Why do some countries sustain democracy while others collapse into authoritarian rule? Why do Scandinavian welfare states look so different from American social policy, even though both emerged from industrialized capitalist economies with competitive party systems? Why did France, Russia, and China produce social revolutions while structurally similar societies did not? These are among the central questions of comparative politics -- the subfield of political science that attempts to explain variation in political outcomes by systematically comparing cases across countries and historical periods.
The comparative method is not simply reading about different countries. It is the disciplined application of research designs that can distinguish causes from correlations, identify the mechanisms through which institutional arrangements shape outcomes, and generate findings that travel beyond single cases. From John Stuart Mill's nineteenth-century articulation of the methods of agreement and difference, to the large-N statistical work enabled by the V-Dem project's two centuries of cross-national democracy data, comparative politics has developed a sophisticated methodological toolkit for addressing questions that neither pure theory nor single-country expertise can answer alone.
This article covers the methodological foundations of comparative politics, the major empirical research programs on democracy, electoral systems, welfare states, state capacity, party systems, and revolutions, and the contemporary debates about democratic backsliding, competitive authoritarianism, and the comparative assessment of contentious politics.
"Political scientists have traditionally studied politics in individual countries. Comparative politics is the attempt to move beyond that constraint by studying variation across countries and drawing general conclusions from it." -- Arend Lijphart, 'Comparative Politics and the Comparative Method' (1971)
| Research Question | Comparative Method | Example Studies |
|---|---|---|
| Why do democracies survive or fail? | Cross-national statistical analysis | Lipset (1959), Przeworski (2000) |
| How do electoral systems affect governance? | Paired country comparisons | Lijphart (1999), Proportional vs. majoritarian systems |
| What causes civil war? | Large-N quantitative datasets | Fearon & Laitin (2003) |
| Why do some states develop faster? | Historical case studies | Acemoglu & Robinson (2012) |
| How does federalism work? | Federal vs. unitary system comparison | Riker (1964) |
Key Definitions
Polyarchy: Robert Dahl's term for real-world democratic approximations, defined by meaningful opposition rights, regular free elections, and guaranteed civil liberties. Distinguished from democracy as a full ideal.
Most similar systems design: A comparative method that compares cases alike on most variables but differing on the outcome, isolating the variable responsible for the difference.
Most different systems design: A comparative method that compares cases differing on most variables but sharing the outcome, identifying what they have in common as a potential cause.
Process tracing: A method for identifying causal mechanisms within cases by reconstructing the sequence of events through which a cause produces an effect.
Decommodification: Esping-Andersen's concept of the degree to which individuals can maintain a socially acceptable living standard independent of market participation, as a function of welfare state generosity.
The Comparative Method
Mill's Methods and Their Application
John Stuart Mill's 'A System of Logic' (1843) articulated two methodological designs that remain foundational to comparative politics. The method of agreement -- corresponding to what Arend Lijphart later called the most-different systems design -- compares cases that share an outcome despite differing on most other variables. If democratic stability is found across countries as culturally, economically, and geographically diverse as India, Japan, and Germany, the features these countries share are candidate causes. The method of difference -- corresponding to the most-similar systems design -- compares cases that are similar on most variables but differ on the outcome. If two neighboring countries with similar economic structures, colonial histories, and cultural backgrounds diverge dramatically in their political trajectories, the variables on which they differ are candidates for explanation.
Lijphart's 1971 article 'Comparative Politics and the Comparative Method' in the American Political Science Review formalized these designs for the discipline and argued that structured case comparisons could provide causal leverage comparable to that of large-N statistical analysis, provided cases are selected carefully to control for confounding variables. His own work on consensus versus majoritarian democracies, synthesized in 'Patterns of Democracy' (1999), compared 36 democracies over four decades, finding consistent associations between institutional design and policy outcomes.
Process Tracing and Within-Case Analysis
Process tracing, developed as a methodology by Alexander George and Andrew Bennett and others, addresses a limitation of cross-national comparison: even when cases differ on the outcome and on a candidate cause, this correlation does not establish that the cause produced the outcome through any specific mechanism. Process tracing traces the sequence of events between cause and outcome within individual cases, looking for observable implications of the proposed causal pathway -- the 'smoking gun' that confirms a specific mechanism as opposed to competing alternatives. It is particularly valuable in historical and area studies work, where deep knowledge of particular cases allows reconstruction of causal chains in detail that aggregate statistical analysis cannot provide.
Large-N Statistical Analysis and Cross-National Datasets
The development of systematic cross-national datasets has transformed what comparative politics can demonstrate. The Freedom House ratings (launched 1972), Polity project scores, and above all the V-Dem project at the University of Gothenburg (launched 2014 by Michael Coppedge, John Gerring, Staffan Lindberg, and colleagues) have made it possible to trace patterns in democratic quality, civil liberties, electoral integrity, and judicial independence across 180 countries over 200 years. These datasets enable regression analysis that tests whether proposed causal variables predict outcomes after controlling for potential confounders. The persistent challenges of endogeneity (political outcomes also influence the institutional variables proposed to cause them), reverse causation, and omitted variable bias mean that statistical findings in comparative politics require theoretical interpretation and ideally triangulation with qualitative evidence.
Democracy: Measurement, Causes, and Decline
Dahl, Polyarchy, and the Measurement Enterprise
Robert Dahl's 'Polyarchy: Participation and Opposition' (1971) provided the conceptual foundation for systematic democracy measurement. Dahl argued that full democracy -- complete popular self-governance -- is an unrealized ideal, and that the empirically relevant variable is polyarchy: the degree to which governments face meaningful opposition, citizens can organize and vote, and civil liberties are protected. Two fundamental dimensions structure the space: the extent to which political opposition is permitted (liberalization) and the proportion of the population entitled to participate (inclusiveness). Different historical pathways to polyarchy -- democratization of oligarchies, extension of participation in closed hegemonies, combined liberalization and extension -- produce different vulnerabilities and strengths. Dahl's framework guided Freedom House's operationalization and influenced the V-Dem project's multi-dimensional measurement approach.
Larry Diamond's work, including 'The Spirit of Democracy' (2008), documented what he termed the democratic recession beginning around 2006: after the third wave of democratization that swept Southern Europe, Latin America, and Eastern Europe from the mid-1970s to the early 2000s, the global trend reversed. More countries each year were experiencing democratic decline than democratic advance, a pattern that the V-Dem project's data confirmed with increasing precision.
Competitive Authoritarianism and Democratic Backsliding
Steven Levitsky and Lucan Way's 'Competitive Authoritarianism' (2010) identified a distinctive regime type that became common after the Cold War: hybrid regimes that maintained the formal institutions of democracy -- elections, opposition parties, independent media -- but in which the playing field was systematically tilted in favor of incumbents through media control, selective prosecution of opponents, manipulation of electoral rules, and intimidation of civil society. These regimes were neither fully democratic nor fully authoritarian; they were competitive in that the outcome of elections was not entirely predetermined, but competitive in conditions sufficiently unfair that democratic accountability was severely compromised.
Anna Lührmann and Staffan Lindberg's 2019 analysis using V-Dem data found that 25 countries were experiencing autocratization as of their study period, with the average level of democracy in these countries declining significantly. Hungary under Viktor Orban after 2010 and Turkey under Recep Tayyip Erdogan after 2013 became the most extensively studied cases of democratic backsliding, illustrating how governments elected democratically could systematically erode judicial independence, media pluralism, civil society autonomy, and electoral integrity without formally abolishing democratic institutions. Levitsky and Daniel Ziblatt's 'How Democracies Die' (2018) argued that this kind of gradual erosion is the modal form of contemporary democratic death, driven by the abandonment of informal norms of mutual toleration and institutional forbearance.
Electoral Systems and Institutional Design
Duverger's Law
Maurice Duverger's observation in 'Political Parties' (1951), formalized as Duverger's Law, holds that plurality electoral systems -- those in which the candidate with the most votes wins regardless of whether a majority is obtained -- tend to produce two-party systems. Two mechanisms drive this tendency. The mechanical effect: because plurality rules disproportionately penalize third parties whose votes are spread across many constituencies, they win far fewer seats than their vote share would justify under proportional rules. The psychological effect: voters and donors anticipating wasted votes converge on the two frontrunners, reducing support for smaller parties before votes are cast. The law is probabilistic rather than deterministic -- India's plurality system produced a multiparty system, and the United States' system has seen significant third-party activity in different eras -- but the overall pattern is well-supported across comparative evidence.
Proportional representation (PR) systems allocate seats to parties roughly in proportion to vote shares, enabling multiple parties to win representation and typically producing multiparty systems. Germany's mixed-member proportional system, which combines single-member districts with party list seats to achieve overall proportionality, is often cited as combining local constituency accountability with systemic representational fairness.
Lijphart's Consensus vs. Majoritarian Democracies
Arend Lijphart's 'Patterns of Democracy' (1999) provided the most comprehensive systematic comparison of democratic institutional designs. Lijphart distinguished majoritarian democracies -- in which decision-making authority is concentrated in a single-party majority government elected by plurality rules -- from consensus democracies -- in which power is shared through coalition governments, proportional representation, bicameral legislatures, strong federalism, and independent central banks. The Westminster model (United Kingdom, New Zealand pre-1996) exemplifies majoritarian democracy; Switzerland, Belgium, and the Netherlands exemplify consensus democracy. Lijphart found that consensus democracies achieved better performance on a range of social indicators: higher voter turnout, more income redistribution, stronger environmental regulation, lower incarceration rates, and more generous social spending. Whether these differences reflect the causal effects of institutions or correlated cultural and economic factors is an ongoing debate, but the cross-national associations are striking and have been replicated in subsequent work.
Welfare States and Varieties of Capitalism
Esping-Andersen's Three Worlds
Gosta Esping-Andersen's 'The Three Worlds of Welfare Capitalism' (1990) fundamentally changed comparative welfare state research by arguing that welfare states differ not merely in spending levels but in their underlying logic of social organization. His concept of decommodification -- the degree to which workers can maintain a living standard independent of the labor market -- captured a qualitative dimension invisible in expenditure comparisons alone. His three-regime typology distinguished liberal welfare states (the United States, United Kingdom, Canada), which provide modest, means-tested benefits designed to preserve market incentives; social-democratic welfare states (Sweden, Denmark, Norway), which provide generous universal benefits that substantially decommodify labor and promote cross-class solidarity; and conservative-corporatist welfare states (Germany, France, Austria), which provide generous but status-preserving benefits organized around occupational categories, maintaining gender and class hierarchies while insuring workers against market risk.
Feminist critics of Esping-Andersen, including Ann Shola Orloff and Julia O'Connor, noted that the framework neglected unpaid care work and treated women's relationship to welfare states as secondary to men's labor market participation. Subsequent research added familialist dimensions to the typology and identified a fourth Southern European cluster (characterized by family as the primary welfare provider) and potential East Asian developmental welfare state variants.
Hall and Soskice: Varieties of Capitalism
Peter Hall and David Soskice's 'Varieties of Capitalism' (2001) provided an institutional analysis of advanced economies that explained why different capitalist countries specialize in different kinds of economic activity. Their framework contrasts liberal market economies (LMEs) -- the United States, United Kingdom, Australia -- which coordinate economic activity primarily through competitive markets and prioritize labor flexibility, with coordinated market economies (CMEs) -- Germany, Japan, Sweden -- which coordinate through non-market institutions including employer associations, long-term bank-firm relationships, and cooperative labor relations. LMEs are suited to radical innovation and the rapid reallocation of resources to new activities, giving them advantages in sectors like high technology and finance. CMEs are suited to incremental innovation and high-quality manufacturing that requires stable long-term relationships and patient capital, giving them advantages in precision engineering, specialty chemicals, and machine tools. The framework has been influential and contested: critics have questioned whether the varieties are as distinct as the typology suggests, whether they are stable under globalization pressures, and whether there are not more than two coherent types.
State Capacity and Political Development
Migdal, Evans, and the State in Developing Countries
Joel Migdal's 'Strong Societies and Weak States' (1988) argued that in many post-colonial states, formal state institutions compete with a wide array of social organizations -- landlords, ethnic associations, village authorities, local strongmen -- that resist state penetration and provide alternative sources of order and identity. The result is organizational fragmentation: official state institutions are formally present throughout the territory but are hollow in practice, with officials whose effective loyalties lie with local patrons rather than with the national apparatus. The imagery of a strong state imposing order on society reverses the actual relationship in many cases.
Peter Evans's 'Embedded Autonomy' (1995) examined the contrasting cases of East Asian developmental states. The most effective developmental states, Evans argued, combine two apparently contradictory features: meritocratic bureaucratic autonomy (protection from capture by particular social groups, enabling coherent long-run policy) and embeddedness (dense network ties to productive sectors of the economy, providing the information and cooperation necessary to implement policy effectively). South Korea's Economic Planning Board and Japan's MITI during their high-growth periods exemplified this combination. Brazil's industrial policy was more inconsistent, reflecting less coherent embeddedness. India's bureaucracy showed strong autonomy without adequate embeddedness.
Fukuyama on Political Order and Decay
Francis Fukuyama's two-volume study, 'The Origins of Political Order' (2011) and 'Political Order and Political Decay' (2014), offered the most ambitious synthetic account of political development in comparative politics in recent decades. Fukuyama argued that successful modern states require three components: capable state institutions with monopoly on legitimate violence, rule of law (constraint of the state by legally enforceable norms), and accountability to citizens. The historical challenge is that these three components emerged in different sequences in different societies, producing distinctive institutional configurations. England achieved rule of law before it achieved a strong state or democracy, while China developed a strong state very early but without rule of law or democracy. The United States combined rule of law and early democracy but only developed a strong bureaucratic state relatively late. Fukuyama's concept of political decay describes the tendency of well-functioning political systems to become re-patrimonializated as vested interests capture institutions and use them to serve narrow constituencies rather than the public good.
Party Systems and Political Parties
Lipset-Rokkan and Cleavage Theory
Seymour Martin Lipset and Stein Rokkan's classic 1967 essay 'Cleavage Structures, Party Systems, and Voter Alignments' argued that the party systems of Western democracies were shaped by four fundamental social cleavages that emerged from two historical revolutions. The National Revolution produced center-periphery cleavages (between central national culture and ethnic, linguistic, or regional peripheries) and state-church cleavages (between secular state authority and religious institutional claims). The Industrial Revolution produced urban-rural cleavages (between primary producers and industrial interests) and class cleavages (between owners and workers). Different countries made these cleavages politically salient in different ways and at different times, producing the distinctive party system structures -- multiparty systems, two-party systems, religious parties, agrarian parties -- that characterized mid-twentieth-century democracies. The 'freezing hypothesis' held that party systems reflected the cleavage structures of the 1920s and remained relatively stable thereafter.
Giovanni Sartori's party system typology, developed in 'Parties and Party Systems' (1976), classified systems by the number of parties and the ideological distance between them, distinguishing two-party systems, moderate pluralism (three to five parties without extreme flanks), polarized pluralism (five or more parties with extreme parties and centrifugal competition), and predominant party systems (one party consistently winning majorities without authoritarianism). Richard Katz and Peter Mair (1995) proposed that mainstream parties in advanced democracies had evolved into cartel parties: collusive arrangements that use state resources and access to insulate themselves from electoral competition, raising barriers to entry for new challengers. This model helps explain the simultaneous decline in party membership, increase in party state funding, and rise of populist challengers.
Revolutions and Contentious Politics
Skocpol's Structural Theory
Theda Skocpol's 'States and Social Revolutions' (1979) is among the most cited books in comparative politics. Analyzing the French, Russian, and Chinese revolutions comparatively, Skocpol argued against both voluntarist accounts that attributed revolutions to ideological movements and Marxist accounts that centered class struggle. Her structural analysis identified a specific combination of conditions necessary for successful social revolution: military-fiscal crisis of the old regime (often triggered by interstate competition that overwhelmed extraction capacity), political conflict within the agrarian elite that prevented unified suppression of unrest, and agrarian peasant revolt that was possible because existing village community structures enabled collective action. Crucially, Skocpol argued that revolutions were not made by revolutionary movements; rather, state breakdown created the structural opportunities within which existing social conflicts produced revolutionary outcomes. The comparison with Prussia, Japan, and England -- cases with similar pressures but no revolution -- tested and confirmed the structural model.
Tilly, Tarrow, and Contentious Politics
Charles Tilly's research program, from 'The Contentious French' (1986) through 'Contentious Performances' (2008), reconceived revolutions as a subset of a broader domain of contentious politics encompassing strikes, social movements, riots, demonstrations, and collective violence. Tilly introduced the concept of repertoires of contention: the historically specific, culturally learned menu of forms of collective action available to groups in a given era. The modern repertoire -- marches, demonstrations, petitions, strikes -- is historically recent, emerging in eighteenth and nineteenth-century Europe and diffusing globally. Earlier repertoires involved different forms: grain seizures, shaming rituals, community celebrations turned into political events. Repertoires change slowly because they are learned through practice and constrained by what authorities will and will not respond to.
Sidney Tarrow's 'Power in Movement' (1994) developed the concept of political opportunity structure as the key variable explaining when social movements emerge, how much they grow, and whether they succeed. Political opportunities include the openness or closure of institutionalized political systems, the stability or instability of elite alignments, the presence or absence of elite allies, and the state's capacity and propensity for repression. The Arab Spring of 2010-2012, beginning with the Tunisian revolution, provided a natural comparative experiment: uprisings spread across North Africa and the Middle East through a contagion process, but produced dramatically different outcomes -- democratic transition in Tunisia, military-backed restoration in Egypt, civil war in Libya and Syria, suppression in Bahrain and Jordan. Comparative analysis has pointed to military cohesion, elite defection, economic structure, and the degree of cross-class coalition as key variables.
References
Dahl, Robert A. 'Polyarchy: Participation and Opposition.' Yale University Press, 1971.
Lijphart, Arend. 'Comparative Politics and the Comparative Method.' American Political Science Review 65(3), 682-693, 1971.
Lijphart, Arend. 'Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries.' Yale University Press, 1999.
Esping-Andersen, Gosta. 'The Three Worlds of Welfare Capitalism.' Polity Press, 1990.
Hall, Peter A., and David Soskice, eds. 'Varieties of Capitalism: The Institutional Foundations of Comparative Advantage.' Oxford University Press, 2001.
Levitsky, Steven, and Lucan A. Way. 'Competitive Authoritarianism: Hybrid Regimes After the Cold War.' Cambridge University Press, 2010.
Skocpol, Theda. 'States and Social Revolutions: A Comparative Analysis of France, Russia, and China.' Cambridge University Press, 1979.
Tilly, Charles. 'The Contentious French.' Harvard University Press, 1986.
Tarrow, Sidney. 'Power in Movement: Social Movements, Collective Action and Politics.' Cambridge University Press, 1994.
Evans, Peter. 'Embedded Autonomy: States and Industrial Transformation.' Princeton University Press, 1995.
Fukuyama, Francis. 'The Origins of Political Order: From Prehuman Times to the French Revolution.' Farrar, Straus and Giroux, 2011.
Lührmann, Anna, and Staffan I. Lindberg. 'A Third Wave of Autocratization is Here: What is New About It?' Democratization 26(7), 1095-1113, 2019.