Table of Contents
1. Comparative Research
2. The Three worlds of welfare capitalism
2.2 Social Stratification
3. The three welfare state regimes
3.1 The social democratic regime type
3.2 The conservative regime type
3.3 The liberal regime type
5.1 Three welfare-state regimes?
5.2 Drawing conclusions
5.3 Problems of empirical identification
Since the end of the 20th century, welfare states emerged in every industrialized society to protect its citizens against certain risks and disadvantages. The national differences of these social protection systems are admittedly considerable. This diversity has been ever since an issue for historical and social science and poses challenging questions: What are the differences of the different nation systems? Where do they come from – do they exist due to political, cultural or economic reasons?
In 1990 Gøsta Esping-Andersen published “The Three Worlds of Welfare Capitalism” which became a milestone concerning the field of comparative welfare state research. Since the The Three Worlds of Welfare Capitalism was published, it is extremely difficult to find even one article on welfare states in which Esping-Andersen’s approach is not at least mentioned. Reams of reviews, articles, critiques and additions were composed. An issue that is discussed ever since and which is still a topic of controversial debates is the number of regime types which should be distinguished. As the title already shows, Esping-Andersen suggests three distinctive welfare-types, but his empirical argumentation in particular caused a debate concerning the actual number of types.
In this term paper I want to take a closer look at an empirical critique that asks for additional regime-types. The discussed criticism was exemplary chosen from Obinger and Wagschal (1998). The first point of the term-paper is a short introduction to social comparative research in the context of cross-national studies. On the one side because it is important to place Esping-Andersen’s work in the field of research methods and on the other hand because The Three Worlds of Welfare Capitalism set a milestone in this branch of research. Secondly Esping-Andersen’s approach will be illustrated: The re-specification of the welfare state including the concepts of de-commodification and social stratification. The following section addresses the three ideal-typical regime-types: The social democratic model, the liberal type and the conservative regime-type. After a short conclusion Obinger and Wagschal’s criticism on Esping-Andersen’s statistical approach will be introduced followed by a discussion whether their critique is valid and in which way it affects the theory.
1. Comparative Research
The questions posed in the introduction are often topics of the field of comparative social science. Comparative researches have to be causally analytic, interpretative, generalizable as well as historical concrete (q.v. Ragin 1987. P.34). This means that an “account for significant historical outcomes in a sensible chronological sequence” (Janoski & Hicks 1994, p.6) is not sufficient but that the researcher has to come up with some causal mechanism. The scientist must also resolve the tension between nomothetic and idiographic explanations by finding generalizations (q.v. Ragin 1987, p. 35).
Comparative research in the field of cross-national studies is of course just a specification of various methods. Qualitative approaches (e.g. Boolean) as well as quantitative approaches (e.g. cross sectional analysis, time-series analysis, event-history method…) were applied (q.v. Janoski & Hicks 1994, p.1). Ragin uses variable-based (deductive) and case-based (inductive) approaches to define the broad range of methods (although he concentrates mainly on his own Boolean approach). The common denominator of these methods is that they all try to come up with an answer for the existence of the national differences. Basic problems of former large scale quantitative studies is that they assumed a linear model in which the level of expenditures reflect “a state’s commitment to welfare” (Esping-Andersen 1990, p.19) which clearly neglects the specific structures and organizational features of national social-systems. Another problem is the sole understanding of the welfare state as an independent variable. In the nineteen eighties, studies came up, which connected a detailed historic-quantitative analysis with a large scale empirical data. Until now the most influential study of this method was conducted by the Scandinavian social scientist Gøsta Esping-Andersen (q.v. Kohl 1993, p.66).
Why comparative social science?
Before continuing it is crucial to give an answer to why comparative science is applied so frequently in cross-national studies at all. The answer can be summarized in two points: (as discussed in Janoski and Hicks 1994, p. 6-7)
1. Experiments on societies are not possible; Scientists are left with the comparative method to find sociological rules or generalizations about societies by comparing them. Amongst others, variables can be: Class structure, deviance politics, skill formation, etc.
2. A comparative method helps to overcome cultural hegemony or ethnocentrism – Everybody is in a way culturally biased by the language and society one was socialized in (which is of course not a sole problem of comparative research).
More general: It is assumed that a full understanding of developments in any particular country is only possible if its experience is set in the context of the experiences of other countries. It is also increasingly acknowledged that developments in any single country cannot be explained without setting them in the context of wider changes (q.v. Cochrane, Clarke, Gewirtz 2001, p.2).
 Referring to Therborn’s definition of a welfare state (1983) proposing a historical transformation of state activities: In a genuine welfare state, minimally the majority of its daily routine has to be devoted to servicing the welfare needs of households. The consequence of this approach is that no state can be regarded as a real welfare state until the 1970s. (q.v. Esping-Andersen 1990, p.20)