents.

Do parties sustain and stabilise democracy in developing societies?

• The stabilising case.
  • Parties channel social conflict into electoral competition rather than street violence or coup, which is precisely what praetorian politics lacks.
  • They perform an integration function that has no counterpart in the advanced-industrial list: building a national political community out of regional, ethnic, linguistic and caste segments.
    • Kothari‘s account of the Congress system is the classic demonstration — a dominant party with genuine internal factional competition, functioning as an arena of consensus that absorbed and co-opted opposition demands rather than defeating them.
  • They recruit and socialise leadership where other channels are closed, and provide the only ladder from local to national politics for people without wealth or lineage.
  • They make alternation possible, which is the operative test of consolidation — Huntington’s two-turnover test.
• The destabilising case.
  • Personal vehicles with no organisation independent of a leader mean alternation changes who holds office without consolidating anything.
  • Ethnic parties can harden cleavages rather than aggregate across them, and where the party system maps onto a single reinforcing cleavage, elections become censuses rather than choices.
  • Clientelism substitutes patronage for programme, so accountability runs through brokers rather than policy.
  • Absent internal democracy turns candidate selection into a market, raising the cost of entry and the incentive to recoup it.
  • Decisively, parties in office are now the principal agents of democratic erosion.
    • Nancy Bermeo shows that contemporary backsliding takes the form of executive aggrandisement — incremental, legal and led from within — rather than the coup and self-coup of the twentieth century.
    • Levitsky and Ziblatt trace the sequence by which elected leaders capture the referees, sideline opponents and rewrite the rules of the game.
    • Kim Lane Scheppele‘s autocratic legalism names the technique of doing all of it through law, so that each step is individually defensible.
• The measured verdict: parties are necessary and insufficient. What distinguishes the cases is institutionalisation — where parties have roots and organisation independent of leaders, democracy survives; where politics runs through personal vehicles, elections continue while the substance of democracy erodes.

The Indian party system: neither western nor indigenous

• The proposition is a claim about hybridity, and the evidence sits on both sides.
• What is western about it: competitive multiparty elections, universal adult franchise from the first election, a Westminster parliamentary frame, the language of ideology and programme, and — for the Communist and Socialist parties — organisational models imported wholesale.
• What is indigenous about it:
  • Mobilisation runs through caste, community, language and region, categories with no European counterpart.
  • Morris-Jones identified three idioms in Indian politics operating simultaneously — the modern (constitutional, secular, of the courts and parliament), the traditional (of caste, kinship and locality) and the saintly (of moral appeal and renunciation) — and parties speak all three.
  • Rudolph and Rudolph‘s modernity of tradition is the general argument: traditional structures were not swept aside by modernisation but became the vehicles of it, with caste associations turning into interest groups and vote banks.
  • The Congress system is not a category in any Western typology: a dominant party with real internal competition, in which opposition parties functioned as parties of pressure working on Congress factions rather than as alternative governments.
• What is neither: the Duverger exception — a first-past-the-post system that has produced sustained multipartism nationally while remaining broadly bipolar within states and constituencies — is a genuinely distinct configuration that the imported theory did not predict and the indigenous account cannot explain alone.
• The prismatic reading holds the two halves together: the form is modern and imported, the content is traditional and local, and the friction between them is the system’s characteristic behaviour rather than a transitional defect.

Contemporary developments

• Party finance and regulation have become the central institutional battleground worldwide — state funding in Europe, disclosure regimes, spending caps, and the Indian Supreme Court’s 2024 judgment striking down the electoral bonds scheme as violative of the right to information.
• Digital and platform organisation has changed what membership means: online sign-up, plebiscitary consultation of members, micro-targeted advertising, and now generative AI in campaign content, which raises unresolved questions about synthetic media in elections.
• Movement-to-party conversion is being tested in real time.
  • Nepal‘s Gen Z uprising of September 2025 toppled a government within days and produced an interim administration.
  • The March 2026 election that followed gave the anti-establishment Rastriya Swatantra Party a near two-thirds parliamentary majority.
  • The case will settle, one way or the other, whether a leaderless movement can institutionalise into a governing party — the transition at which Five Star, Podemos and Syriza each faltered in different ways.
• Party system institutionalisation is falling in places that had it: several European systems now show volatility levels once associated only with new democracies, which suggests the advanced-industrial and developing categories are converging from the top down rather than the bottom up.
• Women’s representation inside parties remains the sharpest gap between the descriptive claims parties make and their internal practice, and the mechanism that has moved it fastest everywhere is candidate quotas — voluntary, statutory or reserved-seat.

Populism and the party form

• Populism is not an ideology of the same order as liberalism or socialism, and treating it as one produces bad comparative analysis.
  • Cas Mudde‘s thin-centred ideology formulation is the standard: society divides into “the pure people” against “the corrupt elite”, and politics should express the general will.
  • Being thin, it must attach to a host ideology — nativist on the right, redistributive on the left — which is why populisms of opposite content share a grammar.
• What makes it a party-system phenomenon rather than merely a rhetorical style is where it grows.
  • It grows in the space cartelisation created. If the governing alternatives have converged, the only available axis of differentiation is establishment versus outsider, and the populist owns it by construction.
  • It grows where representation has thinned — Mair’s void — because the claim “they do not represent you” is empirically defensible before it is rhetorically effective.
• Organisationally, populist parties tend toward a distinctive form.
  • Low intermediate structure and a direct, unmediated bond between leader and followers, since intermediaries are themselves suspect.
  • Plebiscitary internal procedures — online votes, direct consultation — which look like radical democracy and function as leader ratification, a digital restatement of Michels.
  • Rapid organisational fragility: parties built without structure fracture when the leader falters, which is why populist vote shares are volatile even where they are large.
• The left and right variants behave differently. Right populism stresses nativism, sovereignty and cultural threat; Latin American left populism — Laclau‘s theoretical home ground, with his empty signifier and chain of equivalence — stresses redistribution and the incorporation of excluded majorities.
• The comparative caution is that populism is not confined to developing societies, and never was. Its recent centre of gravity has been in wealthy, long-established democracies, which is exactly what the modernisation thesis — affluence breeds stable democracy — did not predict.

Party finance in comparative perspective

• Finance is where the abstract typologies become concrete, because who pays for a party determines whose demands it processes.
• Four models are visible across systems, and most real systems mix them.
  • Membership dues — the mass-party model, now residual almost everywhere as membership collapses.
  • Interest-group and corporate funding — the catch-all model, which buys access rather than instruction, and which produces the access-asymmetry critique that organised business is heard continuously and the unorganised never.
  • Public funding — the cartel-party model, which stabilises parties, insulates them from donors, and simultaneously detaches them from society and raises barriers against new entrants, since eligibility rules are written by incumbents.
  • Personal wealth of the candidate — the dominant model where regulation is weak, and the mechanism by which the cost of entry filters who can stand at all.
• Regulatory instruments differ by system: contribution caps, expenditure ceilings, disclosure requirements, bans on foreign or corporate donations, and free broadcast time. Enforcement, not the statute, is the variable that matters.
• India is the sharpest recent case.
  • The electoral bonds scheme permitted unlimited, anonymous corporate donation routed through a state bank.
  • The Supreme Court struck it down in February 2024 as violating the voter’s right to information under Article 19(1)(a).
  • The instructive point is institutional: a court, not a legislature, resolved a finance question that incumbents of every party had an interest in leaving unresolved.

Parties, gender and descriptive representation

• Parties are the gatekeepers of representation, because in every system with parties the decisive selection happens before the voter sees a ballot.
  • This is why measures aimed at the electorate rarely move descriptive representation, and measures aimed at candidate selection do.
• Three mechanisms, in ascending order of bindingness:
  • Voluntary party quotas — a party commits internally to a share of women candidates. Effective where party discipline is high, and reversible at will.
  • Legislative candidate quotas — statute requires a share of candidates, sometimes with placement mandates so women are not clustered in unwinnable seats.
  • Reserved seats — a share of seats is set aside, the strongest instrument and the one that most constrains competition.
• The finding that recurs comparatively is that descriptive representation shifts substantive outcomes, but not automatically: it works where the representatives acquire an independent political base rather than remaining nominees of whoever selected them.
• The Indian anchor is the contrast between panchayats and Parliament.
  • Reservation at the local tier delivered a third of seats, and more in several states, within a single electoral cycle.
  • The national legislature moved slowly until the 106th Amendment (Nari Shakti Vandan Adhiniyam), 2023, whose implementation is tied to a future census and delimitation.

Conclusion

The comparative study of parties has reached an ironic destination. The advanced industrial democracies produced the theory — the mass party, the frozen cleavage, the catch-all transition — and are now the cases where that theory least applies.

• Their parties have shed the members, the ideological distinctiveness and the social roots the theory assumed, and hold office without holding society.
• The developing societies, which imported the form, are where parties still do the heaviest work — integrating segmented societies, recruiting leadership, and constituting political communities that nothing else constitutes.
• The variable that decides outcomes in both worlds is the same, and it is neither ideology nor wealth nor constitutional design but institutionalisation: whether a party exists independently of the person currently leading it.
• Where it does, competition consolidates. Where it does not, elections continue while representation quietly stops working — which is precisely the condition the backsliding literature is describing.