"Rational choice? A two-level analysis of the Nordic EU referendums"

Moses, Jonathon. (1997) "Rational choice? A two-level analysis of the Nordic EU referendums". In: UNSPECIFIED, Seattle, WA. (Unpublished)

Preview

PDF Download (1590Kb) | Preview

Abstract

The impetus for this paper is an empirical puzzle. In a previous work, my co-author and I employed a single, fairly parsimonious, model to explain political behavior at two distinct levels of analysis: the individual and the country (i.e., a higher level of aggregation). The puzzle began here: the model did not fit comfortably at both levels of analysis. To overcome this lack of fit, I began a series of alterations. As will become more clear, my first response was to try again, with no luck. I then began to relax assumptions; this only introduced more problems. Finally, we are now trying to develop an even grander, more complex model. As work in the latter direction continues, I have begun a more radial (and speculative) query to the problem at hand. This paper airs that query with a hope at sparking a dialogue, and eventually (hopefully) reconciling the initial dilemma. The query (argument is too strong a word) contains two parts. In the first section I introduce some of the empirical results which initiated the problem. Here I present some data from Moses and Jenssen (1997) which show how a narrowly defined rational choice argument goes some distance in explaining county-level referendum outcomes, but finds very little support in the individual-level data. I refer to this explanatory gap as the inconsistency problem. Our empirical work studies variance in Nordic (Finnish, Swedish and Norwegian) referendum outcomes on EU membership in 1994. Our data are from cross-national polls and county-level aggregate data.

Actions (login required)