The SME Initiative and What We Don't Know

Last week, Innovations for Poverty Action’s SME Initiative brought together researchers and practitioners to discuss recent research on SMEs (Small and Medium Enterprises), mostly in the developing world.

Why the growing interest in SMEs? Partly it’s a reaction to the murky results on the impacts of microfinance. Evidence is increasingly suggesting that microenterprises do not tend to grow much and their impacts on income and consumption are up for debate. Against that, supporting SMEs may be a more effective way to provide jobs and reduce poverty.

Given that, what struck me most about the conference is not what was discussed, but what was not discussed. Many big-picture questions that underlie the focus on SMEs were not explicitly raised:

•    What jobs are created by SMEs, and for whom? How do SME support policies and program change these characteristics? Generating employment is a key impact, but the poverty-reduction power of SMEs depends on who is being employed by SMEs and who is being hired as SMEs grow with the support that is provided to them: does the marginal employee live in a poor household? Would s/he otherwise be self-employed in a microenterprise and receive microfinance services?

•    While the –necessary and excellent– research presented at the conference suggests that supporting SMEs produces positive impacts, the ultimate question is about the cost-effectiveness of supporting SMEs. Is supporting SMEs more cost-effective than other interventions (microfinance, cash transfer programs, education, health, etc.) to generate various outcomes/impacts?

These questions are critically important for donors and policymakers deciding how to invest, grant funds, or define policy frameworks with the objective of having the largest impact on poverty. Jonathan Morduch and I recently finished a paper that uses non-experimental data on Bangladeshi SMEs and microcredit borrowers to start getting at these kinds of questions. Answering these questions in a more definitive fashion will take time, and rigorous impact studies such as those presented at the conference will help.
Some highpoints from the conference: 

  • MIT’s Antoinette Schoar highlighted that it’s a misconception that there’s a continuum between small subsistence enterprises (self-employed individuals with no employees; typically the target of microfinance ) and “transformational enterprises” (small – but growing – firms that have made the jump to having non-household employees). While many transformational firms started off as microenterprises, most subsistence enterprises will not in fact turn into transformational SMEs. 
  • Many papers at the conference highlighted that SMEs face different constraints from microenterprises. Unlike microfinance- or microenterprise-focused research, the majority of the research presented at the conference investigated human capital constraints. Researchers, for example, partnered with business consultants to help entrepreneurs improve their skills and practices. These skill-building and training interventions work in some firms: sales, profit, product quality and (to a lesser extent) employment increase as a result of the training. But overall, the evidence remains fragmented. There’s not yet a clear picture of what works, for which firms and under what conditions. As Yale’s Dean Karlan highlighted in the context of human capital constraints, we will probably benefit more from a stronger conceptual framework than from specific impact studies at this point. I agree that more big thinking is needed, and not just about human capital constraints.