Versión de HTML Básico
Analíti a
k
9
Revista de Análisis Estadístico
Journal of Statistical Analysis
Analítika, Revista de análisis estadístico, (2015), Vol. 9
Impact of Fertility on Female Labor Supply
7
Impact of Fertility on Female Labor Supply
Andrea Molina Vera
1
Abstract
This paper studies the effect of fertility on Ecuadorian female labor participation.
It uses, as source of exogenous variability in family size, the instrument introduced
by Angrist and Evans in 1998 for the United States: parental preferences for a mixed
sibling-sex composition on their children. The empirical application shows that women
with two boys or two girls are 4 percentage points more likely to have a third child
than women with one boy and one girl (for both samples: all and married women).
2SLS estimations show that a third child causes a negative impact of 9 percentage
points on female labor supply in all work definitions for married women, meanwhile for
all women, there is a negative impact of 8 percentage points in “work outside home”
and “paid work outside home” as measures of labor supply. On the other hand, OLS
estimates are significant and negative for all definitions of work and for both samples.
JEL Classification: J13, J22
Keywords: Causality, Fertility, Female Labor Supply, Developing Countries
1 Introducci´on
For the U.S., Angrist and Evans found that fertility has a negative impact of about 12
percentage points on female labor supply (Angrist and Evans (1998)). Likewise, in Ecuador
there seems to be a negative relationship between fertility and female labor supply, arising
the question: can the U.S. results be extrapolated to Latin American developing countries
as Ecuador?
The differences of labor markets and family contexts between U.S. and Ecuador motivates
this question. Also, it is important to take into account some differences that developing
countries have in comparison to the developed ones: high levels of underemployment and
informal economy, high heterogeneity in work arrangements (Delpiano (2012)), higher fer-
tility, lower levels of female education and fewer facilities for formal childcare (Cruces and
Galiani (2007)).
Also, in empirical works, the problem of endogeneity between fertility and labor market
outcomes is widely known. Angrist and Evans (1998) mention that this problem arises
1
FLACSO-Ecuador. I gratefully acknowledge valuable comments from Hessel Oosterbeek.