7 Summary and conclusions

View the most recent version.

Using the 1983-to-2004 Longitudinal Worker File, this study has examined the post-childbirth employment, job mobility and earnings trajectories of Canadian mothers. We found that the short- term post-childbirth employment rates of Canadian mothers rose from the mid-1980s to 2000 and have declined since then. However, the long-term post-childbirth employment rates of successive cohorts of Canadian mothers kept increasing over time without any substantial diversion. Furthermore, Canadian mothers were also less likely to quit their jobs during the post-childbirth years than their 'non-mother' counterparts, and the differences also increased over time.

Focusing on a group of Canadian women who had strong labour market attachment, we first found that the earnings of mothers did not decline in the pre-childbirth years: a result that casts some doubts on one version of the endogenous motherhood hypothesis. Second, we found that the motherhood-earnings penalties were not spurious: in the year of childbirth and the year after, mothers from our sample experienced about 40% and 30% earnings drops, and they continued to incur earnings losses during the other post-childbirth years. Third, the earnings effects of childbirth were not fixed; rather, they declined over the post-childbirth years: therefore, it may not be plausible to simply estimate a 'single' or an 'average' effect of childbirth on the earnings of the mothers.

Under the more general fixed-trend model specification, we found that the negative earnings effects of childbirth started to disappear from the seventh post-childbirth year, and thus—at least for a group of mothers who had strong labour market attachment—our finding contradicts the results of several studies that claim that earnings losses due to career interruption can never be fully regained.

Furthermore, earnings losses incurred by mothers who returned to, and worked for, their pre- childbirth employers were negligible beyond the second year of childbirth when individual earnings growth trend is controlled. This suggests that firm-specific human capital is likely an important factor in the earnings recovery process for Canadian mothers, but due to the lack of accurate information on labour supply—such as hours worked or full/part-time status—we cannot make a firm inference on the role of firm-specific human capital.

It is therefore desirable to investigate the effect of childbirth on the labour supply decision of the mothers as well as the effect of childbirth on the wage rate of the mothers simultaneously.