By Frank Bandau (University of Bamberg)
Does government partisanship matter when it comes to the size and generosity of the welfare state? While the answer to this question is clearly positive for the ‘golden age’, there is so far no clear-cut answer with regard to the subsequent era of welfare retrenchment. This is by no means due to a lack of research. Quite to the contrary, a substantial number of macro-quantitative studies published over the past 15 years have addressed this very question.
What is obviously missing so far is a comprehensive and systematic summary of this work which:
a) presents the main insights of existing research and,
b) identifies prospects for future research.
The former goal is achieved by grouping 54 empirical studies and the incorporated statistical tests along the lines of the dependent variable as well as individual welfare programs. The main insights are that partisan effects on social spending have declined and even vanished over the course of time while the impact of partisanship persists when we turn to the more sensitive measure of welfare entitlements. Notably, the latter finding does not apply to life-course related welfare programs like pensions but is limited to class-related welfare programs like unemployment benefits.
Concerning the second goal of formulating research prospects, welfare scholars have recently raised a number of methodological issues which serve as excellent starting points for future research. First, a comparison of the most prominent entitlement datasets shows that researchers have to be (more) conscious about choosing data which match their conceptual framework. Second, future research may profit from using more dynamic concepts of partisanship. Using time-variant values of party positions can yield additional insights about the role of parties, e.g. help us to understand if programmatic convergence or external constraints are at the bottom of declining partisan effects. Third, there is fresh evidence that standard panel data analysis discriminates against partisan variables. Shifting the focus from country-years to political entities like cabinets might challenge and even alter the findings on the non-existent impact of parties on social spending.
Full Content: SSRN