Microenterprise growth and the flypaper effect: Evidence from a randomized experiment in Ghana

Women with relatively successful microenterprises increased their business to equal or surpass their male counterparts with in-kind grants, but not with cash grants. Male-owned businesses increased their profits with both cash and in-kind grants. Specifically:

• In-kind grants to women who owned businesses increased average profits by 31-50 cedis/month. However, this increase was solely due to gains made by the female businesses with baseline profits above the median of 181 cedis in capital stock, which was about 40% of the female enterprise sample. These enterprises saw profits increase by 77-96 cedis monthly, while the increase was only 2-5 cedis among baseline low-profit (below-median) businesses.
• In-kind grants to men who owned businesses increased monthly profits by 28-60 cedis. Unlike with the female-owned businesses, both low and high baseline profit male-owned businesses benefited equally from the in-kind grants.
• Cash grants to women who owned businesses had no statistically significant impact on profits. The money was spent or transferred outside the house (to people or toward household expenses), instead of invested in their business. Researchers found that the ways in which these women spent the cash was due to personal preference of use, rather than coercion by extended family or community pressures.
• Cash grants to men who owned businesses increased profits by 5-29 cedis, though cash grants to men were not significantly different between low and high baseline profit businesses.

In short, cash and in-kind grants had markedly varied outcomes for female entrepreneurs. With in-kind grants in the form of equipment, supplies, and merchandise, there was significant growth for businesses that were already exhibiting a higher profit margin. These higher baseline women actually equaled or surpassed average men’s success with in-kind grants, suggesting that the gender gap in business profits could be narrowed by in-kind grants to women who already own relatively successful enterprises. However, this intervention was significantly less effective for enterprises with lower baseline profits, so women who own such businesses cannot expect to benefit in the same way.

Researchers deliberately chose a sample from two cities, Accra and Tema, in Ghana because Ghanaian women are historically nearly as active as men in the labor market. They then randomly selected 70 sample areas in Accra and 30 in Tema. In each area, household surveys were conducted to screen for the following inclusion criteria: 20-55 years old, self-employed working greater than 30 hours per week, no employees, and no business vehicle. From this base sample of individuals, males and females were then randomly chosen from three groups of typical business types: male-dominated, such as construction or vehicle repair; female-dominated, such as food or beauty services; and mixed, such as retail. This narrowed down the groups to 907 individuals, who were given a second survey prior to the experiment.  After this survey, a total of 793 microenterprises represented were chosen to participate in the trial, 479 female and 314 male.