Using defamilisation typologies to study the Confucian welfare regime

Abstract

Since Esping-Andersen classified the 18 Organization for Economic Co-operation and Development (OECD) countries into the “three worlds of welfare capitalism” in 1990, the comparative studies of welfare have been dominated by the responses to this study. This paper focuses on two of these responses. The first response is concerned with the gender insensitivity of Esping-Andersen's way of categorizing the welfare regimes. The second response is concerned with the issue that the East Asian countries are under-represented in the 18 OECD countries. To make contributions to these responses, two analytical tasks are conducted. First, we build new defamilisation typologies covering both East Asian countries and OECD countries. Second, we demonstrate that the evidence generated from the typologies suggest that some of the indispensable conditions for the development of a Confucian welfare regime do not exist.

Keywords:

• East Asia
• defamilisation typologies
• Confucian ideas

Notes on contributors

Sam Yu is Associate Professor in the Department of Social Work at the Hong Kong Baptist University. His research interests include social exclusion, social quality, social capital, welfare ideologies, welfare mix, medical finance, housing and comparative studies of social policy.

C.M. Chau is Independent Researcher. She is formerly Assistant Professor in the Department of Social Work and Social Administration at the University of Hong Kong. Her research focuses on social quality, social exclusion, welfare mix, ethnic minorities in Britain and disabled people in China.

K.M. Lee is a Senior Lecturer in Community College of City University. He is a sociologist. His research interests include financial crisis, social exclusion, working poor and ethnic minority groups.

Notes

1. The 21 countries studied by Bambra (2007) include the 18 OECD countries studied by Esping-Andersen (1990) and three Mediterranean countries (Greece, Portugal and Spain). These three countries have been studied by Ferrera (1996) and Bonoli (1997).

2. There are significant differences between the work done in preparation for this article and the work done by Chau and Yu (2013). First, the authors of this article studied 25 countries rather than 22 countries. Chau and Yu (2013) studied the 18 OECD countries examined by Esping-Andersen (1990) and the four East Asian countries (Hong Kong, South Korea, Singapore and Taiwan). Second, the authors have used five variables rather than three variables to study the concept of defamilisation. The details of the five variables are shown in the text of this article. Third, Chau and Yu (2013) used one statistical technique (cluster analysis) to study the defamilisation patterns of the 21 countries, whereas the authors have used two statistical techniques to interpret data and classify countries – cluster analysis and the index-based regime construction. As a result, the authors have developed two defamilisation typologies rather than one defamilisation typology.

3. There are significant differences between the work done in this article and the work done by Bambra (2007). First, Bambra (2007) studied 21 countries rather than 25 countries. Second, Bambra (2007) used three variables rather than five variables to measure defamilisation. Third, as mentioned in note two, the authors have used two statistical techniques to interpret the data and as a result built two defamilisation typologies. Like Chau and Yu (2013), Bambra (2007) used one statistical technique (cluster analysis) and built one defamilisation typology.

4. For example, the OECD Family Database (OECD, 2012) provides data concerning the childbirth-related leave benefits, gender differences in employment outcomes, educational enrolment by gender and average years spent in formal education. The OECD Employment Database also provides data concerning the labour market outcomes such as labour force participation rates (OECD, 2014).

5. Unlike the method used by Yu (2012), the defamilisaton index developed by the authors gives equal weight to the five variables for measuring defamilisation.