Process Tracing for Causal Explanation and Causal Inference: Comparative Case Study of Welfare Legislation Patterns in South Korea and Japan

The major finding of my research was that the dispersed power structure—divided governments in Korea and coalition/minority governments in Japan—of both nations was the most influential factor behind the transformation of fragmented, indirect, and weak welfare benefits into universal, direct, and generous ones. Unlike the conventional analysis of welfare expenditure data, I showed correlation between the change in power structure and the legislative activities of major political actors using the whole universe of bills—20,000 in total—in both nations during the past two decades. I pushed this statistical correlation to causal relation by examining the causal mechanism with the in-depth analytical narratives of major political actors. By reading documents, labouring through archives, interviewing, and surveying the secondary literature, this intensive testing sought to understand the actors'' preferences, their perceptions, their evaluation of alternatives, the information they possess, the expectations they form, the strategies they adopt, and the constraints that limit their actions. As the result of examining causal mechanism, I revealed the micro-logic behind this macro-association: the electoral competitions among political entrepreneurs which resulted in a ‘race to the top’ and a ‘ratcheting up’ of welfare policy reform promises. When bargaining among a number of parties prevailed, policy tended to be premised on the lowest common denominator—the median voters such as consumers, women, labourers who often were unorganized and not geographically concentrated. In addition to the clarification of the micro-logic, the causal mechanism also helped to fine-tune the hitherto accepted theory of welfare state development by shedding the light on the context of the micro-logic. The “race to the top” was especially favourable in both nations as i) they were not in a welfare state retrenchment but in a welfare state expansion stage (politics was more about “credit-claiming,” not “blame-avoiding”); ii) the experience of growth with equity in Japan and Korea also meant that the political system was not a priori structured along class lines; iii) and office-seeking, not policy-seeking, politicians offered welfare policy promises mainly to win seats without internalizing the long-term fiscal implications of programmatic expansion. Obviously, this challenges the well-accepted proposition advanced by power-resource scholars that strong and encompassing trade unions and leftist parties are necessary institutional preconditions of universal welfare expansion Going further, I carefully traced the historical structural factors giving rise to the frequent occurrence of dispersed power structure—the independent variable of research—and found the following causal paths: i) the dispersed structure was the result of a 2.5 party system coupled with the electoral cycle in both nations; ii) the convergence of a 2.5 party system in both nations can be attributed to the convergence in electoral proportionality and the level of party discipline in both nations; iii) the change in electoral proportionality and to the level of party discipline was the result of the change in production system and accompanied political corruption; iv) the change in production system can be explained by exogenous conditions such as globalization, change in technology, and liberalization pressure from U.S and international organizations.