Doing Fieldwork in Japan

Fact-Rich, Data-Poor: Japan as Sociologists’ Heaven and Hell

Social scientists do not always make good predictions. The scholars who supervised my dissertation research later told me that they had been sure my research plan would fail. Fortunately, they were wrong. But unfortunately , their prediction was based on sound reasoning: they were cognizant of a major impediment facing social scientists attempting to do the type of research in Japan that I had proposed. That impediment is the virtual unavailability in Japan of individual-level data—“raw” data, as social scientists often call it—with which to carry out statistical analyses. This problem has hampered American research on Japan in a number of substantive and theoretical areas in sociology, political science, and economics. In the case of my own discipline, sociology, it has hindered the training of graduate students who are interested in joining the ranks of the handful of academic researchers who claim dual loyalty to Japanese studies and to sociology as a discipline.1 In the extreme, the unavailability of data so discourages Ph.D. students that they make the unfortunately pragmatic choice of abandoning m a r y c . b r i n t o n Fact-Rich, Data-Poor: Japan as Sociologists’ Heaven and Hell Mary C. Brinton consulting with a research assistant. the study of Japan altogether in favor of a research site (often the United States) where high-quality, individual-level data are accessible. This seems a rash choice indeed, but it is a survival strategy in those fields in sociology that more or less require practitioners to carry out empirical research with large survey data sets. The situation is even worse in labor economics. The Japanese government produces and releases some of the best GNP, corporate financial, and inputoutput data in the world. But microlevel labor data are available only to a handful of Japanese researchers and very rarely to any foreign researchers. This chapter outlines the set of problems facing social scientists and would-be social scientists of Japan who need quantitative data for their research and suggests some strategies for overcoming these problems. I also discuss recent promising developments in the Japanese social science research community that will make some types of quantitative data more accessible to all researchers. While I speak from the vantage point of a sociologist , the issues and solutions I discuss are relevant to all researchers, academic and nonacademic alike, who find themselves in the position of needing individual-level data on which to carry out analyses. My purpose is not to try to convince readers that the best social scientific studies of Japan are necessarily quantitative in nature. Rather, I seek to deal with the “data problem ” in areas that more or less require quantitative data. These include fields such as social inequality, labor markets, gender stratification, education, demographic processes, and political and social attitudes. Who Needs Data? Individual-level data refer to objective and subjective information on individuals that is typically gathered through interviews or written surveys, often of large samples of individuals representative of particular populations (e.g., national samples of Japanese men and women, regional samples of the voting population, etc.). I use “objective” and “subjective” in the typical senses of those words, with “objective” referring to characteristics of the individual such as sex, education, birthplace, father’s education, income, and so on, and “subjective” referring to an individual’s opinions and attitudes, selfperceived social class, and so forth. Individual-level data are important for the analysis of any number of social science problems where one is trying to tease out causal processes. Suppose one is interested in the wage gap between working men and women 196 | m a r y c . b r i n t o n in Japan and the causes behind that gap. Japan has consistently had the dubious distinction of having one of the largest male-female wage gaps of any highly industrialized country (Brinton 1993; Ogawa and Clark 1995). Within East Asia, it is outdone on this dimension only by South Korea. Gaining an understanding of this phenomenon and its intransigence is thus an important sociological endeavor. When I went to Japan as an energetic Ph.D. candidate there were no data available to allow me to answer in a comprehensive manner the question of why the Japanese gender wage gap is so large, let alone answer the other questions about women in the labor market that I had proposed to study. I...