Center for the Study of Democracy

The diversity of issue interests and party options in multiparty systems makes individual electoral decisions increasingly complex. Voters are challenged to find a political party that represents their own political views in this more complex political space. This research offers a new methodological approach to studying voting choice in a multidimensional party space. We integrate the issue preferences of European voters and the issue preferences of party elites in a two-dimensional model of electoral choice. A common space of political competition for citizens and party elites is defined by the economic and cultural cleavages using data from the 2009 European Elections Studies (EES). Our innovation is to employ multilevel structural equation modeling to address the unique statistical challenges of a multi-dimensional party space, mass-elite comparisons, and cross-national analysis. This new approach generates results that are distinctly different from previous studies—even those using the same dataset. By factoring in the measurement of issue dimensions, economic issues have a stronger impact than recognized using previous methodologies, with more modest cultural influences on voting. Moreover, there are significant cross-over effects of the two cleavages in voters’ choices. The results reveal the complexity of realignment between voters and political parties in Western Europe.

The 2018 elections in Brazil caused profound disturbance to political scientists, political analysts, jurists, and activists.1 The electoral outcome shocked the establishment. The winner was an anti-system politician, an irrelevant (and burlesque) representative, with no significant party, no experience in positions in the Executive Branch, and with a frankly pro-military authoritarianism against minorities discourse. Bolsonaro reached strong popular support and was elected President. Politicians, scholars, and part of the voters immediately ran out for an explanation for such an electoral outcome.

In today’s world of intense ideological conflict at the elite level, the nature of mass belief systems has changed dramatically since the last time Converse’s famous levels of conceptualization (Campbell et al., 1960; Converse, 1964) were coded in 2000.  This paper shows that the percentage with well-developed belief systems based on a clear understanding of public policy choices has increased substantially since then.  It also introduces a new category termed “policy wonks” to reflect a sub-category that Converse only referred to in passing but which is now quite common.  Unlike respondents whom I classify as “concept ideologues” in this paper, policy wonks do not employ overarching concepts such as liberalism/conservatism or the scope of government.  Rather, policy wonks just refer to at least three public policy stands when asked what they like and dislike about the major parties and presidential candidates.  Although it was very rare for citizens in the 1950s to show a clear belief system based on the specific choices of government action, today’s highly intense and polarized policy debates have made programmatic-oriented belief systems quite common.  A close examination of policy wonks shows that they are just as politically knowledgeable and consistent on issue dimensions as concept ideologues (i.e., those who employ ideological terms).  Hence, policy wonks possess a well-defined belief system based on employing an understanding of public policy, thereby befitting Converse’s criteria for classification at the top level of conceptualization.   

The substantial increases in both concept ideologues and policy wonks accounts for virtually all of the increase since the 1980s in respondents whose partisanship matches their ideology (i.e., conservative Republicans and liberal Democrats).  Not only are respondents at the top of levels of conceptualization more numerous than they used to be, but being more consistent than they used to be has led to a marked increase in the overall correspondence between partisanship and ideology.  On the other hand, the decrease in ideologically inconsistent partisans (i.e., liberal Republicans and conservative Democrats) has occurred across all conceptualization levels.  Thus, party polarization is a combination of:  1) better-developed belief systems increasing ideological-partisan consistency; and 2) partisan sorting decreasing partisans who are out step with their party’s ideological stance.

Past research has shown that Republicans are substantially more likely to be ideologues whereas Democrats are much more inclined to conceptualize politics in terms of group benefits.  This pattern was quite evident in the 2008 and 2012 American National Election Study (ANES) responses that I personally coded.  However, two developments occurred in 2016 that dramatically reshaped the partisan nature of belief systems.  First, the Bernie Sanders wing of the Democratic Party evidenced a great deal of ideological thinking, thereby pushing Democrats to a record percentage at the top level of ideological conceptualization.  Second, the voters who supported Trump in the Republican primaries were much less likely to be ideologues or policy wonks than those who supported more traditional Republican candidates.  These developments combined to make Democrats and Republicans more similar than ever before in terms of ideological conceptualization in 2016. 

Partisan dealignment is recurrently presented in the literature as a main driver of the “personalization of politics”. Yet, on the one hand, the claim that leader effects on voting behaviour are increasing across time is short on comparative evidence. On the other hand, there is limited empirical evidence that such increase is due to dealignment. This article addresses these claims, exploring the longitudinal relationship between dealignment and the determinants of vote choice through a novel dataset pooling 90 national election surveys from 14 Western European parliamentary democracies in the period 1961-2016. The results suggest that both critics and proponents of the personalization thesis got it partially right. Leader effects did not increase over time, but their relative importance did: leader images came to matter more as party attachments came to matter less. Partisan dealignment is the key contextual dynamic in downplaying the electoral impact of partisan attachments vis-à-vis leaders evaluations.

For years, many political scientists and research institutes endorsed South Korea (Korea hereafter) as a fully consolidated liberal democracy. This non-Western icon of liberal democracy recently underwent a series of setbacks due to the restoration of autocratic governance by the President Park Geun-hye government. Why did liberal democracy backslide in the highly globalized and modernized country, contrary to what is expected from modernization and other prominent theories of democratization? To explore this question, we propose a cultural theory of democratic deconsolidation, and test it with the latest wave of the Asian Barometer Survey conducted in Korea in 2015. The analysis indicates that socioeconomic development under the sponsorship of the state and big businesses has failed to “emancipate” both the ruling class and the masses from the Confucian legacies of political paternalism and social harmony. Moreover, it has failed to instill them with “the bourgeois impulse” to become a free and equal being. As the habits of their hearts and minds, these legacies powerfully motivate both groups to reembrace or condone the resurgence of autocratic political practices. Theoretically, therefore, the deconsolidation of liberal democracy in Korea and the prevalence of affinity for paternalistic autocracy among its people can be considered two solid pieces of evidence confirming the thesis of “No bourgeois, no democracy”. They can further be considered to support the orthodox Asian Values Thesis that Confucianism is inherently incompatible with liberal democracy.

What does democracy mean? This question is difficult to answer - theoretically, we find different ideas of a “good” democracy, and also empirically, democracy is a multidimensional concept: Across countries, democracy varies both in its quality and in the way it is realized. Yet, when researchers for example analyze if citizens are satisfied with “the way democracy works” in their country, they suppose that democracy means the same for individuals all over the world. I argue that in order to be able to analyze support for democracy in a more nuanced way, we need to take a step back and ask what democracy actually means to citizens and how such expectations are formed. Based on the theoretical and empirical literature on varieties of democracy, I suppose that individual expectations from democracy differ across countries, and that they are influenced by two factors: The democratic culture, consisting of age and quality of institutions as well as authoritarian legacies, and the prevalent democratic model. Hence, the specific democratic context in which a citizen lives matters - due to socialization and democratic learning, individuals acquire democratic preferences and value those dimensions more which they experience in their own democracy. Using individual-level data from the European Social Survey (ESS) Round 6 as well as country-level data from the Democracy Barometer, I test how the national democratic context in 27 European democracies influences these individual democratic ideals. Indeed, I find evidence for both socialization and participation effects of the democratic context on citizens’ democratic ideals.

Most federal second chambers give subunits equal representation. A few apply per capita representation, like most first chambers. Only Germany and Canada compromise between territorial and per capita representation. Both broadly allocate seats following S_i=SP_i ⁿ/∑P_k ⁿ, the only simple format without internal inconsistencies. Two values have been proposed for n. The rigid n=0.5 approximates the Canadian pattern but does not fit the German system. The flexible n=[1/logT-1/logS]/[1/logT-1/logP] takes into account the number of subunits (T) and total seats (S), for given total population (P). The flexible model better predicts seat allocation both in Canada and Germany. This model has been shown to apply to the European Parliament and the EU Council. Hence it may express what countries intuitively grope for when trying to strike a compromise between representations per capita and per subunit. As such, it does not fit the seat allocation of administrative subunits in unitary states, France and Italy.

What types of political systems do people in East Asia favor most and least? Throughout the region, do most people uniformly prefer democracy to nondemocratic systems, as advocates of universal democratization theses claim? If they do, do they prefer liberal democracy to non-liberal democracy? If they do not favor democracy more than other types of political systems, what type do they favor most? Is it meritocracy or a hybrid system, for which proponents of Confucian democracy or the Asian Values Thesis have recently advocated? To address these questions, I first review previous studies on democratic system support and highlight their limitations in unraveling the meanings of avowed democratic system support and comparing its levels across different countries and regions of the world. Then I propose a new typology of citizen preferences for a variety of political systems, including democracy and autocracy. Unlike all other typologies, it ascertain in sequence the types and subtypes people prefer without using the word “democracy” (“the D-word” hereafter). Finally, I attempt to evaluate the relevance of universal and liberal democratization theses in the context of East Asia, analyzing the 3^rd wave of the Asian Barometer Survey conducted in 13 democratic and nondemocratic countries. The analysis reveals that these theses serve merely as prodemocracy rhetoric, not as theoretically meaningful propositions.

The 2008 European recession has been linked to higher political unconventionality across countries in recent studies. Research on the impact of the economic downturn on people’s engagement in protest has focused on data mostly from 2008 to 2012. Recent findings have supported primarily a relative deprivation theory based explanation of why Europeans choose to participate in street marches, suggesting a change has taken place in the way the economy affects political contention. This article assesses the relationship between the economy and protest in 2014, six years after the crisis took place, a long enough period for countries to have improved their economic situation and for people’s interpretations of the economy to be more optimistic. Does the economy still matter to explain protest if it is not as salient any longer? This research employs data for 13 European Union member states from the 2008 and 2014 European Social Survey to test the importance of national level objective economic indicators as well as individual level evaluations of financial wellbeing to study the link between confrontational activism and economic variables. Some of the findings suggest a limited relevance of the economy in the explanation of protest, for objective economic variables, yet a more salient role for personal interpretations of economic wellbeing. Yet, a combination of relative deprivation and resources theories is needed to understand why citizens choose protest in light of the economic situation. The link between the economy and confrontational activism in 2008 and 2014 looks in the end very similar, dismissing any serious long term change in the relationship.

A growing number of political scientists have recently claimed that democracy has emerged as a universal value, and that it is also becoming the universally preferred system of government. Is the whole world becoming democratic, as these proponents of global democratization claim? To test the validity of these claims, this study critically reviews the voluminous literature on citizen conceptions of democracy and identifies the limitations of previous public opinion research on democratization. In an attempt to overcome those limitations, it proposes a two-dimensional notion of informed democratic understanding, and thereby reanalyzes the World Values Surveys conducted in 2005-8. Results of the analysis reveal that two-thirds of global citizenries are either uninformed or misinformed about the fundamental characteristics of democracy and its alternatives. In every region except for the old-democratic West, moreover, the well-informed constitute minorities of its avowed supporters. On the basis of these findings, the study contends that for much of the world today, democracy represents little more than an appealing rhetorical political symbol voiced in regimes that still retain authoritarian practices. It also contends that contrary to increasingly popular theses of global democratization and neo-modernization, liberal democracy is not likely to stand at the “end of History.”