Show Me the Data: Self-Employment, AKA Microenterprise
Economic growth practitioners are familiar with the dilemma. We want to provide youth, women, rural residents, or some other group with remunerative livelihoods that will outlast our three-year programs. If we invested the time and resources to ask them what they wanted, they would say one word, “jobs.” However, the sleepy job market in their country, province, or municipality is not going to pick up for the foreseeable future. What’s an aid worker to do?
A popular solution is to help people help themselves — and “buy” them a job running their own business. We call these microenterprises, or self-employment, because only the client and sometimes one or two family members work on the particular business.
The default ways of supporting these ventures are to provide capital and business training. The question is: Does this always work? As one of my fellow economists likes to quip, “In God we trust. All others must show their data.”
So let’s look at the evidence that’s out there.
It seems impossible to mention microenterprise without talking about microcredit, typically group loans to women. David Roodman's “An Impertinent Inquiry” concluded that such microcredit has no net effect on poverty, and it potentially harms women's empowerment. More recent work by Innovation for Poverty Action and the Abdul Latif Jameel Poverty Action Lab researchers shows similar results: Traditional microcredit increases investment in and the start-up of microenterprise but has little to no impact on household income, children’s schooling, or women’s empowerment.
Generally, women benefit less than men in microenterprise/self-employment, according to this World Bank systematic review. Part of it may be that women’s business constraints and options differ from men’s. Rohini Pande has found that, when given a two-month grace period before repayment starts, women divert the capital to their husband (or son)'s enterprise, typically in a higher-risk, higher-return line of business than theirs. In Pande’s India study, this grace period increased incomes particularly for households with a “male” enterprise. This study also suggests that tweaking the credit product may well yield different results.
Part of the trouble is that business training varies wildly in content, format, and quality, though it usually covers financial management, human resources, and business plan development. Generally, research shows that business training by itself may change the behavior of women-owned microenterprises, but it usually has no net impact on firm performance (sales, profits, number of employees). Like microcredit, business training does benefit male entrepreneurs more than female ones. See, for example, this overview, which I peer-reviewed last year, and this World Bank review on business training.
The bottom line is similar for entrepreneurship by youth and the poor: Business training may increase their self-employment rates, but not their income. This was even true for college graduates in Tunisia, where there was no impact on overall employment rates, only a slight increase in self-employment.
But wait — context also changes what works.
Unlike mainstream urban microcredit, rural microcredit has increased incomes in Morocco and Mongolia, according to the World Bank. But this review presents more of a mixed picture for microcredit in Sub-Saharan Africa, mostly from evaluations of rural programs. A Pakistan study found that access to larger loans (USD 1,700, still micro in some countries) had no impact on rural borrowers.
Rural Business Training
Here, the impact also varies. A study of a two-day business training in rural Mexico found positive impact (higher revenue and sales) and changed behavior (more accounting, registration with government). The World Bank Pakistan study just cited finds mostly improved behaviors and increased household expenditures among rural male borrowers, but not much improvement for rural female ones.
Beyond Microcredit and Business Training
This is not to say, by any means, that donors should abandon microenterprise. In fact, USAID will continue to work towards poverty alleviation through a variety of approaches, including supporting microenterprise. The evidence suggests that we can and should try other solutions besides traditional microcredit and business training to facilitate livelihoods for the economically active poor. These impact evaluations are extremely valuable to our work and in shaping future programming. USAID is starting to incorporate this evidence into our project designs.
What else can we try to support self-employment and microenterprise? Based on this USAID-commissioned evidence gap map, especially the cell for employment assistance/income and consumption, one could look into productive asset support, market-driven technical skills training, and perhaps savings and conditional cash transfers for vocational training in conflict-affected settings. One could certainly try market linkages—we just don’t know their impact yet.
Ultimately, research shows that net benefit and sustained impact on both behavior and income in microenterprise programs don't come from business or financial literacy training, wage subsidies, or guaranteed employment. Also, tailored technical assistance and consulting for firms are only effective at the SME level.
There are probably dozens of other interventions that you, dear reader looking for new ideas, could try. Please create them, pair each one with an impact evaluation after you beta test it, and show us the data.