Estimating the Effects of Institutions on Behavior: Correcting for Selection Effects Using CSES Data

John Aldrich
11.10.2016, GESIS Mannheim (B2,8 rechts), 13:45

Abstract
CSES, like many cross-national survey projects, provides key insights into the effects of institutional and social structures on political behavior.  It, like the others, yields estimates from comparative designs such as hierarchical modeling, and similar techniques.  But it, like the others, does not have 100% participation from the various eligible nations.  Indeed, considering CSES in particular, there are some easy to observe biases in the set of nations that do and the set of nations that do not participate in CSES.  In this paper, we examine several existing papers to assess the consequences of non-random selection-to-participate decisions.  We employ several techniques for such corrections, apply them to these studies and reach appropriate conclusions.

Zur Person
Professor John Aldrich (PhD Rochester) is the Pfizer-Pratt University Professor of Political Science, Duke University, the Chair of the Comparative Study of Electoral Systems (CSES) Planning Committee for Module 5, and is visiting GESIS on the invitation of the CSES. Professor Aldrich is a past President of the Southern Political Science Association, the Midwest Political Science Association, and the American Political Science Association, has been a Guggenheim Fellow and is a Fellow of the American Academy of Arts and Sciences.
His works have appeared in leading political science journals including the American Journal of Political Science and the American Political Science Review and he is the (co)-author of several leading texts in the field of political behavior and methodology including Why Parties, Before the Convention, and Linear Probability, Logit, and Probit Models. His forthcoming work (2017, with John Griffin) is Why Parties Matter: Political Competition and Democracy in the American South 1832-2012.