Accounting for Productivity Growth When Technical Change is Biased

Abstract

Solow (1957) decomposed labor productivity growth into two components that are independent under Hicks neutrality: input growth and the residual, representing technical change. However, when technical change is Hicks biased, input growth is no longer independent of technical change, leading to ambiguous interpretation. Using Solow's model, I decompose output per worker into globally independent sources and show that technical bias directly contributes to labor productivity growth above what is captured in the Solow residual. This contribution is sometimes large, generating rates of total technical change that substantially exceed the Solow residual, prompting a reinterpretation of some well-known studies.

Keywords: technical change, productivity growth, technical bias, innovation

JEL Classification: O33, O47, N61