How important is human capital misallocation across sectors for aggregate productivity differences? Evidence beyond Cobb–Douglas

Abstract

Misallocation of workers and their human capital across sectors has negative implications for productivity. The literature examining this type of labor misallocation has mostly assumed a Cobb–Douglas production function, which is at odds with important macroeconomic frameworks and empirical evidence. Our paper hence suggests a new twist to revisit the question: to what extent does human capital misallocation across sectors explain why some countries are so much richer than others? Our innovation consists of incorporating a more flexible and empirically plausible CES production structure with different labor skill types into the misallocation literature. The results from this analysis indicate that human capital misallocation can explain approximately 15% of output per worker variation across countries, which is about one fourth less than under the conventional Cobb–Douglas specification (21%). This suggests that labor market efficiency may be less of a problem for aggregate productivity differences than previously thought.