We propose conceptual and empirical solutions to problems in Satisfaction with Democracy (SWD) research. They are: 1) an unproductive debate about the meaning of the SWD survey question; 2) inadequate acknowledgment that the SWD question is an individual-level attitude requiring individual-level theory grounded in political behaviour and psychology literatures; and 3) no study of SWD uses an appropriate statistical model where institutional and contextual factors only influence SWD through a complete set of individual-level determinants of satisfaction, and not alongside them.
We provide an encompassing theoretical model of institutional and individual determinants of SWD and estimate it using state-of-the-art multilevel mediation models (Preacher et al. 2010) on a merged dataset of ESS and country-level contextual data. Since there are many possible attitudinal mediators through which institutions affect SWD, the first step is theoretical: specifying these relationships from contextual factors through individual-level mediators, to the outcome of interest – SWD.
