What Helps Citizens to Have Consistent Attitudes? A Cross Country Analysis of the Individual and Contextual Determinants of Attitude Constraint

The inconsistency of the beliefs citizens have about political issues is neither a new nor a surprising finding for political scientists (Converse 1964). Still at least from a normative point of view attitude constraint is a desideratum that is important for the quality of electoral decisions and ultimately for the quality of democracy (Friedman 2006). Under these circumstances it is not surprising that several studies examined the constraint between the attitudes of the citizens (Converse 1964; Converse and Pierce 1986; Peffley and Hurwitz 1985; Zaller 1990; 1992 Sturgis, Roberts and Allum 2005; Granberg and Holmberg 2006). One factor that is especially important is the level of political knowledge, more informed voters should have higher levels of attitude constraint as they are better able to identify their preferences (Althaus 1998; Carmines and Stimson 1980; Delli Carpini and Keeter 1996; Lau and Redlawsk 1997; 2006; Downs 1957; Dahl 1989; Converse 1964). But at the same time as most voters are politically ignorant, a simple heuristic such as having a party ID might help them have consistent political attitudes, compensating for lower levels of political knowledge (Lupia 1994; Popkin 1994; Lau and Redlawsk 2001; Zaller 2004). This paper will further investigate the determinants of attitude/issue constraint with an emphasis on political knowledge and party ID, as a simple heuristic. Its major contribution is that it will go further than the single country environment in which this was studied before. Using the 2009 European Election study, will also allow me to bring context into the picture by analyzing how the level of attitude constraint varies across the institution rich environment provided by the 27 member states of the EU. The empirical analysis shows that while political knowledge and the constraint of elites have a consistent effect of attitude constraint, the effect of polarization and partisanship is at best limited.