Grants Double Income but not Empowerment for Ultra-poor in Uganda

new paper by Chris Blattman (Columbia) and co-authors provides optimistic new evidence on the returns to providing cash grants to impoverished women in northern Uganda.  The new experiment varied whether the ultra-poor, largely women, were offered a business grant worth $150, training and supervision, and found dramatic impacts of the cash grant on entrepreneurship, hours worked, individual earnings, and household consumption.

The paper stands out from previous studies in that it finds strong positive impacts for women, and that it does so among the most impoverished people in the village.  Only those people identified by a local nonprofit as the poorest fifteen people in each village (86 percent of whom were women) were eligible for the study.  Previous studies of cash and in-kind small enterprise grants delivered to women in Sri Lanka and in Ghana find more mixed effects.  Grants to female-owned microenterprises had, on average, no impact in Sri Lanka, and in Ghana, only in-kind grants or grants made to initially more profitable female microenterprises appeared to benefit recipients.

The authors of the Uganda paper find that entrepreneurship among ultra-poor village members receiving grants doubled, increasing from 39 percent to 80 percent; grant recipients increased average weekly work hours from 15 hours to 24 hours; individual cash earnings increased by 95 percent; and household consumption increased by roughly one-third. 

In addition, income and consumption increased significantly in response to grants to both male and female recipients.  Women’s earnings increased by 92 percent relative to the control group mean of roughly $7 per month, and men’s earnings increased by 74 percent relative to the control group mean of roughly $11 per month.  Effects on consumption do look slightly larger for men, but the authors note that they’re not able to statistically differentiate between effects of grants to men and women, even for this measure.

Why did results for women in this context differ from previous studies?  In stark contrast to previous studies on cash grants and women’s entrepreneurship, which find no measurable benefits or impacts on business outcomes, this study explicitly tied receipt of the grant to approval of a business plan.  In the studies mentioned earlier on female entrepreneurs in Sri Lanka and Ghana, cash grants were either explicitly described to recipients as unconditional on use, or framed as compensation for participation in a survey, in either case leaving it to the recipient to decide what the highest priority use of the money should be.  The explicit framing of the grant as earmarked for business use in this study may have helped to overcome some of the factors that economists believe lead women to have a tougher time growing their businesses, such as family pressure to distribute gains that could otherwise be used for business investment.  Another difference in context, which the authors mention, is that the Uganda experiment took place in a post-conflict setting, in which returns to capital for all individuals might be unusually high.

Overall, this paper provides a new and valuable perspective on whether providing capital to and encouraging entrepreneurship among women may be a good path forward in pursuing poverty reduction and advancing development.  The authors conclude, however, on a dark note, noting that they find no impact of these one-time transfers on measures of women’s empowerment, as measured by questions about autonomy in economic decisions and the incidence of abuse.  This is in contrast to experimental evidence from the Philippines showing that access to savings accounts increases women’s empowerment.  There’s more to be said regarding the role of financial access in improving the lives of women in developing countries, and we look forward to contributing to this discussion with our mobile banking study in Bangladesh.

TopicsGenderRandomized Control Trials (RCTs)TrainingTransfers & SubsidiesUltra Poor and Women