Work and family practices in Japanese firms: their scope, nature and impact on employee turnover

Abstract

Using firm-level data from Japan, this study examines the effects of four commonly used work and family practices on employee turnover: flextime, maternity leave, child care leave, and nursing care leave. Overall, we find statistically significant associations between work and family practices and female employee turnover in Japan. In stark contrast, we do not find such a statistically significant linkage between work and family practices and male employee turnover. As such, this study highlights the potential moderating effect of individual characteristics such as gender on the relationship between work and family practices and employee attitudes and behaviours.

Keywords:

• Japanese firms
• voluntary turnover
• work and family practices

Notes

1. Law on Securing, Etc. of Equal Opportunity and Treatment between Men and Women in Employment was enacted in 1985 and became in effective 1986.

2. Statistics in 2005. Riding capacity is defined as the number of seats and hand straps.

3. When it was enacted in 1991, the law stipulated only a child care leave. In 1995 amendment was made to include a nursing care leave.

4. Although the law reads ‘family care’ we use nursing care because it is more commonly used in relevant literature.

5. Strictly speaking, it is possible that some firms fail to mention flextime as a non-traditional work arrangement in the survey in spite of the presence of flextime in the firm. For example, some firms may not consider flextime a noteworthy ‘non-traditional work arrangement’ and hence neglect to mention it in the survey. We think this possible underreporting is probably not too serious since flextime has been attracting national attention in Japan in recent years and it is highly unlikely that Japanese firms consider their use of flextime trivial. For reference's sake, we compared the proportion of firms that list flextime as a non-traditional work arrangement in our final sample with government statistics. As we will show in the results section, 47.0% of firms had adopted flextime in our sample, which consist of most major Japanese firms. The government report shows that 32.8% of firms with 1,000 employees or more had adopted flextime as of 2002 (Ministry of Health, Labor, and Welfare). Given that our figure was higher than the corresponding government figure, it is unlikely that firms in our sample failed to report flextime to SSJ.

6. PROC LIFEREG in SAS yields parameter estimates of the distribution of uncensored data (SAS Help and Document).

7. For the traditional employment system and its recent changes, see for example Kato (2001).