In rich democracies, about one-fifth of government expenditure goes to the mainstate prerogatives – general public services (including debt service),defense, public order, and safety. About one-half goes toward social policynarrowly defined – social protection and health. And more thanthree-fifths (almost two-thirds) is allocated to social policy more broadlydefined – social protection, health, and education (OECD 2011, tables 5.1and 5.3). In other words, government expenditures on welfare functions dwarfthose on night-watchman functions. In all rich democracies, governments forcetheir citizens to be part of mandatory social insurance systems covering allmain risks. Taxing their citizens and spending huge amounts of money on socialpolicy are the main activities of governments today.

It is inconceivable that, in democracies, the single most prominent governmentactivity – social policy – can be performed without widespreadsupport by citizens. Yet, the potential importance of public opinion for socialpolicy making has largely been neglected. Clem Brooks and Jeff Manza, who haveperhaps most systematically addressed this shortcoming, argue that existingaccounts are “reaching for, but not grasping, public opinion”:“Analysts within multiple theoretical traditions have anticipated thepossibility that mass opinion is a factor relevant to shaping socialpolicymaking, and perhaps in accounting for differences between countries. Butnone of these scholars has taken the next steps of explicitly theorizing andattempting to measure its policy impact” (Brooks and Manza 2007,25–26). Analyzing the potential policy impact of public opinion –as in the seminal work by Clem Brooks and Jeff Manza – is a step in theright direction, but it begs the important question: why does popular supportfor social policy differ across space and time? We have to turn to theindividual level to explore this topic.

Despite its shared focus on democracies, the welfare state literature has, forthe longest time, paid very little attention to citizen preferences. Onlyrecently have scholars investigated the topic, but existing studies typicallyexclusively concentrate on the micro-level. A joint perspective of micro-leveldeterminants of social policy preferences and the latter's impact onsocial policy making has yet to be conducted. It is such a perspective that Idevelop in this chapter.