Political Participation, Representation and Electoral Systems in Comparative Perspective

Representation is the device by which a large society governs itself without assembling. Participation is what citizens do to make the device work. Both were once treated as solved problems — the franchise had been won, elections were held, and the remaining questions were technical. That confidence has not survived. Turnout is falling where democracy is oldest, rising where it is newest; the elected are less like the electorate than the theory assumed; and elected governments have become the principal threat to the institutions that elected them.

Representation: what the word carries

PSIR Optional Notes {PDF}

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

Discounted Price | Buy Now @ ₹1000 →

• To represent is to make present again something that is not literally present. Every dispute about representative democracy is a dispute about what exactly is being made present — a person, a body of opinion, a social group, or an interest.
• Hanna Pitkin‘s The Concept of Representation remains the organising scheme, and its value is that it shows the four senses to be distinct and often in conflict.
  • Formalistic representation concerns the arrangements by which someone acquires and loses the right to act for others. It has two halves: authorisation — how the representative was empowered — and accountability — how the represented can call them to answer. A body can be perfectly authorised and wholly unaccountable, and vice versa.
  • Descriptive representation concerns resemblance — whether the representative body mirrors the represented in social composition. On this view a legislature is a map or a sample of the nation, and its accuracy is what makes it representative.
  • Symbolic representation concerns what the representative stands for in the eyes of the represented — the affective and identificatory dimension that explains why a flag or a head of state represents without acting.
  • Substantive representation concerns acting in the interest of the represented, responsively and with independent judgement. Pitkin’s own conclusion is that this is the sense that matters most.
• The mandate–independence controversy is Pitkin’s name for the oldest quarrel in the field, and she treats it as genuinely unresolved rather than as a puzzle with an answer.
  • The delegate model holds that the representative transmits instructions and is a conduit for a will formed elsewhere.
  • The trustee model, given classic form in Burke‘s Speech to the Electors of Bristol, holds that the representative owes the constituency judgement rather than obedience.
    • On this view the representative betrays the electorate precisely by sacrificing judgement to opinion.
    • Parliament, in Burke’s phrase, is “a deliberative assembly of one nation, with one interest, that of the whole” — not a congress of ambassadors from rival and hostile interests.
  • The party mandate model, which describes actual practice in most systems, holds that the representative is bound neither to constituents nor to conscience but to a programme on which a party was elected — which relocates the problem rather than solving it.
• Jane Mansbridge extended the scheme with four contemporary forms that fit modern politics better than the two-way choice.
  • Promissory — the classical model, in which the representative keeps promises made at election time.
  • Anticipatory — the representative acts on what voters will approve at the next election, which makes representation prospective rather than retrospective.
  • Gyroscopic — the representative acts from internal orientations, and voters choose the gyroscope rather than the instructions.
  • Surrogate — a representative speaks for people outside their own constituency who have no one else, which is how minorities dispersed too thinly to win anywhere get represented at all.
• Michael Saward made the further move of treating representation as a claim rather than a fact: someone puts themselves forward as speaking for a constituency, and the constituency either accepts the claim or does not. This explains how movements, NGOs and celebrities represent without election, and why their representativeness is always contestable.

Descriptive representation and why it matters

• Anne Phillips framed the distinction, and her argument is not that presence replaces ideas but that ideas alone have demonstrably failed for groups whose concerns were never on any party’s agenda in the first place.
• The empirical question is whether descriptive representation produces substantive representation, and the answer is a qualified yes.
  • The critical mass hypothesis — associated with Drude Dahlerup — holds that a threshold proportion, often put near a third, must be reached before a minority in a legislature can change outcomes rather than merely being present.
  • Sarah Childs and Mona Lena Krook shifted the emphasis from critical mass to critical actors: what matters is whether individual representatives act, and whether the institutional conditions let them, rather than whether an arithmetic threshold has been crossed.
  • The consistent comparative finding is that descriptive representation shifts substantive outcomes but not automatically, and works best where the representatives have an independent political base rather than depending on whoever selected them.
• Symbolic effects are real and measurable: the presence of representatives from an excluded group raises political engagement, aspiration and trust among that group, independently of any policy change.

The struggle for representation in industrial societies

• The franchise was not granted; it was fought for, and the fight is what made industrial democracies both democratic and stable.
• The British sequence is the standard illustration of gradual, contested extension.
  • The Reform Act of 1832 redistributed seats and modestly widened a property-based franchise, disappointing the movements that had pressed for it.
  • Chartism, from 1838, was the first mass working-class political movement, and its Six Points — universal male suffrage, secret ballot, equal constituencies, payment of members, abolition of the property qualification, and annual parliaments — were rejected in their own time and, apart from the last, enacted within a century.
  • The Reform Acts of 1867 and 1884 brought in the urban and then the rural working class; the Representation of the People Act 1918 enfranchised all men and women over thirty with a property qualification; 1928 achieved equal universal suffrage.
• The American sequence shows that formal enfranchisement and effective enfranchisement can be separated by a century — the Fifteenth Amendment of 1870 was followed by disenfranchisement through literacy tests, poll taxes and violence, and was made effective only by the Voting Rights Act of 1965, itself won by a movement.
• Women’s suffrage followed its own long and uneven course, from New Zealand in 1893 to Switzerland’s federal enfranchisement of women in 1971 — a reminder that neither wealth nor stability predicts inclusion.
• Rokkan’s four thresholds describe what any excluded group had to cross, and they are still the right checklist for assessing inclusion anywhere.
  • Legitimation — is opposition tolerated at all?
  • Incorporation — can members of the group vote?
  • Representation — can they win seats without allying with established forces?
  • Executive power — can they actually govern?

What the struggle did to the quality of democracy

• It transformed the political party. A mass electorate could not be organised by caucuses of notables, so extension produced the mass party with its branches, dues and programme — and produced, with it, the first genuinely programmatic politics.
• It produced the welfare state. Enfranchised workers voted for redistribution, and the association between franchise extension and the growth of social spending is one of the most robust findings in political economy.
• It converted class conflict into electoral competition. The historic significance of extension is that it offered organised labour a route to power that did not require revolution, and the societies that extended earliest were the ones least disrupted by revolutionary movements.
• It deepened citizenship. T. H. Marshall‘s sequence — civil rights in the eighteenth century, political rights in the nineteenth, social rights in the twentieth — treats political rights as the hinge: they were won using civil rights, and they were used to win social rights.
• It raised state capacity and legitimacy, since a state that must be re-authorised by everyone has both a stronger claim to obedience and a stronger need to deliver.
• The counter-case deserves its place, because the story is not one of steady improvement.
  • Michels predicted that extension would produce oligarchy inside the new mass organisations, and largely it did.
  • The elite theorists — Mosca and Pareto — held that extension changes which elite rules and never whether an elite rules.
  • Schumpeter accepted the whole critique and redefined democracy to fit it: a method for selecting leaders through competitive struggle for the vote, in which the people’s role is to produce a government, not to govern.
  • Universal franchise did not deliver equal influence. Turnout, donation, lobbying and office-holding all remain steeply graded by class, which is why the twentieth-century agenda shifted from who may vote to whose vote counts.

The franchise settled who is entitled to participate; a century later the open question is why participation still tracks income as closely as it does.

“Parliament is not a congress of ambassadors from different and hostile interests… but a deliberative assembly of one nation, with one interest, that of the whole.” — Edmund Burke

Political participation

• Political participation is conventionally defined, following Verba and Nie, as activity by private citizens aimed at influencing the selection of governmental personnel or the actions they take. Three features of the definition are contested and worth noticing.
  • It covers acts, not attitudes — interest and knowledge are not participation.
  • It covers acts by private citizens — the activity of officials is governing, not participating.
  • Whether it should cover only legal and conventional acts is exactly what the study of protest reopened.
• Milbrath‘s hierarchy remains a useful first sorting, ordered by the cost and intensity of the act.
  • Gladiators — a small minority who campaign, contest, fund and organise.
  • Spectators — the large majority who vote, discuss politics and attend occasionally.
  • Apathetics — those who do none of it.
• Verba, Nie and Kim replaced the single ladder with distinct modes of participation that recruit different kinds of people and produce different kinds of influence: voting, campaign activity, communal activity and particularised contacting of officials about a personal problem.
• Barnes and Kaase‘s Political Action established that unconventional participation — petitions, demonstrations, boycotts, occupations — is not the resort of the alienated but is practised by the well-educated and politically confident, alongside conventional participation rather than instead of it.
• Pippa Norris‘s critical citizens and democratic phoenix arguments complete the correction: citizens in advanced democracies have become more critical of institutions while remaining committed to democratic values, and their participation has migrated from parties to protest, single-issue campaigning and digital mobilisation.

Why some people participate and others do not

• The socio-economic status model is the oldest and most robust finding: participation rises with education, income and occupational status, and education is the strongest single predictor almost everywhere.
• The civic voluntarism model of Verba, Schlozman and Brady explains why, through three answers to the question of why people do not take part.
  • “Because they can’t” — they lack resources: time, money and civic skills.
  • “Because they don’t want to” — they lack engagement: interest, efficacy, information, partisan attachment.
  • “Because nobody asked” — they are outside the recruitment networks of parties, unions, churches and associations that mobilise people into politics.
  • The third is the most policy-relevant, because mobilisation is the variable an organisation can change and social structure is not.
• Political efficacy — the belief that one’s action can matter, and that the system will respond — is the crucial attitudinal variable, and it is the one that erodes fastest under backsliding.
• Institutional context matters as much as individual characteristics: compulsory voting, registration rules, proportionality, electoral competitiveness and the number of elections held all move participation independently of who the citizens are.
• Participatory democratic theory treats all of this as more than a technical matter.
  • Carole Pateman argued that participation is educative: the capacity for democratic citizenship is developed by exercising it, not possessed in advance.
  • It follows that participation in workplaces and local institutions is a precondition of competent participation at the national level, which is why she treats industrial democracy as a democratic question.
  • C. B. Macpherson made participation the corrective to the thinness of the liberal-democratic model, which he regarded as having secured the form of self-government while abandoning its substance.

Voter turnout: the comparative puzzle

• The theoretical starting point is that voting is hard to explain at all.
  • Downs‘s paradox of voting: the probability of a single vote being decisive is effectively zero while the cost of voting is real, so the strictly rational citizen abstains.
  • Riker and Ordeshook rescued the act by adding a D term — the intrinsic value of voting itself, as duty, expression and identity.
  • The concession embedded in that fix is significant: turnout is not explained by instrumental calculation, which is why institutional and normative variables outperform economic ones in every empirical model.
• Institutional determinants are the most powerful and the most manipulable.
  • Compulsory voting, enforced, raises turnout by a very large margin — Australia, Belgium and Brazil being the standard cases.
  • Proportional representation raises turnout relative to plurality systems, because fewer votes are wasted and fewer constituencies are foregone conclusions.
  • Registration rules — automatic or state-initiated registration produces higher turnout than voter-initiated registration, which places the cost on the citizen.
  • Timing — weekend or holiday polling, multi-day polling and postal or early voting all raise turnout.
  • Concurrent elections raise turnout for the lower-salience contest, and the reverse — too many separate elections — produces measurable voter fatigue.
  • Closeness and salience — turnout rises where the outcome is genuinely uncertain and where the office matters, which is why second-order elections consistently underperform.
• Sociological determinants: education, income, age (turnout rises with age and has fallen most among the young), residential stability, union and associational membership, and religious observance.
• Attitudinal determinants: efficacy, trust, party identification, and satisfaction with the available choices.
• Why turnout has fallen in the advanced democracies — the explanations are the same ones that explain dealignment.
  • Generational replacement — cohorts socialised into voting as a duty are being replaced by cohorts who treat it as a choice.
  • Weakening party identification removes the habitual anchor that carried marginal voters to the poll.
  • Programmatic convergence reduces the perceived stake in the outcome.
  • The decline of mobilising organisations — unions, churches, mass-membership parties — removes the “nobody asked” mechanism at scale.
  • Globalisation and delegation to central banks, courts and supranational bodies narrow what a national election is believed to decide.
• The scale: global turnout has fallen by close to a tenth over the last fifteen years, and the fall is concentrated in established democracies. The 2024 electoral super-cycle, in which countries containing roughly half the world’s population went to the polls, produced high participation in absolute numbers and continued the downward trend in rates.

Unequal turnout and its consequences

• The most consequential fact about low turnout is not its level but its distribution: in most advanced democracies, abstention is concentrated among the young, the poor and the less educated.
• Arend Lijphart named this democracy’s unresolved dilemma: unequal turnout produces systematically unequal influence, since parties respond to those who vote, which reinforces the alienation of those who do not, in a self-reinforcing cycle. His remedy is compulsory voting, defended precisely because it is the only measure that equalises participation across class.
• The developing-world pattern frequently inverts this, and the inversion is theoretically important.
  • In India, turnout among the poor, rural, Dalit, Adivasi and less-educated has for decades been at or above that of the urban middle class — the reverse of the Western gradient.
  • The explanation offered by Mukulika Banerjee‘s ethnographic work is that Indian voters treat the act of voting as a duty and a public performance of equal citizenship, valued for what it says about their standing rather than only for what it delivers.
  • Yogendra Yadav‘s two democratic upsurges describe the same phenomenon in aggregate: the entry of the intermediate castes into effective political participation, and then of Dalits, Adivasis, and the very poor — a participatory deepening without a parallel in the West.
• Remedies worth listing: compulsory voting; automatic registration; polling on holidays; multi-day and postal voting; lowering the voting age with civic education; making elections concurrent; and reducing the number of separate polls.

Participation in developing societies

Identity and political participation

• Identity is a more powerful determinant of participation in developing societies than in advanced industrial ones, and the reasons are structural rather than cultural.
  • Identity lowers the cost of mobilisation. Caste, ethnic, kin and religious networks already exist, already have leaders and already carry trust, so a party or movement that works through them avoids the organising costs that Olson’s logic imposes on everyone else.
  • Identity carries information. Kanchan Chandra‘s patronage democracy argument explains why: where the state controls scarce goods and officials exercise wide discretion, a voter cannot evaluate policy promises but can predict who will be favoured, and ethnic identity is the cheapest available signal of that.
  • Identity is often the form a material demand takes. Reservation, quotas, land, recognition of a language and access to public employment are distributive claims made in an identity idiom because that is the idiom the state itself uses to classify people.
  • Colonial classification created the categories. Census enumeration, communal electorates and administrative schedules converted fluid social groupings into countable, politically usable ones.
  • Weak programmatic differentiation leaves identity as the most legible cleavage available, since parties competing on similar economic programmes must differentiate on something.
• The evidence that identity is not everything is equally strong, and any assessment that stops at identity is out of date.
  • Welfare delivery has become a major axis of voting in India and elsewhere, with a large and identifiable constituency of beneficiaries who vote on receipt rather than on identity, and survey evidence shows welfare receipt cutting across caste and community.
  • Identity is a floor, not a ceiling. It predicts the initial disposition and is routinely overridden by leadership, incumbency performance, local factors and coalition arithmetic.
  • Identity blocs are unstable and internally contested: they split, are outbid, and are reconstructed by political entrepreneurs, which is why treating them as fixed produces consistently wrong predictions.
  • Cross-cutting identities — a Dalit woman in an urban informal job holds three politically salient identities that pull in different directions — make single-identity accounts arithmetically impossible.
• The defensible conclusion is that identity structures the entry into participation in developing societies and does not determine its content, and that its salience is a function of how the state distributes goods rather than of any inherent traditionalism.

Does participation by the underprivileged strengthen democracy or produce chaos?

• The pessimistic case has a serious theoretical basis and should be stated at full strength before being answered.
  • Huntington‘s central proposition is that political decay follows when participation grows faster than institutionalisation — mobilised populations entering weakly institutionalised systems produce praetorianism, in which every social force acts directly and none is mediated.
  • Mass entry can raise distributive conflict, since new participants make claims on resources that have not grown.
  • Ascriptive mobilisation can harden cleavages, as Horowitz argues of ethnic party systems in divided societies, converting elections into censuses.
  • Rapid mobilisation without organisation creates the conditions for populist capture by leaders who claim the mobilised as their own.
• The optimistic case is better supported by the evidence, at least in the Indian instance.
  • Democratic deepening is the observable pattern: participation by the previously excluded has produced descriptive representation, policy attention and, in time, changed elites — Jaffrelot‘s silent revolution describing the rise of lower-caste political power through the ballot rather than through revolution.
  • Elections have absorbed conflict that would otherwise have been fought outside institutions; the empirical claim is that democratic participation is a substitute for violence, not a prelude to it.
  • Institutions have proved more elastic than Huntington expected. India expanded participation enormously without collapse, which is the standing counter-example.
  • The conflict that mass entry generates is the conflict democracy is for — the alternative to noisy inclusion is quiet exclusion, which is not stability but suppression.
• Atul Kohli‘s formulation is the best synthesis available: the relationship between mobilisation and stability is an inverted U.
  • The early phase of mass entry is genuinely destabilising, as new groups press claims and old elites resist losing position.
  • As those groups are accommodated within institutions — parties, legislatures, bureaucracies, reservation — conflict subsides, and democracy is stronger for having absorbed them.
  • Whether a system passes through the U depends on the accommodative capacity of its institutions, not on the character of the entrants — which relocates the whole question from the people mobilising to the state receiving them.

Democracy, autocracy and political violence

• The proposition that autocracies restrain violence better than democracies is superficially plausible — they can repress, they do not have to bargain, and they exhibit less visible disorder — and it does not survive comparative testing.
• The finding that does hold is curvilinear. The regimes most prone to internal political violence are neither consolidated democracies nor consolidated autocracies but hybrid regimes — the anocracies of the comparative literature — which combine enough openness to permit mobilisation with enough repression to make it costly and enough institutional weakness to prevent accommodation.
• Why consolidated democracies restrain violence:
  • Channels exist, so grievance has a route that is cheaper than violence.
  • Alternation is possible, so losing is survivable and there is no need to fight the last battle over the last election.
  • Rights protection and independent courts provide redress that does not require the streets.
  • Federalism, proportionality and power-sharing allow accommodation of territorially and socially concentrated groups.
• Why autocracies’ apparent calm is misleading:
  • Repression displaces rather than removes conflict, into insurgency, exile, and long silences that end abruptly.
  • Autocracies monopolise violence rather than reducing it: state violence is not counted as political violence in the intuitive reckoning, though it is the largest category.
  • Succession is unsolved, which makes autocratic transitions systematically more violent than democratic ones.
  • Information is poor — a regime that punishes bad news cannot detect a crisis before it becomes one.
• The important qualification is that democratisation is dangerous even though democracy is safe.
  • Mansfield and Snyder show that states in transition — mobilised populations, weak institutions, elites competing for a new mass audience — are more prone to nationalist conflict than either stable type.
  • This is the anocracy finding stated dynamically, and it is the strongest available version of the case for caution about rapid democratisation.

How democratic politics constructs citizenship

• Citizenship is not a natural fact that democracy recognises; it is an artefact that democratic politics produces, and the production is continuous rather than completed.
• Four dimensions of the concept, all of which democratic politics acts on.
  • Status — formal membership of a political community, with the state deciding who is in.
  • Rights — civil, political and social entitlements attaching to that status.
  • Participation — citizenship as a practice, not a possession.
  • Identity — the sense of belonging that makes the status meaningful.
• T. H. Marshall‘s account is the standard developmental story, and it is a story about sequence.
  • Civil rights established the legal person; political rights gave that person a share in power; social rights made the share usable by guaranteeing a minimum of material security.
  • The sequence is contested. Feminist critics note that it describes the trajectory of the male working class and misdescribes women’s, whose civil rights arrived late and partially.
  • Post-colonial critics note that it inverts outside Europe: many post-colonial states granted universal political rights first, to populations that possessed neither secure civil rights nor social ones.
• The mechanisms by which democratic politics constructs citizenship:
  • Universal franchise constitutes the citizen as an equal. One person one vote is the only institution in most societies that treats every adult as exactly equivalent, and it does so publicly and periodically.
  • Elections perform equality, which is why the ritual matters independently of the outcome — Banerjee’s Indian evidence on voting as an assertion of dignity is the clearest demonstration.
  • Contestation defines the boundary. Who counts as a citizen has been settled and re-settled through political struggle — over property, race, gender, religion and now migration and documentation — and it is never settled permanently.
  • Claims-making creates rights. Movements convert felt needs into demands, demands into entitlements and entitlements into law; India’s movement-to-statute chain converted information, work, food and forest access into justiciable rights that were not citizenship’s content before.
  • Participation produces the capacity for citizenship. Mill, Tocqueville and Pateman all argue in different registers that citizens are made by the exercise of citizenship rather than qualified for it in advance.
  • Political society extends citizenship informally. Partha Chatterjee‘s point is that most of the post-colonial population negotiates with the state not as rights-bearing citizens in civil society but as populations making claims that are quasi-legal — and that these negotiations are how citizenship is actually distributed.
• Differentiated citizenship is the contemporary complication.
  • Iris Marion Young and Will Kymlicka argue that formally identical citizenship can entrench substantive inequality, because identical treatment of unequally placed groups reproduces the inequality.
  • Group-differentiated rights — reservation, territorial autonomy, language guarantees, personal law — may therefore be required for equal citizenship rather than being a departure from it.
  • The counter-argument is that differentiation institutionalises the divisions it was meant to remedy, and creates a standing interest in maintaining the categories.
• The Indian anchor is unusually rich, and every element of it is an argument about the same question.
  • A constitution establishing single, uniform citizenship while simultaneously providing differentiated protections for specified groups.
  • A long-running controversy over documentation, registration and the amended citizenship law, in which the boundary of the community is contested directly.
  • A Supreme Court judgment upholding the special provision governing citizenship in Assam, which fixed a cut-off date for a particular population.
  • What unites them is the question who the democratic community is for — the question democratic politics answers continuously rather than once.

Electoral systems

• Once representative democracy became unavoidable, the electoral system became part of democratic theory rather than a matter of administration, because it decides what a vote turns into.
• What an electoral system does, beyond counting:
  • Translates votes into seats, by a rule that is never neutral.
  • Shapes the party system, through the mechanical and psychological effects that give Duverger’s law its force.
  • Determines the accountability relationship — to a constituency, to a party, or to a national electorate.
  • Decides how much voice minorities have, which is the criterion on which the systems differ most sharply.
• The criteria on which systems are judged, and they trade off against one another.
  • Proportionality — the fit between vote share and seat share, measurable by disproportionality indices.
  • Minimum wastage of votes — the share of votes that elect nobody.
  • Government stability and clarity of responsibility — whether the system produces governments that can act and voters who can identify whom to blame.
  • The local link — whether a voter has an identifiable representative.
  • Minority inclusion — whether persistent minorities can win anything.
  • Simplicity — whether the voter understands what their ballot does.

Majoritarian and plurality systems

• First past the post (simple plurality) — the candidate with the most votes wins, regardless of whether that is a majority.
  • Merits: simplicity; a clear local link; a tendency to produce single-party majorities and therefore stable and identifiable government; and it penalises small extremist parties without a territorial base.
  • Demerits: the least representative of all systems and the one with the greatest wastage of votes; a party can win a majority of seats on a minority of votes; it systematically under-represents parties whose support is spread thinly and over-rewards those whose support is territorially concentrated.
  • Used for: the House of Commons in the United Kingdom, the House of Representatives in the United States, and the Lok Sabha and state legislative assemblies in India.
• The second ballot (two-round system) — a candidate must win an absolute majority; if nobody does, a second round is held, typically between the top two, which guarantees a majority winner.
  • Merits: the winner has majority support; voters can vote sincerely in the first round and strategically in the second; it encourages coalition-building between rounds.
  • Demerits: expense, administrative burden and voter fatigue, with turnout usually falling between rounds.
  • Used for: the presidential elections of France and Sri Lanka, among many others.
• The alternative vote (instant-runoff) — voters rank candidates; if nobody has an absolute majority, the lowest-placed candidate is eliminated and their ballots transferred to the next preference, repeatedly until someone crosses fifty per cent.
  • Merits: delivers a majority winner in a single poll, so it is an improvement on the second ballot in cost and fatigue; it rewards candidates who are broadly acceptable rather than merely first-placed, which Horowitz argues gives it real value in divided societies because it creates an incentive for vote-pooling across communal lines.
  • Demerits: still a majoritarian system applied in single-member districts, so it remains disproportional overall; counting is more complex.
  • Used for: the Australian House of Representatives.

Proportional representation

• The rationale is that even an absolute-majority requirement leaves large minorities entirely unrepresented, and that democracy’s purpose includes the protection of minorities, so a system that gives a persistent minority no seats has failed a test that majority rule alone cannot state.
• List systems — the most proportional family, dominant in continental Europe.
  • The territory is treated as one or a few large multi-member constituencies; each party receives seats in proportion to its vote share and fills them from a list.
  • Closed lists put the order in the party’s hands; open lists let voters influence which listed candidates are elected.
  • Merits: minimum wastage of votes; representation for dispersed minorities without any need for separate quotas; the possibility of including candidates who could not win a constituency; and a reduced role for the individual personality relative to the party and its programme.
  • Demerits: it breaks the local link; it transfers enormous power to party leaderships who control list placement; and it reliably produces coalition government, with the instability and post-election bargaining that follow.
  • Thresholds are the standard corrective against fragmentation — a party must clear a minimum vote share to qualify for any seat, set at 5% in Germany and 3.25% in Israel.
  • Used for: the Knesset in Israel, and, with variations, most of continental Europe.
• The single transferable vote (STV) — proportional representation achieved through ranked preferences in multi-member constituencies.
  • A candidate is elected on reaching a quota, not an absolute majority. The standard formula is the Droop quota: the total valid votes divided by one more than the number of seats, plus one.
  • Surplus votes above the quota are transferred to later preferences at a reduced value; if no one reaches the quota, the lowest candidate is eliminated and their votes transferred; the process repeats until all seats are filled.
  • Merits: proportional and preserves a constituency link; lets voters choose between candidates of the same party; wastes very few votes.
  • Demerits: complex to count; large districts are needed for real proportionality, which weakens the local link it is prized for.
  • The distinction from the alternative vote is worth stating exactly: AV is majoritarian and used in single-member seats, while STV is proportional and used in multi-member seats — the transfer mechanism is shared, the quota and the purpose are not.
• Mixed systems combine the two families and are the most common recent choice for new constitutions.
  • Mixed-member proportional (MMP) — voters cast two votes, one for a constituency candidate and one for a party list, and the list seats are allocated so that the overall result is proportional. Used in Germany and New Zealand, and generally regarded as the best available compromise between the local link and proportionality.
  • Parallel or mixed-member majoritarian systems allocate constituency and list seats independently, so the outcome is semi-proportional at best.

The Indian case

• Direct elections to the Lok Sabha and the state legislative assemblies use first past the post, which is the source of India’s persistent gap between vote share and seat share and of the Duverger exception — plurality rules coexisting with sustained national multipartism.
• Indirect elections use STV. The Rajya Sabha and the state legislative councils are elected by proportional representation by means of the single transferable vote by the elected members of the relevant assemblies.
• The presidential election under Article 55 is described in the Constitution as proportional representation by means of the single transferable vote, and the description carries an anomaly worth stating plainly.
  • Proportional representation requires a multi-member constituency, and the President is a single office, so this cannot be proportional representation in the ordinary sense.
  • What actually operates is a quota system that resolves to an absolute majority of the value of votes polled, with transfer of preferences — making the mechanism substantively an alternative vote carried out with weighted votes, though constitutionally labelled proportional.
  • The election also uses a weighted-value system: MLAs’ votes are weighted by their state’s population and the number of elected members, and MPs’ votes by the total value of MLAs’ votes, so that the electoral college balances Union and states.
• The reform debate has recurred for decades without resolution.
  • The case for change: India’s results are severely disproportional, small and dispersed minorities are under-represented, and governments regularly command large majorities on minority vote shares.
  • The proposals have generally been for a mixed system adding a list element rather than for full proportional representation.
  • The case against: PR would produce chronic coalition instability in a society with India’s number of cleavages, and would sever the constituency link on which Indian representation is popularly understood to rest.
• Delimitation is the live institutional question: constituency boundaries and the distribution of seats between states have been frozen against population change, and the eventual unfreezing raises a genuine federal dilemma between equal representation of persons and equal treatment of states that reduced their population growth.

An electoral system is a constitutional choice disguised as an administrative one: it decides not how votes are counted but which votes are worth casting.

Methods of minority representation

• Every democracy with a persistent minority must answer the question that ordinary elections cannot: how a group that is permanently outvoted can be represented at all.
• Proportional representation — the general solution, and the one most systems adopt. It works within a joint electorate, so it produces representation without partitioning the electorate itself, and it requires no official classification of who belongs to which group.
• Separate electorates — the minority votes in its own roll for its own reserved seats.
  • Introduced in India by the Morley–Minto reforms of 1909 and extended thereafter, and rejected at independence.
  • The objection is that separate electorates make the representative accountable only to the minority, remove any incentive for cross-community appeal, and institutionalise division — which is why they are regarded as a threat to national unity and why proportional representation is preferred as achieving the same end without the same cost.
  • The Poona Pact of 1932 is the decisive Indian episode: the Communal Award’s separate electorates for the Depressed Classes were replaced, after Gandhi’s fast, with an enlarged number of reserved seats within a joint electorate — the model India retained.
• Reservation of seats — seats are set aside for the group but elected by the whole electorate, as with Scheduled Castes and Scheduled Tribes in Indian legislatures, and as New Zealand‘s Māori seats do in a different form.
• Candidate quotas — parties are required, by statute or by their own rules, to field a minimum share of candidates from the group; used extensively for women’s representation, and the mechanism behind India’s 106th Amendment reserving seats for women, whose implementation awaits a census and delimitation.
• Consociational devices — Lijphart‘s four instruments for deeply divided societies: grand coalition of all significant segments, segmental autonomy, proportionality in representation and public appointments, and a mutual veto for each segment on vital matters.
• Centripetal devices — Horowitz‘s alternative, which distrusts consociationalism for freezing group identities. The aim is to build electoral incentives for moderation: preferential systems that require candidates to attract second preferences across communal lines, distribution requirements that oblige a presidential candidate to win support in many regions, and constituency boundaries drawn to be heterogeneous.
• Best-loser and reserved-list systems — Mauritius allocates additional seats to the best-performing unsuccessful candidates from under-represented communities, a technique for correcting outcomes without partitioning the electorate.

How electoral democracy augments participation

• The mechanisms by which the ordinary conduct of elections widens participation are worth setting out, because they are usually assumed rather than stated.
  • The franchise constitutes a mass public. An election is the only occasion on which a state must ask its entire adult population a question and act on the answer.
  • Competition compels mobilisation. Parties in close contests must find, register, inform and transport voters they would otherwise ignore, so competitiveness is itself a participation-generating mechanism.
  • Periodicity creates a rhythm of political attention, campaigning and argument, which pulls in people who are otherwise inattentive.
  • Elections create secondary participation — canvassing, campaigning, party membership, polling agency, observation, and the vast informal activity around a campaign.
  • Local and tiered elections multiply the points of entry. India’s panchayat and municipal elections create hundreds of thousands of elected positions, and with reservation they have brought into office people who would never have contested at any other level.
  • Elections generalise political language. Rights, entitlements and grievances acquire a public vocabulary through campaigning, which is a precondition of claim-making outside elections too.
• The limits should be stated with equal clarity.
  • Elections deliver episodic and blunt participation — a single undifferentiated judgement on a bundle of issues every few years.
  • Electoral participation is not equal participation where turnout is class-graded, and where money, media access and candidacy costs are not.
  • Electoral authoritarianism demonstrates that elections can be held routinely while participation is a ritual — which is why the contemporary literature measures the quality of elections rather than their occurrence.
  • The activity that elections generate can crowd out continuous participation, by concentrating political energy into campaign periods.

Democratic backsliding and democratic backlash

• The contemporary world exhibits two trends that appear to contradict each other and in fact reinforce each other, which is what makes the pairing a paradox rather than an inconsistency.

Backsliding: what it is and how it works now

• Democratic backsliding is the state-led debilitation or elimination of the political institutions that sustain an existing democracy — decline driven from inside the state by people who acquired power legitimately.
• Three subtypes are usefully distinguished.
  • Democratic erosion — gradual, piecemeal weakening of established practices: judicial independence, media freedom, the neutrality of the civil service, the integrity of the electoral administration.
  • Democratic breakdown — rapid, multidimensional collapse: dissolving branches of government, banning opposition parties, suspending the constitution.
  • Autocratic deepening — the narrowing of what little space remains in states that were already authoritarian.
• Nancy Bermeo‘s central contribution is the observation that the forms have changed, and that measuring today’s decline with yesterday’s indicators makes it invisible.
  • The classic coup d’état, election-day fraud and the executive coup — the twentieth century’s characteristic instruments — have all declined sharply.
  • What has risen instead is executive aggrandisement: the incremental, legal weakening of checks by an elected executive, each step defensible in isolation, none of them a rupture that would trigger a response.
  • Strategic electoral manipulation — controlling media, harassing opponents, altering rules, using state resources — has replaced ballot-box fraud, because it produces the same result while surviving observation.
  • Promissory coups — seizures justified as restoring democracy and promising elections — have replaced open seizures of power.
• Steven Levitsky and Daniel Ziblatt describe the sequence: elected leaders capture the referees (courts, prosecutors, electoral bodies, regulators), sideline the opposition through legal and financial pressure, and rewrite the rules to entrench themselves — while the informal norms of mutual toleration and institutional forbearance that made the written rules work quietly erode.
• Kim Lane Scheppele‘s autocratic legalism names the technique that makes it hard to resist: the entire process is conducted through law, using constitutional amendment, ordinary legislation and judicial appointment, so that every step is formally lawful and no single step is a coup.
• The comparative record is stark, and the figures discipline any complacent reading.
  • Roughly a quarter of the world’s countries are currently assessed as backsliding or autocratising, against a much smaller number democratising.
  • The global level of democracy has been pushed back to approximately where it stood in the late 1970s, before the third wave began.
  • Freedom of expression is the most common first target of autocratising leaders over the last twenty-five years, and shows the sharpest global decline of any indicator; rule of law is deteriorating in more than twenty countries.
  • Most strikingly, a majority of the newly autocratising cases are now in Europe and North America — which ends the assumption that backsliding is something that happens to developing societies.

Backlash: two meanings, both real

• The term carries two distinct meanings, and the paradox in the pairing dissolves once they are separated — and then reappears in a more interesting form.
• Backlash against liberal democracy — a popular and elite reaction against the liberal component of liberal democracy: minority rights, judicial constraint, supranational authority, cultural pluralism and technocratic governance.
  • Its drivers are those of the populist right generally: economic insecurity, Norris and Inglehart’s cultural backlash, migration salience, and the convergence of governing alternatives.
  • It is majoritarian rather than anti-democratic in its own self-understanding: it claims to be restoring popular sovereignty against institutions that frustrate it, which is why it wins elections and why the language of democracy is available to it.
• Backlash against backsliding — pro-democracy mobilisation by voters, courts, oppositions, civil society and international actors, which has been substantial and in several cases successful.
  • Poland in 2023, where an opposition coalition framed the election as a referendum on democracy and removed a backsliding government.
  • Brazil in 2022, where an incumbent was defeated and an attempt to overturn the result failed, with courts and a very broad civil-society coalition — hundreds of thousands signing a public defence of the democratic rule of law — playing decisive roles.
  • Zambia in 2021 and Senegal in 2024, where an attempt to postpone an election was blocked by the constitutional court and mass mobilisation.
  • The recovery playbook that emerges from these cases has four components: restoring democratic norms at the top, reinstating civic freedoms, pursuing accountability for corruption, and reforming the institutions that were captured — the last being the slowest and the most often abandoned.
  • The characteristic danger of recovery is that the incoming government inherits the concentrated powers its predecessor built, and finds reasons not to dismantle them.

Explaining the paradox

• The two processes have different agents and different levels. Backsliding is elite-led and institutional; backlash is mass and mobilisational. There is no contradiction in an executive concentrating power while a public resists it — that is what a contested political process looks like.
• They target different things.
  • Support for democracy as an idea remains extremely high almost everywhere, including among voters who elect anti-pluralist parties.
  • Those voters are not choosing autocracy: they are choosing a party on other grounds and discounting or disbelieving the democratic cost that observers attribute to it.
  • This is why backsliding is possible in societies with overwhelming pro-democracy majorities, and why appeals to defend democracy in the abstract change so few votes.
• The mechanism of backsliding generates its own opposition. Because erosion is legal and gradual, it produces no single moment of rupture around which resistance can coordinate — but it accumulates grievances continuously, so resistance builds without a trigger and eventually arrives at an election.
• Elections are the instrument of both. The same institution that backsliders use to legitimate themselves is the one that removes them, which is why the crucial variable is whether elections remain competitive enough to be lost. Every recovery case is a case where they were.
• Backlash against liberalism and backlash against backsliding feed one another. Illiberal governance provokes pro-democracy mobilisation; pro-democracy mobilisation is framed by incumbents as elite obstruction of the popular will, which strengthens the illiberal claim — a loop that produces polarisation as the stable outcome.
• The measured conclusion: democracy is not simply declining, it is being contested more intensely, on more fronts, and in more places than at any point since the third wave began.
  • The outcome in each country turns less on economic development or political culture than on institutional capacity to resist — whether courts, electoral administrations and oppositions remain able to act.
  • The second variable is coalitional: whether a broad enough alliance can be assembled to defend procedures that no single group owns and every group needs.

Backsliding and backlash are not opposite trends but two halves of one process: erosion works precisely because it is gradual and legal, and it is gradual and legal precisely because resistance is real.

Conclusion

The classical theory of representative democracy assumed that the hard problems were at the entrance — who may vote, and how votes become seats. Those problems were solved, and harder ones appeared behind them.

• Representation turns out to have four incompatible meanings, and a system can satisfy any one of them while failing the others; descriptive representation matters because the politics of ideas demonstrably did not deliver for the excluded.
• Participation has not simply declined; it has become more unequal in the advanced democracies and more equal in some developing ones, which reverses the expected relationship between wealth and democratic quality and is the most interesting comparative finding in this field.
• Electoral systems are not administrative details but constitutional choices, and every one of them buys some goods at the cost of others — proportionality against accountability, minority inclusion against governability, the local link against the fit between votes and seats.
• The current danger is not that elections stop being held but that they stop deciding anything, which is what executive aggrandisement is designed to achieve without ever taking the step that would be recognised as a coup.
• The recovery cases establish the one hopeful proposition available: backsliding is reversible where elections remain competitive enough to be lost, which makes the integrity of the electoral system — the technical question the classical theory thought it had settled — the decisive political question of the present.