Women's Work Choices in Kenya: The Role of Social Institutions and Household Gender Attitudes

ABSTRACT

This study considers the factors that influence women's work behavior in Kenya. In particular, it examines whether gender attitudes and certain types of social institution influence the probability of employment or type of employment for women. Using data from the Demographic and Health Survey of 2008–9, it finds that religion and ethnicity are significant determinants of women's employment in Kenya. While personal experience of female genital mutilation is insignificant, spousal age and education differences, as well as marital status (which reflect attitudes both in women's natal and marital families), are significant determinants of women's employment choices.

KEYWORDS:

• Women's labor force participation
• institutions
• intrahousehold bargaining

JEL Codes:

• J01
• O12
• O15

View correction statement:

Erratum

NOTES ON CONTRIBUTORS

Giovanna De Giusti graduated with a PhD from the University of Reading in 2012. She worked on women's entrepreneurship and sociocultural institutions in Kenya. Having worked in an experimental laboratory for behavioral studies in Kenya for a year, Giovanna took up a post as Lecturer in Development Studies at the University of Maseno where she continues to work on women's autonomy and entrepreneurship. In addition, she works on international projects relating to the socioeconomic and institutional aspects of climate change in East Africa.

Uma Sarada Kambhampati is Professor of Economics working on development economics at the University of Reading. She graduated with a PhD from the University of Cambridge in 1992. Uma's research over the years has encompassed a number of areas – applied industrial economics, child labor and schooling, well-being and happiness, and institutions and development. In recent years, she has worked on various aspects of women's autonomy and its impact on household welfare.

ACKNOWLEDGMENTS

We are grateful to seminar participants in the Economics and Geography departments at the University of Reading for comments on an earlier version of this paper. We are also grateful to Sarah Cardey, Augustin Fosu, Bereket Kebede, Wim Naude, Zahra Siddique, and Jackie Wahba for commenting on the work on which this paper is based.

Notes

1 Of course, it is possible that religion itself is not entirely exogenous, given that an individual's choice of religion and a household's choice to perform FGM may be correlated with each other and with conservatism more generally (Joshua D. Angrist and Jörn-Steffen Pischke 2009).

2 Research conducted by the ILO stresses that women's economic autonomy is considered to be highest when they are engaged in wage and salaried work or are employers, lower if they are own-account workers, and lowest when they are unpaid family workers (ILO 2008).

3 While information on whether her mother was educated and/or worked would have been useful in considering the influence on the women's own attitudes, these data are not made available in the dataset.

4 The supplemental online Appendix is available under the Supplemental Tab on the publisher's website.