“A scientific idea is ‘not even wrong’ if it is so incomplete that it cannot be used to make predictions that could be compared to observations to see if the idea is wrong” Peter Woit, (2006).
In spite of its widespread use, the concept of the aggregate production function has long been subject to serious theoretical and empirical problems. Even if well-behaved micro-production functions exist, there are well-established formal proofs that they cannot be summed to give aggregate production functions. The standard justification for the continued use of the aggregate production function is that they give good statistical fits to the data. However, the production function is a technological relationship that should be expressed using physical measures, and not estimated using constant price value data. Yet value data are always used. This is not innocuous as it can be shown that all that statistical estimates of aggregate production functions are picking up are a transformation of an underlying accounting identity. As such, the results have no implications at all about the underlying conditions of production and the measurement of technological change. Herbert Simon thought this result so important that he mentioned it in his Nobel Prize lecture. However, this problem is still almost totally ignored in the literature.
This seminar presents an overview of these issues and assesses their implications for modern macroeconomics.
John McCombie is Professor of Regional and Applied Economics in the Department of Land Economy at the University of Cambridge and Fellow in Economics at Downing College. He is Director of the Cambridge Centre for Economic and Public Policy. He was previously a member of the Department of Economics at the University of Hull and at the University of Melbourne, Australia. The culmination of his work in this area, with a colleague, may be found in Felipe, J. and McCombie, J.S.L. (2013), The Aggregate Production Function and Technical Change: “Not Even Wrong”, published by Edward Elgar.