Working Paper: NBER ID: w23005
Authors: Suresh De Mel; David McKenzie; Christopher Woodruff
Abstract: The majority of enterprises in developing countries have no paid workers. Is this optimal, or the result of frictions in labor markets? We conduct an experiment providing wage subsidies to randomly chosen microenterprises in Sri Lanka. In the presence of frictions, a short-term subsidy could have a lasting impact on employment. We find the subsidy induced firms to hire, but there was no lasting impact on employment, profitability, or sales. Analysis rules out several theoretical mechanisms that could result in sub-optimally low employment. We conclude that labor market frictions are not the reason own-account workers do not become employers.
Keywords: wage subsidy; microenterprises; employment; labor market frictions; Sri Lanka
JEL Codes: C93; D22; L26; O12; O17
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| employment levels (J23) | profitability (L21) |
| employment levels (J23) | sales (M31) |
| wage subsidy (J38) | employment levels (J23) |
| wage subsidy (J38) | survival rates for firms with low capital and low profitability (D25) |
| wage subsidy (J38) | number of paid workers hired (J39) |
| wage subsidy (J38) | employment gap diminishing between treated and control firms (J68) |