Constraints to women’s entrepreneurship and welfare in developing countries

Women in developing countries invest a larger part of their income in their children’s nutrition, health and education than men (Hoddinott et al., 1995; Strauss et al., 2000; Gammage, 2006; Quisumbing et al., 2006). As a result, financial resources acquired by women bring forth a long-time reduction in fertility and birth-spacing as well as an allocation of resources in favor of children which generates improvements in welfare (Dyson et al., 1983; Caldwell et al., 1987; Hogan, et al., 1999; Eswaran, 2002; Schady et al., 2008; Mukhopadhyay et al., 2011). Furthermore, according to the World Bank (2011), almost 35% of the world’s firms are owned by women. And a large percentage of these companies are in developing countries (Jones, 2012). However, few empirical studies estimate the causal effect of women’s entrepreneurship on economic outcomes. Some authors examine constraints to entrepreneurship in general (Brunetti et al., 1997; Batra et al., 2003; Banerjee et al., 2004; Aterido et al., 2009). Others study specific constraints to women’s entrepreneurship (Hampel-Milagrosa, 2010; Aterido et al., 2011; Jones, 2012). In general, these constraints are economic, financial, institutional or structural, and may be more restrictive for small businesses. But, to the best of our knowledge, there are no empirical evidence on the impact of women entrepreneurs on welfare for a large group of countries. This is largely due to difficulties raised by endogeneity issues. Most likely, causality runs in both directions. Indeed, an increase in the proportion of women entrepreneurs may cause an increase in welfare due to the aforementioned reasons. But an increase in welfare may also generate an increase in women’s access to the workplace through institutional development promoting women’s emancipation. Also, there exists a joint determination problem in that the proportion of women entrepreneurs and the level of welfare may be the result of a third, unobserved, factor. The main objective of this study is to examine whether women’s income-generating activities significantly improve welfare in developing countries. A second objective is to identify constraints to women’s entrepreneurship which could then serve as aid targets to improve women’s status and their economic impact in developing countries.In this study, we examine the effect of the variation in the proportion of women entrepreneurs on welfare in 120 countries between 2005 and 2010. We use different proxies for welfare: the rate of infant and child mortality, Gross Domestic Product (GDP) per capita, Human Development Index (HDI) and the level of education indicator of the HDI. The data on women entrepreneurs comes from the World Business Environment Survey (WBES) and is based on formal manufacturing firms. Our approach uses instrumental variables to take into account the endogenous relationship between women’s entrepreneurship and welfare. We use as instruments a set of constraints to women’s entrepreneurship that we identify as key in a woman’s decision to become an entrepreneur. These instruments respect the exclusion restriction as these constraints are only correlated to the dependent variable (welfare) through their direct effect on the proportion of women entrepreneurs. Indeed, these constraints act as a sorting mechanism between male and female entrepreneurs, and only then the outcome of this sorting affects welfare through the channels suggested above. In order to take into account the problem of joint determination, we control for fixed effects at the country level. We also examine certain characteristics previously identified as important in explaining the effect of women on welfare namely health and education expenses and some macroeconomic variables like inflation and lagged GDP per capita. We also examine the interaction between women’s entrepreneurship and the level of aid received by a country and seek among the different constraints to women’s entrepreneurship which are more prone to be responsive aid targets.Based on preliminary results, we find that an aid-policy encouraging an increase of 1% in the proportion of women entrepreneurs could decrease the rate of infant mortality by as much as 18%. We also find a one-for-one relationship between the proportion of women entrepreneurs and GDP per capita.