The integrated community care system (ICCS) in Japan has been discussed for more than a decade and considered one of the most important developments in health care policy during that time. Yet the discussion seldom addresses issues beyond the scope of home care for the frail elderly. The ICCS is not perceived as including a much wider system that covers all care modalities for all health problems in all age groups-namely, primary care. Japan is among a handful of countries where the system of primary care remains underdeveloped. Primary care is a patient-centered, family-oriented, community-based, and cost-effective service provided through a continuing partnership with users, not only in the management of common health problems but also in the appropriate use of health care resources, prevention, and health maintenance and promotion. Primary care functions as a hub of coordination that creates a network within the community being served and with outside partners.

To date, Japanese medical education has lacked national systems to accredit postgraduate training programs and to certify specialist doctors in all medical and surgical disciplines. Having recognized the urgent need to address this issue, the government has finally decided to introduce appropriate systems starting in 2017 academic year. Constructive discussion is therefore needed to make these systems effective.

Historically, several names have been used in Japan to describe doctors working in communities but these lack sufficient consideration regarding their defined roles and functions in primary care. Although the patient-centered clinical method (PCCM) is well known among family doctors around the world as being the most important core competency of their work, not many Japanese doctors understand the values associated with the method. The PCCM has been extensively researched over the past 30 years and a large-scale study is now underway in Canada to examine whether the PCCM can reduce health care costs. Given this background, the future Japanese training program for primary care should include the PCCM within its curriculum.

Research is another key element that can promote primary care in Japan. For family medicine to become a truly independent discipline in the country, however, there are clearly many research questions that need to be addressed within this field. As health care reforms toward well-functioning primary care are a crucial part of the health care policy agenda, I would like to call on researchers in health economics to conduct collaborative researches on this issue with us.

The rapid expansion of psychiatric hospitalization in post-war Japan during 1950s and 1960s was financed by three types of public expenditure programs, each of which was based on different institutional characteristics. The first type of program was compulsory hospitalization under the Mental Hygiene Act, which was "public custody" implemented by local governments. The second type was hospitalization with medical public assistance, which was primarily viewed as a means-tested "public assistance" program for the poor. The third and final type was healthcare benefits from the public health insurance programs that had been formulated as part of the nation-wide "social insurance" system.

In this paper, we address the following question: why and how did these three types of public expenditures come to play crucial fiscal roles in the formation of post-war large-scale psychiatric institutionalization? This question is important because previous studies have tended to focus on the extreme expansion of compulsory hospitalization by "public custody" in the 1960s and have often failed to investigate the significant contributions of public expenditures via "public assistance" and "social insurance" to the post-war increase in psychiatric inpatients. In fact, these latter two public spending schemes have persistently and significantly contributed to the increase in psychiatric inpatients from the early 1950s to the present, whereas compulsory hospitalization by "public custody" was responsible for the largest number of psychiatric hospitalizations for only a few years int he 1960s.

The above question can be divided into two research topics. First, we investigate the hypothesis that public expenditures on psychiatric hospitalization through the three different schemes, that is "public custody", "public assistance" and "social insurance", have their institutional origins in the pre-war era (hypothesis I). Second, we analyze the hypothesis that several important post-war reforms of the three types of public expenditure resulted in the rapid post-war expansion of psychiatric hospitalization (hypothesis II).

Regarding hypothesis I, our investigation based on historical documents and statistics reveals that the post-war systems of the three public expenditure schemes outlined above have identifiable origins in pre-war hospitalization schemes for people with mental illnesses. When it comes to hypothesis II, patterns of increases in psychiatric hospitalization between the 1930s and the 1970s provide some evidence that this hypothesis is plausible. In addition, we find that "public assistance" and "social insurance" had more significant roles than "public custody", at least in terms of providing fiscal resources for psychiatric hospitalization.

Overall, our study makes it clear that the post-war patterns of public expenditure for psychiatric hospitalization were not the consequence of strong political or policy initiatives and that the above three types of public expenditure schemes have their origins in pre-war institutions. In addition, while several. post-war reforms of these three types of programs resulted in the rapid expansion of psychiatric hospitalization, in some cases this was not necessarily their intended result.

In Japan, the rapid aging of the population has lead to an increase in people needing long-term care. This has increased the number of working-age people who care for their parents or parents-in-law and their number is expected to increase with the continued aging of the population. When family members require care, working-age caregivers may have to leave their jobs or reduce their working hours. Previous research in Japan in this area assumed that care was exogenously determined. However, the decisions to work and to provide care might be simultaneously determined and treating the care burden as an exogenous variable might cause bias in the estimates. Moreover, the effects of care on work might differ according to the type of care, such as physical care or housework support. However, this was not considered in the previous studies in Japan. Hence, this study used an instrumental variable method to consider the care burden's endogeneity. Also, for our estimates, we used not only the average number of care hours per week as a care burden indicator, but also hours spent on physical care and housework support. The data used was "Survey on the At-home Care Costs and Caregivers." The respondents live with their parents or parents-in-law aged 65 and over. Our estimates showed that providing care decreased the probability of working, working hours, and income, and for men and women the effects were larger for physical care and housework support than for total-care hours. This is thought to show that the physical and mental burdens imposed by physical care and housework support are greater than those of other types of care. Care had a larger effect on work for women than for men, but it had a larger effect on leisure hours for men than for women. This is considered to be because as men are often the main bread winner, even if their care hours increase, they do not decrease their working hours, but instead decrease their leisure hours. In the case of men, the effects of physical care on the probability of working, working hours, and income were greater than the effects of housework support. On the contrary, in the case of women, the effects of housework support on these variables were larger than the effects of physical care. Except for the analyses of leisure and women's income, care burden was an exogenous variable.