Testing for Credit Constraints in Entrepreneurship

Common tests for credit constraints are highly sensitive to econometric assumptions. Tests of the relationship between wealth and entry into entrepreneurship require strong assumptions about the distribution of unobserved ability, and the results change when the definition of entrepreneurship changes. Tests of the relationship between entrepreneurs’ productivity and wealth can produce spurious evidence of credit constraints if heterogeneity in technology is ignored. Using data from Nicaragua, I illustrate these problems, show how to make the productivity test more robust, and demonstrate a semiparametric method for measuring the implicit interest rates faced by entrepreneurs at different wealth levels. As many as 80% of Nicaraguan entrepreneurs face financial constraints. The implicit interest rate for high-wealth entrepreneurs is at least 35% below the rate at the median wealth level. ∗Department of Economics, University of Chicago, [email protected]. I thank Robert Townsend and Sergio Urzua for helpful comments.