Objectives　This study aimed to examine employment status differences in the marriage and fertility rates of Japanese women via an age-period-cohort (APC) analysis.

Methods　We used data collected from 1995 to 2015 in Japan based on the government's “Report of Vital Statistics: Occupational and Industrial Aspects,” which recorded the marriage rates of unmarried adults and fertility rates of married adults-according to their employment status. A Bayesian APC analysis was performed to identify changes in marriage and fertility rates based on three effects: age, period, and cohort. Finally, we calculated the marriage and fertility rate ratios between non-employed and employed women for each age group, period, and cohort.

Results　The APC analyses showed that the period effect on marriage rates for non-employed women decreased during the periods analyzed, while that for employed women increased from 2005. Meanwhile, the period effect on fertility rates increased regardless of employment status, albeit to a larger degree for employed women. The cohort effect on marriage rates began to decrease from cohorts born in the 1960s for non-employed women, and from cohorts born in the 1970s for employed women. And the degree of the decrease was larger among non-employed women than those employed. Meanwhile, the marriage rate ratio increased from 0.46 (95% CI: 0.21, 0.90) in the cohort born between 1946 and 1950 to 1.00 (95% CI: 0.45, 1.92) in the cohort born between 1991 and 1995. Finally, the fertility rate ratio increased from 0.31 (95% CI: 0.12, 0.69) in the cohort born between 1946 and 1950 to 0.38 (95% CI: 0.14, 1.81) in the cohort born between 1991 and 1995.

Conclusion　Employment status differences in the marriage rates of unmarried adults and fertility rates of married adults decreased among younger Japanese cohorts and in recent years. By contrast, there were statistically significant differences in fertility rates of married adults based on employment status, even in cohorts born more recently.