Democratic Regimes

Core Concepts

What Is a Democracy, Exactly?

Liberal democracy is the dominant reference frame in comparative politics, but it is not a single thing. It is more useful to think of it as a layered structure: a set of procedural guarantees at the bottom, institutional arrangements built on top, and structural preconditions that determine whether the whole edifice holds.

Three classic definitions divide what "democracy" means:

Schumpeter's minimal definition. Democracy is an institutional arrangement through which individuals acquire political decision-making authority through competitive struggle for votes. Schumpeter strips away any claim about popular will or collective reason — what matters is whether elites rotate through contested elections. This remains the baseline for most empirical measures.

Dahl's polyarchy. Robert Dahl argued that actual existing democracies are best called "polyarchies" — approximations of the democratic ideal rather than its realization. He identified seven institutional guarantees required to operationalize liberal democracy: inclusive suffrage, the right to run for office, free and fair elections, freedom of expression, access to alternative information sources, associational autonomy (including legitimate opposition parties), and governmental responsiveness. These operationalize five deeper procedural requirements: effective participation, voting equality, enlightened understanding, control of the agenda, and inclusion of all adults. Crucially, Dahl distinguishes two dimensions of democratization: liberalization (the expansion of contestation and rights) and inclusiveness (the expansion of the franchise). Regimes can be more or less advanced on either dimension independently.

The V-Dem multidimensional approach. The Varieties of Democracy project operationalizes democracy across five distinct principles — electoral, liberal, participatory, deliberative, and egalitarian — collected through approximately 3,500 country experts coding more than 450 indicators. The Electoral Democracy Index covers free elections, expression, association, and suffrage. The Liberal Democracy Index adds checks on the executive, civil liberties, rule of law, and judicial independence. The key insight from V-Dem's methodological design is that these dimensions do not move together: economic development may correlate with electoral democracy while correlating inversely with egalitarian democracy as inequality rises.

Dahl's polyarchy is not a description of how democracies work. It is a minimum specification of what they must have to work at all.

The Institutional Stack

Liberal democracy is not reducible to elections. It requires an institutional stack, each layer supporting the one above it.

Free and fair elections are the necessary base. They require universal adult suffrage, transparent procedures, protection against fraud, freedom to campaign, freedom of association, and access to independent information. Elections authorize representatives and create accountability through the credible threat of electoral loss. But elections alone are insufficient — they can be held in contexts so distorted by intimidation, media monopoly, or gerrymandering that they produce no meaningful accountability.

Multi-party competition gives elections substance. Legitimate political opposition — parties with legal recognition, ballot access, and protection from persecution — distinguishes liberal democracy from competitive authoritarianism, where nominal multi-party systems exist but real contestation is suppressed. Electoral systems shape this layer: proportional representation tends to sustain stable multi-party competition; first-past-the-post tends toward two-party dynamics.

Separation of powers provides the checks that prevent winners from making themselves permanent. Montesquieu's formulation — liberty is threatened when powers concentrate — became constitutional doctrine across liberal democracies, dividing government into legislative, executive, and judicial branches that can each constrain the others.

Judicial independence makes separation of powers real. A court that rules against the government can only do so if it is genuinely independent. Both individual independence (security of tenure, protection from removal) and institutional independence (autonomy from executive and legislative branches) are required. Judicial independence is not an end in itself but a precondition for impartial enforcement of constitutional constraints.

Rule of law binds both government and citizen to enforceable legal frameworks. It implies an independent judiciary, equal protection of rights, and civil liberties protections. Critically, rule of law is distinct from rule by law — a government can use law instrumentally to consolidate power while gutting its constraining function.

Constitutional constraints cap the stack. Rigid constitutions protected by judicial review restrict temporary majorities from altering fundamental institutions or infringing basic liberties. Supermajority amendment requirements make it costly to constitutionalize changes that serve particular factions. Constitutional design varies: some systems use hard-written constitutions; others rely on constitutional conventions and parliamentary sovereignty. But in all variants, constitutional principles establish limits on governmental power.

Civil society is the external enforcement mechanism. Voluntary associations, a free press, NGOs, unions, and professional organizations independent of the state monitor government conduct, organize collective action, and create deliberative space outside formal institutions. Civil society is not a nice-to-have: it is a structural precondition for democratic accountability.

Two Models: Westminster vs. Consensus Democracy

Liberal democracies do not all implement the same institutional design. Lijphart's classic typology distinguishes two poles.

The Westminster model (majoritarian democracy) concentrates power in a single-party majority executive responsible to parliament. Its defining feature is executive accountability through votes of no confidence: the government survives only as long as it retains the confidence of the legislature. Parliamentary debate is adversarial — opposition parties challenge the government directly on the floor. The system originated in seventeenth-century England and spread through the British Empire. It produces decisive government but amplifies majority preferences, potentially marginalizing substantial minorities.

Consensus democracy disperses power through institutional arrangements designed to require broad agreement. Rather than simple majority rule, consensus systems use supermajority thresholds, power-sharing arrangements, and veto points to ensure that a wider range of interests must be accommodated before major decisions are implemented. The Northwest Territories and Nunavut in Canada operate formal consensus government models; proportional representation systems in continental Europe create multi-party coalitions that function as de facto consensus-forcing mechanisms.

Fig 1

The tradeoff is real and not resolvable by theory alone. Westminster systems decide faster and assign accountability more cleanly. Consensus systems include more interests but can deadlock, and diffuse accountability makes it harder for voters to identify who is responsible for outcomes.

Two Theories of Legitimacy: Electoral vs. Deliberative

Liberal democracy usually grounds legitimacy in elections — the governed authorize rulers through periodic competitive voting. But a second tradition argues this is insufficient.

Electoral legitimacy (the Schumpeterian baseline) holds that a government is legitimate if it won a free and fair election. The mechanism is retrospective: governments that disappoint can be voted out. This is procedurally clean and empirically observable, but it has a well-known deficiency: it aggregates preferences without requiring that those preferences be informed or reasoned.

Deliberative legitimacy holds that legitimate laws derive their authority primarily from public deliberation — authentic reasoning conducted under conditions of equality, inclusion, and reciprocity — rather than electoral outcomes alone. Habermas's ideal speech situation formalizes the minimum conditions: all competent parties may participate; any assertion can be questioned; no speaker is prevented by coercion from exercising these rights. Joshua Cohen extends this into a political theory: democratic power should be tied to conditions of inclusive reason-giving where all affected parties have equal standing.

The two theories are not simply opposed — most functioning democracies use elections to select decision-makers while relying on deliberative processes (legislative debate, judicial review, public consultation) to govern what those decision-makers can actually do.

Social Capital: The Invisible Precondition

Robert Putnam's research adds a layer that formal institutional accounts tend to miss. In Making Democracy Work (1993), his analysis of Italian regional governments showed that dense horizontal civic associations — networks, norms of civic engagement, and interpersonal trust — correlate strongly with institutional performance. Regions with high social capital governed more effectively, not primarily because of formal institutional differences, but because of the fabric of cooperation underlying formal institutions.

In Bowling Alone (2000), Putnam documented a substantial collapse in American civic participation since the 1960s and linked it to weakened democratic institutions and declining political participation.

The theory distinguishes:

• Bridging social capital: ties across diverse groups, across class, ethnicity, religion — most important for democratic stability because it creates the cross-cutting trust that allows political losers to accept outcomes and remain within the system.
• Bonding social capital: inward-looking ties within homogenous groups, which can reinforce group identity and mutual aid but do not build the cross-group trust democracy requires.

The limits of Putnam's framework are real: it can be read as blaming democratic failure on civil society rather than on structural political failures, and it does not explain how social capital is built from scratch in new polities.

---

Worked Example: Democratic Backsliding as Institutional Targeting

Understanding backsliding concretely requires moving past the idea that democracies simply "collapse." Contemporary scholarship shows that backsliding operates through incremental, targeted attacks on specific institutional pillars while maintaining the surface appearance of democratic procedure.

The Playbook

The pattern is consistent across cases. Backsliding executives use formally legal mechanisms — statutory reappointments, constitutional amendments, emergency powers declarations — to achieve institutional subordination. The targets follow a logic: attack the institutions that could constrain you, while retaining the institutions that legitimate your authority.

Step 1: Judicial capture. Replace judges with loyalists, restructure or pack courts, use constitutional amendments to alter appointment processes. Once courts are captured, the rule of law constraint is removed and further steps become easier.

Step 2: Media restriction. Use licensing restrictions, ownership consolidation, defamation laws, and tax audits of critical outlets. Without independent media, alternative information sources — one of Dahl's seven requirements — degrade, making enlightened understanding impossible.

Step 3: Electoral distortion. Gerrymander constituencies, capture electoral commissions, introduce voter registration barriers, or restrict campaign access for opposition. This maintains the electoral form while gutting the competitive content.

Step 4: Civil service politicization. Purge career officials and replace them with loyalists. This disables the independent bureaucratic capacity that normally implements law neutrally and provides an institutional resistance point.

The Three Pathways

Recent research identifies three institutional pathways through which backsliding culminates in executive aggrandizement:

1. Legislative capture: Weakening parliament's oversight capacity, converting it from a check into a rubber stamp.
2. Plebiscitary override: Bypassing existing institutions through direct popular appeals — referenda, plebiscites — that claim democratic legitimacy while circumventing constitutional procedures.
3. Direct power grabs: Using emergency powers, security crises, or constitutional moments to concentrate authority outside normal channels.

The Structural Context

These elite maneuvers do not happen in a vacuum. Four structural conditions consistently create environments where backsliding is feasible: political polarization, racism and nativism, economic inequality, and excessive existing executive power. Income inequality increases polarization, which increases the likelihood of backsliding through a compounding mechanism.

Nearly half of all autocratizing governments — 21 of 45 countries in active backsliding episodes as of 2024-2025 — are actively spreading disinformation. Political polarization is rising in approximately a quarter of all countries. Disinformation and polarization may be symptoms rather than root causes, but they reduce public resistance by eroding the shared epistemic basis that democratic contestation requires.

The Scale

V-Dem's 2024-2025 data shows 45 countries in active autocratization episodes, against only 19 experiencing democratization. Of the 45 autocratizing countries, 27 were democracies at the start of their episodes, and 18 of those have already completed the transition to autocracy — a roughly 70% fatality rate once backsliding begins. By 2026, global democracy levels had returned to their 1978 baseline, erasing most post-Cold War democratic progress.

The Defense: Constitutional Courts

Constitutional courts function as critical institutional buffers when they maintain secure tenure and sufficient independence from the executive. Romania's Constitutional Court nullified a presidential election after declassified intelligence documented foreign interference. Brazil's judicial system constrained executive overreach during a period of sustained pressure. The condition for court resilience is the same condition backsliding targets first: judicial independence. This creates a race: the executive tries to capture courts before courts can block the executive.

---

Common Misconceptions

"Democracy causes development"

The intuitive claim — that democracies outperform authoritarian regimes economically — has weak causal support. Przeworski's research inverts the causal direction: economic development does not explain when democracies emerge, but it does explain whether democracies, once established for other reasons, survive. Wealthier societies sustain democracy more reliably — not because wealth creates democracy, but because it stabilizes it. Democratic emergence is path-dependent; democratic survival is wealth-dependent. These are separate questions with separate answers.

This matters for a polity designer: building democratic institutions in a low-income, low-social-capital context is not the same problem as consolidating democracy in a wealthy society. The empirical record on democratic survival in poor countries is sobering.

"Elections are democracy"

Elections are necessary but not sufficient. The full Dahl polyarchy framework specifies seven conditions, and elections are one. Without alternative information sources, associational autonomy, and governmental responsiveness, elections produce what Zakaria famously called "illiberal democracy" — majority-empowered governments unrestrained by rights or institutional checks. V-Dem's indices explicitly disaggregate electoral from liberal democracy for this reason: many countries score high on the former while scoring low on the latter.

"Consolidation means resilience"

A consolidated democracy is not automatically a resilient one. Recent scholarship distinguishes the two: consolidation is the establishment of enduring democratic institutions; resilience is the capacity to absorb, recover, and adapt under pressure. A formally consolidated democracy with thin civil society and weakly independent courts may prove brittle; a less consolidated but civic-rich democracy may absorb significant stress. Six of the ten newly autocratizing countries in 2025-2026 are in Europe and North America — previously considered consolidated — demonstrating that consolidation does not confer immunity.

"Strong civil society prevents backsliding"

Putnam's framework shows that civil society supports democratic performance, but civil society is not a neutral actor. In weakly institutionalized democracies, political elites can capture civil society organizations, activating their mobilizing capacity in anti-democratic directions. Civil society organizations that generate bonding rather than bridging social capital can intensify polarization. A dense civil society is a democratic asset under normal conditions, but it does not guarantee democratic outcomes when elites successfully weaponize parts of it.

---

Compare & Contrast

Westminster vs. Consensus Democracy

Dimension	Westminster (Majoritarian)	Consensus
Power distribution	Concentrated in single-party majority executive	Dispersed across coalition partners and veto players
Decision speed	Fast; majority governs decisively	Slow; requires broad agreement
Accountability	Clear; voters can identify who is responsible	Diffuse; coalition arrangements obscure responsibility
Minority inclusion	Losers fully excluded from power	Losers often included in coalition arrangements
Backsliding risk	Higher; power concentration accelerates capture	Lower; veto players create friction
Examples	UK, New Zealand (pre-MMP), Australia	Netherlands, Switzerland, Nordic states

Sources: Westminster model; Consensus model

Electoral Legitimacy vs. Deliberative Legitimacy

Dimension	Electoral	Deliberative
Source of authority	Competitive vote; majority preference	Public reason-giving under conditions of equality and inclusion
What it validates	Who governs	What rules are acceptable to govern all
Minority status	Outvoted minorities must accept outcomes	Minorities retain standing to contest decisions through reason
Empirical basis	Measurable, observable	Hard to operationalize; ideal conditions rarely met
Failure mode	Elected government with no deliberative check enacts majoritarian preferences against minorities	Deliberative forums can be captured by the articulate and already-empowered
Canonical theorists	Schumpeter	Habermas, Cohen

Schumpeter vs. Dahl: What Counts as Democracy?

Schumpeter's definition — competitive elite struggle for votes — sets the minimum. Dahl's polyarchy adds five requirements beyond contested elections: freedom of expression, alternative information, and associational autonomy (i.e., real civil and political liberties). The practical difference is significant: a country can satisfy Schumpeter's definition while severely restricting press freedom, opposition organizing, and civil liberties. It cannot satisfy Dahl's definition under those conditions. Most contemporary measurement frameworks — including V-Dem's Liberal Democracy Index — are Dahlian in spirit, not Schumpeterian.

---

Boundary Conditions

When Does the Polyarchy Framework Not Apply?

Dahl's framework was built from Western liberal democracies. Several conditions mark its limits:

New or post-conflict polities with thin institutional foundations. The polyarchy framework assumes the prior existence of a functioning state, a legal system capable of enforcing rules, and some baseline of social trust. In states where these prerequisites are absent, installing democratic institutions without building the underlying substrate tends to produce unstable hybrids. Przeworski's insight about wealth and survival is particularly sharp here: the probability that a democratic regime survives at very low income levels is low regardless of institutional design quality.

Deeply divided societies. Where political identity maps entirely onto ethnic, religious, or regional cleavages, majoritarian democratic arrangements institutionalize permanent winners and losers. Lijphart's own argument is that consensus democracy is better suited to plural societies — but even consensus arrangements can fail when cleavages are too deep for any coalition to bridge.

The Global South legitimacy gap. Post-independence democracies were premised on two promises: individual rights combined with markets would produce inclusive growth, and electoral voting would translate citizen needs into policy. Across much of the Global South, neither promise has reliably materialized. Elite capture, accountability gaps, and state capacity deficits mean that formal democratic procedures coexist with substantive democratic failure. This is not a malfunction of the polyarchy framework — it is a boundary condition marking where that framework requires additional structural support to produce the outcomes it promises.

When Does Public Choice Analysis Change the Picture?

The public choice tradition (Buchanan and Tullock) applies economic analysis to political decision-making and reveals systematic pathologies inside democratic institutions that neither Dahl's framework nor deliberative theory fully addresses.

Politicians respond to incentives, not public interest. Political actors pursue career advancement, electoral support, and personal benefit. Government failure — like market failure — is predictable and structural, not the result of poor personnel choices. Any polity design that assumes benevolent, well-informed officials will execute public policy in the general interest is operating on an empirically false premise.

Organized minorities beat disorganized majorities. Small, organized interests almost always outcompete large diffuse interests in the policymaking process. Concentrated interests face favorable organizational incentives; diffuse majorities face collective action problems. This is not incidental to democratic politics — it is structural, following logically from the asymmetry between concentrated and diffuse stakes.

Rent-seeking is pervasive. Rational actors allocate resources to political influence rather than productive innovation when the expected return on lobbying exceeds the expected return on production. Regulatory capture — where agencies adopt policies that favor incumbent industries — is a predictable consequence of political incentive structures, not an aberration.

Constitutional constraints are the public choice response to public choice problems. Buchanan's constitutional economics argues that the solution to systematic political failure is not better politicians but better rules. Constitutional frameworks act as pre-commitment devices: by constraining what governments can do, they reduce the scope for opportunistic behavior. The logic of designing constitutions under uncertainty — where designers do not know their future position under the rules they create — is that rational self-interest aligns with choosing general rules that benefit all parties.

The Structural Precondition Question

Three major structural accounts of how democracies emerge — Moore, Skocpol, Acemoglu-Robinson — agree on one thing: democratic institutions do not emerge from institutional design alone. They emerge from structural conditions that determine whether those institutions are sustainable.

Barrington Moore argued that class configuration during industrialization determines the path: a strong, independent bourgeoisie is a necessary condition for liberal democracy ("no bourgeoisie, no democracy"). Skocpol showed that state structures have autonomous capacity independent of class — state crises driven by international competitive pressures matter independently of class dynamics. Acemoglu and Robinson reframe the question from class to institutions: inclusive political and economic institutions — those that enforce property rights, create level playing fields, and distribute power broadly — produce both prosperity and democratic stability, while extractive institutions create vicious cycles of authoritarianism and stagnation.

For a polity designer, the structural accounts are both sobering and useful. They are sobering because they suggest that institutional design can only accomplish so much in the absence of the structural preconditions. They are useful because Acemoglu and Robinson's framework is explicitly about design: the choice is between extractive and inclusive institutional architectures, and that choice has predictable long-run consequences.