Social Security as Viewed through Micro-data

Part 3: Japan's Approach

SHIMIZUTANI Satoshi Consulting Fellow, RIETI

A series of problems have been pointed out about economics surveys conducted in Japan. It is often said that: Japanese surveys use questionnaires focusing on a specific aspect in a lopsided manner; data obtained do not allow a comparison of longitudinal changes; sufficient consideration is not given to international comparability; and survey results can be used only by specific researchers and are not accessible to researchers in general. Surveys of middle-aged and elderly people are no exception, and it has been difficult to evaluate the effects of social security policies. Amid such circumstances, Japan has been lagging behind other countries in efforts to construct "global standard" panel data on middle-aged and elderly people.

In the United States, a panel survey called the Health and Retirement Study (HRS) was launched in 1992. This was followed by the English Longitudinal Study of Ageing (ELSA) in the United Kingdom in 2002 and the Survey of Health, Ageing and Retirement in Europe (SHARE) in continental Europe in 2004, and similar initiatives were also started in Mexico and Korea around the same time.

In light of such developments overseas, we launched a project called the Japanese Study of Aging and Retirement (JSTAR) in 2005 under the leadership of Dr. Masaru Yoshitomi, then president of the Research Institute of Economy, Trade and Industry (RIETI). The first wave of the survey was conducted in 2007, with the third one ongoing in 2011. The JSTAR has benefited from collaboration with other research projects such as the Project on Intergenerational Equity (PIE) which is supported by the Japan Society for the Promotion of Science's grant-in-aid for specially-promoted research.

The JSTAR survey has been developed using the same basic design as those of the aforementioned surveys in the United States and Europe through close cooperation with research teams engaged in those surveys. For instance, it employs the computer-aided personal interviewing (CAPI) method, in which face-to-face interviews are conducted using computers with built-in questionnaires. The initial wave of the survey involved asking respondents numerous questions covering the entire aspects of their lives. This process, which took an average of two hours per respondent, would have taken much longer if we had not used CAPI. The CAPI questionnaire is also designed in such a way to minimize non-responses, in particular, to questions concerning important economic variables such as the amounts of income, consumption, and assets. For instance, a respondent who is unsure about the amount of income will be shown some specific values and asked whether his or her figure exceeds each threshold, thereby narrowing down the range. In the second wave of survey onward, we are able to ask respondents questions based on their respective responses in the previous wave.

Overseas researchers are showing keen interest in the JSTAR as they expect that this initiative may lead to the elucidation of some excellent characteristics of Japanese society such as the longevity of the Japanese people, relatively late retirement, and low medical expenses as a percentage of GDP. Based on the global standard design, the JSTAR can offer knowledge and expertise useful to solving aging problems around the world. In Part 4 of this series, I will present some facts derived from the data collected to date, including findings from the ongoing research.

>> Original text in Japanese

* Translated by RIETI from the original Japanese "Yasashii Keizaigaku" column in the September 14, 2011 issue of Nihon Keizai Shimbun.

September 14, 2011