Citizens’ demand for social insurance is as emphatic as is its supply. Surveys reveal that citizens overwhelmingly agree that it should be the government's responsibility to provide health care for the sick (98 percent); to provide a decent standard of living for the old (96.1 percent); and to provide a decent standard of living for the unemployed (67.8 percent). And, indeed, every rich democracy has social policy programs addressing these (and other) risks. But support for social insurance varies across countries, as does its supply. In fact, there is a fairly close correlation between how supportive citizens are of social policy and how generous social policy programs are. This correlation is well documented (see also Figure 2.2 and Figure 2.3), especially in the work by Clem Brooks and Jeff Manza (Arts and Gelissen 2001; Brooks and Manza 2006a, 2006b, 2007; Burstein 1998, 2003; Coughlin 1980; Flavin 2012; Gelissen 2002; Gilens 2005, 2009). But the relationship between public opinion and social policy making is not well understood: “mass opinion is … the great black box of contemporary welfare state politics” (Brooks and Manza 2007, 16).

The long neglect of citizen preferences as a core ingredient of social policy making is surprising, for a variety of reasons. First, the welfare state literature typically restricts its sample to (rich) democracies. Therefore, the idea that public policy reflects citizen preferences should not be controversial, at least with respect to important, visible, and durable policies such as those underpinning welfare states. In fact, congruence between citizen preferences and policy output is a core part of many definitions of democracy. Second, and as previously mentioned, Clem Brooks and Jeff Manza have convincingly argued that existing accounts are “reaching for, but not grasping, public opinion” (Brooks and Manza 2007, 25). Although citizen preferences are greatly understudied in the welfare state literature (or were until very recently), all main welfare state theories are, in principle, compatible with the conjecture that social policy making reflects citizen preferences; some theories even heavily rely on a link between mass preferences and outcomes (the Power Resources Theory, for example, assumes that parties aggregate and implement distinct citizen preferences).