This paper examines how class closures affect the academic achievements of Japanese students in primary and middle schools, with a special focus on the heterogeneous effects of the socioeconomic backgrounds of students' households. Utilizing the administrative data of students from a city in the Tokyo Metropolitan Area, we estimated the effects of class closures due to flu epidemics, on the students' language and math test scores. We find that class closures adversely affect math test scores of economically disadvantaged students. The magnitudes of the negative effects on disadvantaged students are heterogeneous by subject, grade in school, gender, timing of class closures, and students' achievements at the beginning of the school year. Male students from economically disadvantaged households are more susceptible to class closures, and those with relatively low achievements before class closures suffer more seriously from them. The deleterious effects among economically disadvantaged male students are driven not only by reductions in class hours in school, but also by increases in time spent watching TV and playing video games. We also find that school resources can mitigate the negative impact of class closure among economically disadvantaged students. These results indicate the importance of public programs in preventing a negative temporal shock to student learning environments.

We evaluate the causal effect of class size (number of students in a classroom) on incidence of class closure due to the flu, as an outcome of an infectious disease epidemic. For identification of causal effects, we apply a regression discontinuity design using discontinuous variation of class sizes, around the class size cap set by regulation, to administrative data of public primary and middle school students in one of the largest municipalities within the Tokyo metropolitan area from 2015 to 2017. Most classrooms in Japan are constructed in accordance with a standard of classroom area, 63 square meters; class size reduction improves social distancing among students in a classroom. We find that class size reduction is effective in reducing class closures due to the flu: a one-unit reduction of class size decreases class closure by about 5 percent. Additionally, forming small classes with 27 students at most, satisfying the social distancing of 1.5 meters recommended to prevent droplet infection including influenza and COVID-19, reduces class closure by about 90 percent. Moreover, we find that the older the students, the larger the effects of class size reduction. Our findings provide evidence for the effectiveness of social distancing policy in primary and middle schools to protect students from droplet infectious disease spread, including COVID-19.

BACKGROUND: Japan is facing a rapid increase in the number of reported child maltreatment cases. Child maltreatment has long-term consequences for the victims, and unemployment rate is considered a strong predictor of it. However, only few studies have analyzed the causal relation between child maltreatment and the unemployment rate-particularly the effects of the latter on the former-in Japan. METHODS: Using prefecture-level longitudinal data from 2005 to 2016, we employed a fixed effects instrumental variable estimation. The estimation included a weighted average of the national unemployment rate across industries by industrial structures in 2005 as an instrument to identify the causal effects. RESULTS: The average local unemployment rate changed by approximately 50% from the peak to the bottom in the sample period. A 50% increase in local unemployment rates increased the number of reported child neglect cases and child deaths by 80% and 70% (statistically significant at the 5% level), respectively. Further, it increased cases of death due to external causes, unintentional injuries, and unintentional drowning by 146%, 217%, and 315% (statistically significant at the 5% level), respectively. CONCLUSION: The local unemployment rate is a risk factor for child maltreatment, resulting in children's death, especially as a result of unintentional drowning-the common cause of death due to child neglect. When the local unemployment rates rise, governments should allocate more financial and human resources for preventive measures to combat child deaths caused by neglect.

This study aims to better understand the effects of retirement on health outcomes through a large-scale crosscountry study of the changes in health investment behaviors after retirement among the populations of seven developed countries. Much of the literature on retirement health consists of single-country studies which generally find that health investment behaviors are effective predictors of health outcomes. However, using Global Aging Data normalized to facilitate cross-country comparison, and exploiting differences in the financial incentives in the pension systems across countries as our identification strategy, we find that, even with a careful accounting of the differences in baseline retirement ages, the elderly do not uniformly change their health investment behaviors in retirement. Therefore, in a cross-country framework, health investment behaviors are not necessarily good predictors of health in retirement.

This paper investigates the heterogeneous effects of changes in instructional time on student achievement due to differences in instructional quality, using student test scores and data on teachers and schools from the Trends in International Mathematics and Science Study (TIMSS), as well as a natural experiment caused by a 2002 Japanese curriculum standards revision that reduced instruction time for math but not for science for second-year junior high school (grade 8) students. The random matching of teachers and students in Japan and our application of the time dimension of panel data to two subjects, mathematics and science, allow us to control for unobserved heterogeneities such as individual fixed effects and school fixed effects. We confirm that instructional time is more effective in combination with higher-quality teachers, as measured by common indicators such as experience, schooling, and academic major. We also find that these effects are larger for students whose socioeconomic status (SES) is lower.

This paper analyzes the reasons for differences in the estimated effect of retirement on health in previous studies. We investigate these differences by focusing on the analysis methods used by these studies. Using various health indexes, numerous researchers have examined the effects of retirement on health. However, there are no unified views on the impact of retirement on various health indexes. Consequently, we show that the choice of analysis method is one of the key factors in explaining why the estimated results of the effect of retirement on health differ. Moreover, we re-estimate the effect of retirement on health by using a fixed analysis method controlling for individual heterogeneity and endogeneity of the retirement behavior. We analyze the effect of retirement on health parameters, such as cognitive function, self-report of health, activities of daily living (ADL), depression, and body mass index in eight countries. We find that the effects of retirement on self-report of health, depression, and ADL are positive in many of these countries.