Cultural Factors in Women's Labor Force Participation in Chile

Abstract

This article analyzes determinants of female participation in the Chilean labor force using classic determinants such as age, education, marital status, and number of children. The results indicate that the greater a woman's education level, the greater her labor participation; that older women participate more, though the rate of growth of this effect is decreasing; and the number of children that a woman has is negatively correlated to her decision to participate in the labor force. The article also examines machismo and other cultural values that influence female labor participation. The evidence suggests that the more the women have internalized machista and conservative cultural values, the less they participate in the labor market. Finally, the article concludes that the existence of these cultural factors as a group more than compensates for the positive effect of human capital variables and is statistically associated with low female labor participation in Chile.

Keywords:

• Female labor force participation rate
• social norms
• culture
• machismo

Acknowledgments

The authors would like to thank Carla Lehmann, Ricardo Paredes, Klaus Schmidt-Hebbel, Sergio Urzúa, Salvador Valdés, and three anonymous referees for their valuable comments on a preliminary version of this paper. Special thanks go to Rosalba Todaro for methodological and bibliographic suggestions. We are also grateful for the comments offered by participants in the Meeting of the Chilean Economics Society and the seminar offered by the Universidad de Chile. Finally, we are thankful for the important collaboration of Carla Lehmann and the Centro de Estudios Públicos in gathering the data. Gonzalo Plaza would like to thank Comunidad Mujer for the funding it provided for this project. Any errors or omissions are the exclusive responsibility of the authors. We are also grateful for the funding granted by Iniciativa Cientifica Milenio “Centro de Microdatos,” Proyecto PO7S−023-F.

Notes

¹ ISSP is a continuing, annual program of cross-national collaboration. It brings together pre-existing social science projects and coordinates research goals, thereby adding a cross-national perspective to the individual, national studies. ISSP evolved from a bilateral collaboration between the Allgemeinen Bevolkerungsumfragen der Socialwissenschaften (ALLBUS) of the Zentrum für Umfragen, Methoden, und Analysen (ZUMA) in Mannheim, Germany, and the General Social Survey (GSS) of the National Opinion Research Center (NORC), University of Chicago. Both the ALLBUS and the GSS are replicating, time-series studies. Since 1984, ISSP has grown to include 43 nations around the world, including both developed and developing countries.

² For more details, see section on causality and Bina Agarwal 1997.

³ See section, “Is Chile a machista and conservative country?”

⁴ The ISSP was carried out in Chile by the Centro de Estudios Públicos (Mujer, Trabajo, Familia y Valores2002). The sampling method was randomized and probabilistic in each of the three stages (block-home-interviewee). The response level was 86 percent. Precision was estimated at approximately 3 percent, considering maximum variance and 95 percent confidence level.

⁵ This result is higher than the national average for female labor participation (47 percent according to the 2002 Census in Chile). However, it is important to recognize that the rate is calculated with different samples. The database utilized for this paper covers 144 municipalities spread throughout Chile. Also, the women surveyed here exhibited higher levels of education and were older. However, similar results were obtained utilizing a model that does not include cultural factors, which are mainly based on human capital.

⁶ A series of estimates was also conducted utilizing each of the answers and combinations of answers as explanatory variables. The main result – that is, the importance of cultural factors in women's labor force participation – remains unchanged.

⁷ For a theoretical discussion, see Mark R. Killingsworth (1983); Killingsworth and Heckman (1986).

⁸ The normative-domestic-patriarchal power is only one of the elements that reinforce the subordination of women in the domestic sphere. In classical capitalistic societies that are dominated by men, the worker's role is masculine. The masculine identity of those societies is closely linked to the role of the breadwinner. See Fraser (1990).

⁹ The income data is not available in a continuous manner. It is only possible to obtain them through income categories. As a result, we estimated the nonlabor income variable using a dummy variable that takes value one for a person whose individual income represents a small percentage of the total household income and zero if this is not the case (when the person has a high nonlabor income). In theoretical terms, this variable would have a negative effect on the decision to participate in the labor market. However, the effect was not statistically significant here due to data limitations.

¹⁰ Given that women's labor force participation varies with the level of economic activity, with women withdrawing from the labor force when jobs are tight and reentering when jobs are more plentiful, regional dummies are included to control for regional unemployment rates.

¹¹ This effect may be particularly important for least developed countries. According to the World Bank (2001), women in low-income, gender-stratified societies are at the back of the queue for economic resources. With higher per-capita incomes, proportionally more resources reach the back of the line and lead to changed perceptions about gender roles. This view suggests that growth alone can improve gender equity.

¹² Excluding variables in order to avoid endogeneity also may be problematic. Janet S. Netz and Jon D. Haveman (1999) indicate that omitted variables bias due to excluding family composition variables in labor supply models for the USA can be substantial.

¹³ These techniques include (1) assuming exogeneity; (2) using a standard instrumental variables approach; (3) using a standard instrumental variables approach that also incorporates municipality data; (4) using an instrumental variables approach that exploits the birth of twins as the exogenous instrument for fertility; and (5) using a semi-parametric approach that relies on the inherent nonlinearity of the decision variables.

¹⁴ Having panel data would allow for the use of lagging variables as instruments.

¹⁵ These results are consistent with earlier evidence in Chile (see Contreras, Puentes, and Bravo 2005).

¹⁶ Variants of the questions that measure perceptions of the machista environment and level of conservativeness were used. The results are stable for different definitions of cultural variables.

¹⁷ Results are calculated from individual survey responses on attitudes and values.