The Dynamics of Income Expectations and Redistribution Preferences

Date
-
Event Sponsor
Co-sponsored by the Europe Center
Location
Encina Hall West, Room 400 (Graham Stuart Lounge)
Speaker

Daniel Stegmueller, Assistant Professor of Political Science, Duke University

 

Abstract

Income is a central concept in political economy models of redistribution preferences. How-ever, empirical results regarding the effect of income on preferences are surprisingly incon-clusive. In this paper, I argue for the centrality of individuals’ long-term income expectations. Individuals predict their future income based on past income dynamics and their current economic situation. Income shocks, such as becoming unemployed or getting divorced, are included in this expectation formation process. Thus, individuals form redistribution preferences based on the present-day value of their expected future income and not only based on short-term income fluctuations as captured by cross-sectional surveys. To test my argument I use individual-level panel data from the UK spanning 1991 to 2007. I develop a Bayesian simultaneous dynamic panel model, which jointly models the process of individuals’ income dynamics, their future expected income, and their preferred level of redistribution. I find clear evidence for the central role of income expectations, and only limited effects of current income and short-term income shocks.

 

Biography

Daniel Stegmueller is an Assistant Professor in the Department of Political Science at Duke University. He's also an associate member of Nuffield College, University of Oxford, and of CAGE, University of Warwick. 

His research lies at the intersection of political economy and political behavior. He studies political preferences and choices in advanced industrialized societies, specifically individuals' preferences for redistribution and redistributive voting. He's interested in how these are shaped by social structure and institutions, but also by basic individual characteristics, such as cognitive and non-cognitive skills.

Methodologically, he's interested in applied Bayesian modeling, Bayesian Nonparametrics, hierarchical/multilevel models, measurement and discrete choice models. He's particularly interested in applying these models to problems of comparative/cross-cultural research and to panel data.