The central feature of urban economies in low-income countries is the prevalence of small-scale businesses. Between a quarter and a third of workers are typically self-employed, most often working alone or hiring one or just a few employees. Economic growth depends on job creation by private sector enterprises. This project addresses the barriers to growth among the very large number of very small firms in a low-income country. Scientific evidence on the nature of the barriers is critical to developing policies to stimulate growth.
Recent evidence from the United States and Europe indicates that only around one percent of non-employers create jobs (for others)in any given year. Data from a panel of 618 household enterprises gathered by the PI and collaborators suggests similar transition rates from non-employer to employer status in Sri Lanka. Between June 2005 and September 2007, the share of employers in the sample increased by 3 percentage points. However, because the number of non-employers is very large, the non-employers account for a substantial share of new job creation even though only a small percentage of them grow.
Identifying the constraints to growth is challenging. In the absence of constraints, firms choose the level of capital, wage labor and manager training so that their marginal product equals the marginal cost. For example, firms will increase their capital stock until the return to capital equals the real interest rate (with risk premium) faced by the firm. Therefore the gap between marginal returns to capital and the market interest rate provides a measure of the extent to which capital constraints bind, preventing growth and perhaps job creation. However, the measurement of the returns to business inputs is complicated by the presence of unobserved characteristics of the owner and firm which affect both input use and output and profitability. For example, more able managers may invest in more capital.
The project identifies the returns to capital, wage labor, and management training by randomly altering the price of capital, labor, and business training faced by firms. Altering the price provides exogenous variation at the firm level in the cost of different inputs, enabling identification of the returns to different factors of production, and measurement of the extent to which relaxation of constraints affects the growth of micro-enterprises. The project will provide a detailed and scientific assessment of programs to identify and stimulate high-growth micro-enterprises. Faced with limited budgets for programs aimed at job creation, governments and aid agencies need to focus efforts on those programs and enterprises that are most effective. The project is designed to provide guidance on this issue.
Between one-quarter and one-half of the labor force in most low-income countries is self employed. (By comparison, around 10-15% of the U.S. and European labor forces are self employed.) Most of these enterprises provide goods and services for local markets, many of which are purchased by the poor. Available information indicates that most of these enterprises never hire paid employees. This project examined constraints to expansion of microenterprises in one low-income country, Sri Lanka. We surveyed a sample of just over 1,500 male-owned enterprises with at most 2 employees in urban areas over a period of 3 years. Some of the enterprise owners were randomly selected to receive small business training in the form of a 2-week course based on one developed by the International Labor Organization. Others were randomly selected to receive either a temporary wage subsidy or a matched savings program which encouraged them to accumulate capital to invest in the enterprise. (NSF funds were not used in the latter two interventions.) We find that the savings incentives and the wage subsidies were effective in generating additional employment, while the entrepreneurship training was not. Approximately two years after the training and savings programs were completed, and a year after the wage subsidies ended, about three in ten enterprises offered the temporary wage subsidies employed at least one worker, compared with two in ten of those not offered any of the incentives. Thus, the temporary subsidies generated additional employment in about one in ten of the enterprises, even after the subsidies were removed. The savings program was only slightly less effective in generating employment – with additional employment generated in about one in twelve enterprises offered the program. One interpretation of these results is that the skill of managing employees is not easily taught, but can be learned from experience. Because the changes in capital (from the savings program) and labor (from the wage subsidies and savings programs) were generated in a random subsample of the enterprises, they allow us to make a very accurate estimation of the production function of the microenterprises. This will allow us to have a much better understanding of, for example, the effect of microfinance on income and employment generation. This analysis is ongoing.
