The relationship between fertility and female labour force participation has been a subject
of attention for several decades. Most of the literature focuses on developed countries and
responds to the interest in understanding the evolution of the increasing female role in
labour markets during the last century and the possible effect that controlled fertility may
have on it.
Fertility, however, is a choice variable and thus endogenous. Therefore, disentangling its
impact is not straight forward and several approaches have been proposed over time in
order to isolate it. One of those methods is the use of multiple births occurrence as an
instrument for exogenous fertility which was introduced by Rosenzweig and Wolpin (1980)
and also used by Bronars and Grogger (1994) and Jacobsen et al. (1999) in the context of
developed countries and by Careces-Delpiano (2012) and Ponczek and Souza (2012) for
developing countries.
In this paper we apply the same strategy to study the effects of fertility on labour force
participation and employment in South Africa. We use census data from 2011 which
provides us with enough multiple birth occurrences to identify the effects with precision and
we focus on African women1
. This approach suggests that there is a negative effect of
fertility on employment of 4.9 percentage points per additional child. The effect on labour
force participation is also negative and slightly higher at 5.5 percentage points. Less than 30
percent of women in our sample are employed so this is quite a large effect. Labour force participation is between 60 and 70 percent depending on the definition used to measure it.
Thus, the effect of fertility is not trivial in this case either.
The paper is organized as follows: in the next section we explain in detail the endogeneity
problem and our empirical strategy to deal with it. In section 3 we describe the data and
section 4 summarizes the results.