Exploring the measurement of political trust A multilevel observational analysis of six Swedish public agencies

Abstract

Political trust is a well-used construct and serves both as an explanation and an outcome in the social sciences. Considering the importance of the construct, relatively little attention has been allocated to its measurement. While the exist-ing literature on the measurement of political trust focuses on either developing and validating new scales, or scaling and equivalence assessment of the more gen-eral measures, this article contributes by analysing the contents of the widely used survey items of political trust. Put differently, what is in a typical political trust measure? The analysis uses relevant observational data from Sweden (n = 1760) with repeated questions over six public agencies, where the typically used trust measurement is regressed on theoretically motivated psychological antecedents of trust using a hierarchical heteroskedastic ordered probit model. Results imply that the typical trust measurement contains traces of perceived competence and less so perceived motivation. The results also suggest that political actors do carry meaning beyond the trust construct, influencing both the location and scale of the response distribution.

Keywords

• Political trust
• Measurement
• Bayesian mixed effect scale-location
• Sweden
• Survey data