The meaning of political trust is vague at best. While trust in politicians is usually considered as measuring specific support, trust in institutions implies a diffuse positive orientation towards the democratic political order. The concept’s theoretical vagueness has resulted in modifications of Easton’s original concept of political support, e.g. by Norris (1999). While principal component analyses of item batteries asking for trust in institutions usually uncover a strong first component, stricter requirements for unidimensionality, i.e. a higher percentage of variance explained by the first component or a higher alpha value, indicate its multidimensionality. Moreover, like the ambiguous indicator ''satisfaction with democracy'', the relationship between trust in institutions and other indicators of political support varies considerably across time and space. Furthermore, levels of trust in different types of institutions differ dramatically. Based on surveys of citizens, parliamentarians and journalists in five new (Chile, Poland, South Africa, South Korea, Turkey) and two old democracies (Germany, Sweden), the paper will explore the dimensionality of trust in institutions in different world regions and for groups with different levels of political sophistication. In a next step, the relationship of trust in institutions with perceptions of government performance, support for democracy, satisfaction with democracy as well as personal characteristics (educational level, life satisfaction, inter-personal trust, party affiliation) will be explored to determine if trust in institutions is conditioned by perceptions of political performance or if it is rather part of a syndrome of political satisfaction that primarily depends on personal characteristics.
