Endogeneity bias and other problems in the estimation of production functions

Dr. Jesus Felipe, Asian Development Bank

The possible endogeneity of labor and capital in production functions, and the consequent bias of the estimated elasticities, have been discussed and addressed in the literature in different ways since the 1940s. This paper revisits a very much neglected argument, first outlined in the 1950s, which sees the problem of biased estimates not as one of regressors’ endogeneity. This argument is that the series of output, employment and capital used in empirical exercises are linked through a distribution accounting identity. This identity can be rewritten as a form that resembles exactly a production function (Cobb-Douglas, CES, translog). This happens because the series used are not physical quantities (for output and capital) but value data. The argument has clear predictions about the size of the factor elasticities and about the bias of the estimated elasticities. To test these predictions, we estimate a typical Cobb-Douglas function using five estimators, including GMM and Levinsohn-Petrin, and show that: (i) the identity is responsible for the fact that the elasticities are must be the factor shares; and (ii) the bias of the estimated elasticities (i.e., departure from the factor shares) is, in reality, caused by the omission of a term in the identity. Estimation methods that theoretically deal with endogeneity, including the most recent ones, cannot solve this problem. We conclude that the use of value rather than physical data poses an insoluble problem for the estimation of production functions. This is, consequently, far more serious than any supposed endogeneity problems. 

About the speaker

Jesus Felipe is adviser to the chief economist in the economics and research department of the Asian Development Bank. He has held academic positions at the Hong Kong University of Science and Technology and the Georgia Institute of Technology, Atlanta, and is a research associate at the Cambridge Centre for Economic & Public Policy, University of Cambridge; the Centre of Full Employment and Equity, University of Newcastle; and the Centre for Applied Macroeconomic Analysis, Australian National University. He works on issues relating to long-run growth in Asia, productivity, and technological progress, and his research interests extend to areas such as growth (especially the debate on the sources of growth in East Asia), the functional distribution of income, business cycles, the path of profit rates, and structural change. He is currently leading a project on industrialization and structural change in Central Asia.

Felipe is coauthor and coeditor of Labor Markets in Asia: Issues and Perspectives (Palgrave Macmillan, 2006), named a noteworthy book in industrial relations and labor economics by Princeton University. He is the author of Inclusive Growth, Full Employment, and Structural Change: Implications and Policies for Developing Asia (Anthem Press, 2009), The Aggregate Production Function and the Measurement of Technical Change: A Critique and Evaluation (Edward Elgar, 2013), and Development and Modern Industrial Policy: Issues and Country Experiences (Edward Elgar, 2014). His work has been published in the Cambridge Journal of EconomicsJournal of Comparative EconomicsEastern Economic JournalJournal of Income DistributionInternational Review of Applied EconomicsJournal of Development StudiesWorld Development, and Oxford Development Studies. He also serves on the editorial board of the journal Metroeconomica and is managing editor of the Asian Development Review.

Venue: via Zoom (please contact us at seminars@merit.unu.edu for the Zoom link)

Date: 21 May 2021

Time: 11:00 - 12:00  CEST