Working Papers
Expanding Participation, Strengthening Machines: Evidence from Paraguay’s Registration Reform
Participation-enhancing electoral reforms are often assumed to strengthen democracy, but in clientelistic settings marked by strong organizational asymmetries they can reinforce machine dominance. I argue that when participation barriers fall uniformly, dominant parties can convert lower mobilization costs into greater electoral returns by capturing a disproportionate share of newly incorporated voters. These effects should be strongest where mobilization is more costly and where organizational capacity yields greater returns. Paraguay’s 2012 automatic registration reform—which automatically enrolled citizens upon turning eighteen—provides a useful setting to evaluate this argument. Using a difference-in-differences design with locality-level electoral data, I show the reform increased turnout in treated localities, and the ANR captured a disproportionate share of the expanded electorate, with the largest gains in low-density areas where mobilization costs are highest. Complementary survey evidence shows that, after the reform, younger cohorts became more likely than older voters to report clientelistic contact. Together, these findings show how inclusionary reforms can entrench, rather than erode, dominant machine advantage.
The Historical Roots of Affective Polarization
Research on affective polarization has largely emphasized contemporary sources of partisan hostility, including social media, elite rhetoric, and partisan sorting. This focus, however, overlooks how historical experiences of political violence can shape the emotional foundations of contemporary democratic conflict. I argue that authoritarian repression leaves durable affective legacies by creating moral boundaries of victimhood and responsibility that can become embedded in partisan identities and reactivated by political elites. I test this argument in Spain by combining provincial records of Francoist executions, contemporary survey data, and parliamentary speech data. I show that individuals living in provinces with deeper repression legacies display higher levels of affective polarization, and that this relationship is strengthened during electoral periods when elites increase the salience of the authoritarian past. These findings demonstrate that contemporary partisan hostility is shaped not only by present-day political dynamics but also by the persistence and strategic activation of historical moral conflicts.
Political Polarisation and the Activation of Direct Democracy
Mechanisms of direct democracy are widely available across representative democracies, yet their activation varies across countries and over time. We argue that political polarisation helps explain this variation by reducing the relative attractiveness of representative bargaining and increasing the appeal of submitting conflict to a popular vote. Using a country-year panel of stable democracies since 1945, we estimate random-intercept logistic and negative-binomial models with two measures of polarisation and explore plebiscitary, proactive, and reactive pathways. Both camp and party-system polarisation are positively associated with the occurrence and frequency of discretionary mechanisms. The strongest pathway evidence concerns authority-initiated plebiscites, where the association with polarisation increases with governing-coalition legislative strength. Reactive referendums show more tentative evidence consistent with a substitution logic under fragmented parliamentary opposition. We find no support for the hypothesised moderation of proactive activation by civic-action capacity. The paper links polarisation to institutional choice across distinct routes of direct democracy.
Going Personal: Personalist Parties and Leader-Centered Affective Polarization
Much of the literature on affective polarization examines how strongly citizens favor political in-groups and dislike out-groups, usually with parties as the main reference points. We argue that party personalism changes the relative affective importance of parties and leaders. As authority becomes concentrated in a dominant leader, political affect should become more strongly centered on that leader relative to the party. We combine CSES party and leader like–dislike evaluations with V-Party measures of party personalism across 155 elections in 55 countries. Higher party personalism is associated with a relative shift in affective differentiation toward leaders, and within-party comparisons show the same pattern. This relationship is concentrated primarily in warmer evaluations of favored leaders rather than stronger hostility toward rivals. The findings connect variation in party organization to the political objects around which affective polarization is structured.