Discussion

Erik Hurst followed up on a comment made by the discussant, John Haltiwanger. He posited that the increasing dispersion in revenue productivity across firms might arise because of differential shifts in the composition of the labor force based on human capital accumulation. Haltiwanger added that the source of dispersion in revenues per employee could also be the result of misallocation deriving from a learning process. The adoption of information and communications technologies across firms is a slow process, which would result in staggered revenue productivity gains across firms, he argued. The authors agreed that markup dispersion—themain source of productivity differentials according to their results—provides a good explanation, but it is not exhaustive. Other possible channels, like human capital, and the resulting misallocations are worth investigating, they said. The authors also sympathized with the discussant’s suggestion to extend their analysis to the sectoral level and to employ a nested CES production structure. They agreed that this methodologymight be useful to shed more light on the role of big firms in driving aggregate productivity growth. The discussion ended with a brief question by Haltiwanger on the surge of IPOs in the 1990s. He askedwhether the authors had researched their role in contributing to subsequent growth. The authors answered that they did not perform the exercise but that their framework could be employed for answering the question.