Can Micro-Credit Bring Development?

We examine the long-run effects of micro-credit on development in an occupational choice model very similar to Banerjee and Newman (JPE, 1993). Micro-credit is modeled as a pure improvement in the credit market that opens up self-employment options to some agents who otherwise could only work for wages or subsist. Micro-credit can either raise or lower long-run GDP, since it can lower use of both subsistence and full-scale industrial technologies. It typically lowers long-run inequality and poverty, by making subsistence payoffs less widespread. A case exists, however, in which it both lowers output per capita and raises poverty in the long run. The key to micro-credit's long-run effects is found to be the "graduation rate": the rate at which the self-employed build up enough wealth to start full-scale firms. We distinguish between two avenues for graduation: "winner" graduation (due to supernormal returns) and "saver" graduation (due to accumulation of normal returns). We find that "winner" graduation, however high its rate, cannot bring long-run development. In contrast, if the saving rate and normal returns in self-employment are jointly high enough, then micro-credit can bring an economy from stagnation to full development via "saver" graduation. The lasting effects of micro-credit may thus partially depend on simultaneous facilitation of micro-saving.