Work in Progress
A Global Scale of Economic Left-Right Party Positions: Exploring the Cross-National and Cross-Expert Perceptions of Party Placements
with Ryan Bakker, Ruth Dassonneville, Seth Jolly, Jelle Koedam, Patrick Leslie, Jonathan Polk, Jill Sheppard, and Roi Zur
Abstract: We examine the cross-national comparability of expert placements of political parties on the economic left-right dimension using a novel data set that combines data from Europe, Latin America, Australia, Israel, Canada, and the United States. Using anchoring vignettes and Bayesian Aldrich-McKelvey Scaling (BAM), we evaluate whether there is evidence of geographic or expert-level differential item functioning (DIF), in terms of how experts perceive party placements on an economic left-right dimension. We then explore systematic differences across experts’ perceptual distortion parameters (DIF “shift” and “stretch” terms) using Bayesian Multilevel Regression Models. The models show that there are no substantively interesting systematic biases in perceptions of party placements for either term at the region, country, or expert level. More generally, our results clearly support the claim that the economic left-right dimension travels well across the globe.
Invited to revise and resubmit at the Journal of Politics.
Affect, Not Ideology The Psychological Mechanisms of Partisan Information Processing
with Sam Fuller and Jack T. Rametta
Abstract: In this paper we evaluate the extent to which partisan bias is driven by affect or ideology by theorizing and testing the underpinnings of each of these potential mechanisms. We theorize that partisan bias can be an outcome of two different information processing channels: motivated bias and affect transfer. While motivated bias arises from the tendency to accept information that is consistent with the one stored in memory, affect transfer charges information affectively depending on the characteristics of the messenger. Using a survey experiment designed to disentangle these two mechanisms and a method designed to detect treatment effect heterogeneity and calculate conditional treatment effects (Causal Forest), we find that affect plays a much more important role than ideology in moderating the effect of partisan cues on information processing. This finding contributes to the literature by adding to our understanding of the psychological underpinnings behind partisan bias.
Invited to revise and resubmit at Political Behavior.
Ideological and populist bases of partisan responses to the COVID-19 pandemic in Latin America
with Jonathan Hartlyn and Cecilia Martínez-Gallardo
Abstract: This research note explores variation in policy responses to the COVID-19 pandemic in Latin America, looking particularly at preferences over the trade-off between virus containment and an open economy and over the expansion of executive authority. First, we argue that support for containment was shaped by the ideology of presidents and parties. But populism also had an important—though heterogeneous—effect; while non-populist consistently supported virus containment, populists had more varied responses. In a region marked by weak party systems, we show that the association between pandemic responses, ideology, and populism is stronger for more programmatic parties. Second, we also argue that incumbents and populist favored further concentration of power to address the pandemic. We rely on the Chapel Hill Expert Survey-Latin America (CHES-LA) to test our hypotheses. Overall, we show that key features of party competition – ideology, populism and programmatic linkages –all matter for understanding Latin America’s party systems.
Invited to revise and resubmit at the Journal of Politics in Latin America.
Beyond Partisanship: Political Identity Profiles in Latin America
Abstract: Research on political identities has been dominated by a focus on partisanship. However, it is widely recognized that parties are only one of many political markers with which citizens can identify. I address this gap in the literature by providing a theoretical and empirical framework to study political identification in a comparative perspective. The political identity profiles (PIP) framework proposes that political identities are not bound to partisanship and that context plays a fundamental role in shaping the salience of different forms of political identification. To test this framework, I introduce a multi-item scale tailored to measure different types of political identities. I then provide evidence for the internal and convergent validity of the measures used by focusing on the Latin American case. Results provide strong evidence in favor of the validity of the framework. I conclude by focusing on some of the normative implications of these findings for the study of political behavior and democracy.
Under review.
To blur or not to blur: Disentangling populist ideology and electoral strategy in Europe and Latin America
with Jacob Gunderson
Abstract: Studies of populism in Western Europe associate it with blurred or vague economic positions. But how integral is this strategy for populism outside West European democracies? This paper leverages a comparison between Latin America and Europe to argue that populists tailor their strategies to their competitive context, focusing on party system dimensionality and host ideologies as moderators. We validate new measures of positional blurring from the Chapel Hill Expert Survey to evaluate regional differences in positional blurring by populists and to directly test our moderators across these two regions. We find that populists blur their economic positions in multidimensional systems and when they adopt exclusionary thick ideologies, two characteristics of European populism. In Latin America, where politics is unidimensional and populism is generally inclusive, populists present clear, not blurred, positions. These findings have implications for the study of populism as a global phenomenon and parties’ adaptation to contextual factors.
Under review.
The Case for Non-Partisan Political Cueing: Experimental Evidence from Peru
Abstract: Citizens infer policy information from partisan cues. Yet, the utility of partisan cues varies cross-nationally. In weakly institutionalized democracies with short-lived political parties and volatile party systems, partisan cues are likely of little heuristic value. I argue that in these places citizens can rely on alternative political markers, such as ideological or political movement-related categories, to make sense of the political arena. I test this argument in Peru, a country with an extremely volatile party system and deep distrust in political parties. Using a survey experiment, I show that political cues significantly influence policy preferences even in the absence of clear and strong partisan brands. Findings from this paper underscore the importance of alternative forms of political identification in weakly institutionalized party systems.
Presented at the Midwest Political Science Association 2024 Annual Conference, the Duke/UNC Political Behavior Group Meeting (2024), and at the Congreso Chileno de Ciencia Política 2023.
Political Identities and Belief Networks in Developing Democracies
Abstract: Research in developed democracies suggests that belief networks are structured around political identities. However, we know much less about how belief networks are organized in places with less democratic experience and volatile party systems. Political elites play a key role in structuring belief systems by organizing belief packages and signaling to voters which ideas “ought to be together”. Accordingly, belief systems in less organized party systems can potentially systematically differ from those in highly institutionalized democracies. Building on the social identity perspective (SIP), I argue that political identities play a fundamental role in shaping political belief networks, even in contexts where party competition is less institutionalized. Using data from Chile and Peru and Mixed Graphical Models (MGNs), I show how ideological and political movement markers, as well as partisan ones, play an important role in structuring belief systems. Findings underscore the relevance of alternative forms of political identification, beyond partisanship, in less institutionalized democracies.
Presented at the American Political Science Association 2023 Annual Conference, Los Angeles.
Programmatic and Non-Programmatic Party System Structuration in Latin America
with Jonathan Hartlyn and Cecilia Martínez-Gallardo
Abstract: This paper evaluates the extent to which political parties in contemporary Latin America compete by offering voters distinct policy programs, and the degree to which they rely on alternative, non-programmatic appeals. Using data from CHES-LA, a recent expert survey of party positions in the region, we employ confirmatory factor analysis to show that political parties compete for voters using a mix of programmatic and non-programmatic appeals. We demonstrate that populism and particularism, though related, are two different non-programmatic dimensions, and that they interact in systematic ways with the programmatic dimension. While programmatically polarized parties tend to rely on populist appeals, centrist and right-wing parties are more likely to favor particularism. More generally, we provide a comprehensive view of party competition in the region.
Presented at the American Political Science Association 2023 Annual Conference, Los Angeles.
Beyond an Instrumental Exchange: The Effect of Clientelism on Expressive Partisanship in Mexico
with Rodrigo Castro Cornejo
Abstract: While clientelistic exchanges are typically analyzed from an instrumental perspective, viewing them as rational interactions between actors who aim to maximize their utility, we argue that gift-giving can transcend utilitarian considerations and increase expressive attitudes towards political parties. In particular, we posit that clientelistic exchanges can increase positive feelings towards political parties and even shape levels of partisanship. Identifying the effect of clientelism on citizens’ expressive considerations is particularly challenging because political parties often target loyal voters. We overcome this challenge by leveraging a three-wave panel study that allows us to identify the effect of clientelism on parties/candidates’ evaluations and partisanship controlling by political machines’ pre-existing patterns of resource allocation. Findings show that clientelism can improve evaluations of parties/candidates and partisanship towards the party that distributed the goods and decrease them among those parties that did not target the individuals. Overall, findings from this paper underscore how campaign gift-giving can lead to meaningful attitudinal and affective changes.
Presented at the Midwest Political Science Association 2024 Annual Conference.
Partisan Profiles in Democracies and Non-Democracies: Evidence from 101 Countries
with Courtney Blackington
Abstract: Though previous research consistently highlights the role of partisanship in democracies, we know less about partisans in non-democracies. Do levels of democracy shape the formation of partisan identities, and how do these partisan identities impact attitudes towards political institutions in different regimes? We theorize that levels of democracy shape both who identifies as regime or opposition-party partisans and whether partisans trust political institutions. Using cross-national data from 101 countries, we find that as levels of democracy decrease, partisans and their attitudes systematically differ. In less democratic countries, ruling-party partisans are less educated and greater in number than their opposition-party partisan counterparts; however, no similar gaps appear in democracies. Moreover, as levels of democracy fall, trust in political institutions polarizes. In democracies, partisans trust political institutions equally. In non-democracies, opposition-party partisans trust political institutions markedly less than ruling-party partisans. We show that democratic backsliding correlates with the emergence of these patterns in Hungary and Nicaragua.
Presented at the Southern Political Association 2021 Annual Meeting, Virtual.