Effect of Childbearing on Female Labour Supply: An Empirical Study in Taiwan

ABSTRACT

In society, women generally play the roles of mothers and employees, and those roles are partly substitute to each other; therefore, the contributions of women to the labour market should be considered. This paper applied the widely observed parental preference for children’s sex as an instrumental variable for the number of children in Taiwan and examined the causal effect of childbearing on female labour supply. Ordinary least square estimates revealed that childbearing is associated with reduced female labour supply. However, after using the sex of first and second children as instruments to account for the endogeneity of the number of children, two-stage least square estimates revealed no correlation between the number of children and female labour supply.

KEYWORDS:

• Childbearing
• female labour supply
• instrumental variable

JEL CLASSIFICATION:

• J13
• J22

Acknowledgments

The author would like to thank Yoshiaki Omori, Ayako Kondo, Tsunao Okumura, and anonymous referees for their helpful comments on earlier drafts of this paper. All remaining errors are author's responsibility. The data analyzed in this article were provided by the Center for Survey Research of Academia Sinica. The author appreciates the assistance given by the Center.

ORCID

Hung-Han Lin http://orcid.org/0000-0003-3228-934X

Notes

1 These data were sourced from the Directorate General of Budget, Accounting and Statistics (DGBAS) of Executive Yuan, Taiwan. The birthrate is the total number of live births per 1,000 in a population in a year.

2 Data from the Ministry of the Interior, Executive Yuan, Taiwan.

3 See Montgomery and Trussell (1986) for the cost minimization problem.

4 The original numbers observed subjects are 3,208 for 2009; 5,073 for 2010; and 4,885 for 2011. After excluding subjects who did not respond, were male, or had only one child or no children, the observed subjects totalled 1,070 for 2009; 1,224 for 2010, and 1,215 for 2011.

5 The DGBAS provides the official LFP rates in Taiwan. For females, the rates are 49.62 per cent, 49.89 per cent, and 49.97 per cent for 2009, 2010, and 2011, respectively.

6 In the original study by Angrist and Evans (1998), s₁ and s₂ were the indicators for first- and second-born children; their instrument was given as . The present study focused on preference for sons and therefore simply used d_1id_2i to denote the birth of two daughters.

7 The different numbers of observed subjects between the LFP and WH models occurred because some subjects did not report their WHs in the surveys.