Measurement of capital

The marginal productivity theory of wages has never lacked critics at any stage in its history but, at least until recently, the marginal productivity theory of interest was allowed to pass more or less unscathed. In the 1950s, however, Joan Robinson, soon followed by a number of other Cambridge economists (Cambridge, United Kingdom, that is), launched an entirely new attack on the so-called marginal productivity theory of distribution, directed in particular at the Hicksian two-inputs-one-output simplification of the neoclassical theory of factor pricing. The stock of capital in an economy, it was argued, being a collection of heterogeneous machines rather than a homogeneous fund of purchasing power, cannot be valued in its own technical units, although apparently “labor” and “land” can be so measured; the valuation of capital necessarily presupposes a particular rate of interest, and this means that the rate of interest cannot be determined by the marginal product of capital without reasoning in a circle; hence, marginal productivity theory cannot explain how the rate of interest is determined.

Much of this criticism falls to the ground if we replace the simpliste formulation of marginal productivity theory by the disaggregated Walrasian version, which neither invokes nor implies the concept of aggregate production function, nor indeed the notion of the aggregate capital stock as an economic variable.