We know from existing studies there is a “mood” in the public’s policy preferences. That is, preferences across issue domains tend to move together – in parallel – over time. When preferences change in one domain, they tend to change in other domains. Stimson (1991) first demonstrated this parallel movement in the US. Others have since revealed similar empirical patterns in the UK (Soroka & Wlezien 2005; Bartle et al. 2011), Canada (Soroka & Wlezien 2004), and France (Stimson et al. 2012). The existence of a mood in various countries thus is (increasingly) clear. Its meaning is not, however, as the determinants of the parallel movement of policy preferences remains unexplained.
This study aims to theorize and begin to empirically assess what mood encompasses. We set out to do so by focusing on the possible factors that could explain how (relative) preferences (R) across policy domains j move together: (1) parallel movement in the preferred levels of policy (P*) across domains; (2) parallel movement in actual policy (P) across domains together with thermostatic feedback with each domain; and (3) non-parallel movement in (P) across domains with global thermostatic feedback to collective policy flows, across domains (cf. Wlezien 2002).
Each of these factors could account for the parallelism in relative preferences that we observe in different countries, and there is reason to think that each is at work. An understanding is important, as the patterns structure our interpretations of opinion change and policy responsiveness itself. We nevertheless do not know, as models of mood have not considered – either theoretically or empirically – the different factors. Our paper develops the theoretical model and then begins testing focusing on the countries listed above, for which we have reasonable time series in various domains.
