Plenary: Entrepreneurship, firms and growth in developing countries

Alan Hirsch (IGC) opened the first plenary session of Growth Week 2012 and introduced Chris Woodruff (University of Warwick and IGC) who presented a concise and thorough summary of lessons learned about microenterprise dynamics from randomized control trials. Woodruff organized his survey of the literature using a standard production function, and drew out lessons from a diverse set of interventions targeting capital (K) and labour (L) constraints, as well as those designed to improve total factor productivity (A).

Citing work in Sri Lanka, Mexico and Ghana, Woodruff explained that there is reasonably robust evidence that returns to marginal investments of capital are very high on average in microenterprises, though with a great deal of heterogeneity across firms within samples. Interestingly, these returns are not accompanied by employment growth. While acknowledging that work investigating the impact of microfinance generally finds only insignificant returns, Woodruff questioned whether these studies had enough statistical power to capture an impact. There is not much as much work, in comparison, on the impact of labour constraints on of microenterprise growth, but a study in Sri Lanka found that only a small percentage of firms respond to temporary incentives to hire workers. This study did find some evidence that temporary subsidies may spur further employment growth in a small percentage of firms. Finally, Woodruff highlighted the large amount spent on microenterprise training programs around the world, and particularly in low-income countries, despite mixed evidence of their impact. On average, training appears to have modest effects on microenterprise growth, though Woodruff again noted that in general these studies are under powered. Small sample sizes and heterogeneous samples make it hard to detect significant effects. There is a trade-off, however, as while a more homogenous sample may make it easier to measure a statistically significant effect, it also may reduce the external validity of the study. Finally, Woodruff discussed studies on the impact of formality and highlighted that at the upper tail of the distribution, formality has a large effect on a small number of firms, though this is not observed at the median.

Woodruff’s presentation was followed by a lively question and answer round, during which Woodruff acknowledged that, in many cases, relaxing constraints may not have a large impact on dynamic growth of microenterprises simply because a large portion of firms may not have aspirations grow. However, Woodruff suggested that many micro-entrepreneurs are likely looking to increase their income even if only by a small amount, and questioned why firms weren’t making decisions which would facilitate this. He noted that in the developing the world, the most significant division is between firms which are reliant on own or family labour, versus those that make the jump to hiring non-family labour. Woodruff responded to another question regarding sizes of firms in different sectors and emphasized that while some sectors may be more suited to only large scale enterprises in others, such as retail, a continuous distribution of firm size is observed. When questioned on the implications of this portfolio of work for poverty, Woodruff pointed to the marginal and zero-equivalent effects of capital shocks on enterprises ranked at the bottom of the profit distribution at baseline. He suggested that for these micro enterprises perhaps injections of capital is not enough, and must be matched by investments in human capital as well.