Research
A Dynamic Discrete Choice Approach to the Stability and Consistency of Attitudes
with Steven Stern
This paper proposes a new model for the study of the stability and consistency of attitudes, with a view towards describing the structure of belief systems. We formulate a dynamic discrete choice model of the joint determination of multiple attitudes, with persistent shocks and cross-attitude spillovers. It assigns attitudinal variation to sources that are inherently stable (individual-level permanent heterogeneity and slow-moving observed covariates) and those possibly indicative of instability (time effects, persistent shocks, and cross-attitude spillovers of shocks) and it can detect belief systems, as informed by the measured consistency across attitudes. We apply our analytic approach to study five core political attitudes concerning how people view their role in politics and their evaluations of the wider political world, using the British Election Study Internet Panel. Attitudes are stable, with shocks petering out within months. Most of the variance in the data reflects slow-moving individual observable characteristics and time fixed effects which capture broader political developments. Permanent, individual heterogeneity plays a modest role. Trust in Members of Parliament emerges as the central attitude concerning how people evaluate the external political environment, while neither civic duty nor internal efficacy plays the leading part in how people view their role in politics. Our model is a natural fit for other settings where multiple measures involving ordinary scales evolve and interact over time, more so wherever considerations of stability and spillovers are key.
The Dimensionality of Political Interest and Its Implications for Political Engagement
General political interest is at the core of many of the central questions concerning people’s political habits, knowledge, and voting patterns, as well as candidates’ mobilization efforts. In this paper, I advance the proposal that political interest is inherently multidimensional, that people can be interested in politics in multiple ways, and that individuals will differ in the mix of types of interest they exhibit. I distinguish between three kinds of dispositional interest: utility, attainment, and intrinsic. People are disposed to being interested in politics because it could further their short- or long-term goals (utility), because it might be part of their (ought-)self-image or because it might deliver social rewards (attainment), or because politics itself is enjoyable or engaging, perhaps not unlike a hobby or supporting one’s favorite sports team (intrinsic). I develop and validate measures for the three dimensions of dispositional interest, field them in four survey samples, and find that my measures uncover novel patterns across socio-demographic groups, including respondents’ propensity to engage in various forms of participation. Two factorial vignette survey experiments allowed me to assess the impact of respondents’ configuration of political interest on their perception of the effectiveness of addressing a political issue by either posting on social media or by contacting one’s representative in a way that is independent of the experimentally manipulated characteristics of the action and the behavioral context. I find that the way people are interested in politics influences their assessment of the effectiveness of these two forms of participation and influences which characteristics of the action and behavioral context are considered when evaluating a specific action. Taken together, these results point to substantial heterogeneity in how people take an interest in politics, with salient implications for participation, perceptions of the effectiveness of political actions, and the role of demographic and socio-economic factors in driving inequalities in participation and representation.
Ambiguous Policy Rhetoric & Credit Claiming: Stacking the Deck Against Women Candidates?
with Kerri Milita and Elizabeth Simas
[ABSTRACT]
Materials
My research statement (PDF)