Patterns of Democracy: Government Forms and Performance in Thirty-Six Countries

Lijphart argues that democracies can be mapped onto two opposing models — majoritarian (Westminster) and consensus — defined by ten variables that cluster into two dimensions: executives-parties and federal-unitary [citation:3][citation:6]. He rejects the conventional wisdom that majoritarian democracies like the United States and Great Britain are superior to consensual systems like Switzerland and Israel [citation:5][citation:9]. His empirical finding is that consensual systems stimulate economic growth, control inflation and unemployment, and limit budget deficits just as well as majoritarian democracies do [citation:1][citation:5].

The book introduced the two-dimensional conceptual map of democracy, operationalized through ten variables. The executives-parties dimension includes: concentration vs. sharing of executive power, executive-legislative balance, two-party vs. multiparty systems, majoritarian vs. proportional electoral systems, and pluralist vs. corporatist interest groups [citation:3][citation:6]. The federal-unitary dimension includes: unitary vs. federal government, unicameral vs. bicameral legislatures, flexible vs. rigid constitutions, parliamentary sovereignty vs. judicial review, and dependent vs. independent central banks [citation:3]. This framework is now standard in comparative politics for classifying and comparing democratic institutions [citation:16].

The second edition (2012) updates the data from the original period of 1945–1996 to 1945–2010, adding fifteen years of new evidence [citation:1][citation:3][citation:4]. Lijphart states that the primary change is updating the data; the argument itself remains unchanged [citation:3]. Because the conclusions held with the new data, the author became even more confident in his findings [citation:3]. The second edition also expands the analysis to include updated measures of democratic quality, including rates of incarceration as a new indicator [citation:8].

A significant critique, advanced by Steffen Ganghof, argues that Lijphart's measurement of executive inclusiveness is incoherent and invalid, and that executive inclusiveness is best explained by the interaction of many parties and strong legislative veto points [citation:17]. This implies executive inclusiveness should not be contained in either of Lijphart's two dimensions [citation:17]. Other critics note that the theory used to defend the internal coherence of the ten variables — whether logic, constitutional engineering, or causal relationships — is not fully convincing [citation:6]. The operationalization of several variables is also contested [citation:6].

Lijphart shows that consensus democracies outperform majoritarian systems on measures of political equality, women's representation, citizen participation in elections, and proximity between government policies and voter preferences [citation:5][citation:9]. His updated analysis finds strong correlations between institutional arrangements and quantifiable aspects of democratic quality, including political and economic equality, governmental accountability, rates of incarceration, and gender equality [citation:8]. The phrase 'kinder, gentler' captures the finding that consensus systems tend to produce more inclusive, less punitive, and more representative policy outcomes [citation:5].

A common misconception treats Lijphart's two models as prescriptive recommendations for all countries to adopt consensus democracy. In fact, Lijphart presents them as analytical ideal types that real countries approximate to varying degrees [citation:16]. The framework is descriptive and comparative, not a blueprint. He also does not claim majoritarian systems are undemocratic — merely that consensus systems perform better on specific quality-of-democracy measures [citation:5]. The models are 'pure' types that individual countries do not necessarily meet in all details [citation:16].