We estimate the causal effects of childcare availability on the maternal employment rate using prefecture panel data constructed from the Japanese quinquennial census 1990-2010. We depart from previous contributions by controlling for prefecture fixed effects, without which the estimates can be severely biased upward. We find that the treatment effect is heterogeneous: the employment rate of mothers in nuclear households increases with childcare availability, while that of mothers in three-generation households does not. We apply our estimates to decomposition of the growth of the maternal employment rate from 1990 to 2010. The decomposition indicates that the increase in childcare availability raised the maternal employment rate by about two percentage points. We also find that dissolution of three-generation household lowered the maternal employment rate. This negative effect is more strongly pronounced in small prefectures where the household structure changed dramatically.

Diverging economic inequality has become a common focus of economic debate in developed countries. In particular, the recent experience of Japan has started attracting international attention. We take advantage of a rich micro-level data set from the Basic Survey on Wage Structure (1989Structure ( -2003 to perform an in-depth analysis of the change in the inequality and distribution of the hourly wage. We observe that lower returns to education and years of tenure contribute to diminishing income disparity between groups for both sexes. A larger variance within a group contributes to the wage disparity for males, while an increased heterogeneity of workers' attributes contributes to the wage disparity for females. The Dinardo, Fortin, and Lemieux decomposition also confirms the basic findings from a parametric variance decomposition.

Taking advantage of our first access to micro data from the 2002 Employment Status Survey (ESS) as well as previous years of the ESS, we provide new evidence on changes in the "lifetime employment" practice which serves as a foundation for the Japanese employment system. Overall, we find evidence for the resilience of the Japanese employment system. Job stability of regular employees did not fall much during the first five years of Japan's Great Recession following the bubble. Our cohort analysis of narrowly-defined "lifetime employment" workers (new school graduate hires) also points to the stability of the size of the "lifetime employment" pool over time. To be consistent with the enduring nature of the 'lifetime employment" system, we also find some evidence for the buoyancy of other complementary elements of the Japanese employment system. This, however, does not mean the complete rigidity of the Japanese employment system. Apparently the Japanese employment system has evolved in response to Japan's prolonged stagnation. First, job stability of regular employees eventually fell during the final years of the Great Recession. It took years yet the magnitude of the fall turned out to be notable and for some groups of workers it was large. Second, the burden of downsizing during Japan's prolonged stagnation fell disproportionately on women as opposed to men and on highschool graduates as opposed to college graduates. The largest fall in job stability was found for female regular employees who were mid-career hires and working for large firms. We interpret our findings as a largely rational response of Japanese firms that were in dire need for downward employment adjustment yet were concerned about the cost of reneging on their implicit longterm employment contracts.

No abstract