Cobb–Douglas's last fling?

The use of the malleability assumption and a simple marginal productivity theory of distribution underlies the early post-war work on aggregate production functions: the attempts to sort out from actual statistics the increases in output per man that are due to technical progress, i.e. shifts of an aggregate production function, from those which are due to capital deepening, i.e. movements along a given production function. In this chapter we use Solow's famous 1957 article to illustrate the approach.

Solow assumed a coiistant-returns-to-scale aggregate production function, static expectations and competitive conditions. It followed that paying factors their marginal products exhausted the total product, which consisted of a Clark–Ramsey one all-purpose commodity, see J. B. Clark [1889], Ramsey [1928]. (Capital may then be measured in the same units as output, remembering that one is a stock, the other a flow, see Solow [1956a], p. 101.) Solow did not specify the form of the production function until after he made the empirical fittings when Cobb–Douglas gave the best fit.

Technical progress was assumed to be neutral and completely disembodied, i.e. left all factors unaffected, so that marginal rates of substitution between factors at given factor ratios were unchanged, though, at each ratio, there was a mystical rise of the same proportion in the total output associated with each ratio. All capital goods were treated alike, whether they were newly created and incorporated the latest advances in technical knowledge (and the effects of the pull of expected factor prices) or whether they were fossils inherited from the past, previous years' investments which in fact could be expected to reflect the then prevailing technical conditions, expectations and relative factor prices.