Neoclassical Growth Theory: Technical Change as “Manna from Heaven”

In the neoclassical growth model, growth results from increasing amounts of employ­ment of factors of production and from increasing productivity of these factors. The approach relies on a distinction between (1) a movement along and (2) a shift of the postulated macro production function:

where total output Q is a function of the capital stock K, total employment L and time t; see, in particular, Solow (1956, 1957).

Solow assumed technical change to be Hicks neutral. In this case the production function can be written as:

with A(t) measuring the overall effect of shifts over time. Total output growth splits into employment growth of input factors and the rate of technical change. The latter is computed as a residual contribution to growth not explained by the growth of inputs. This so-called Solow residual Al A is derived from the fundamental equation of “growth accounting”:

Assuming a Cobb-Douglas production function Q = A(t)K^aL1 ^{- α} with constant returns to scale and production factors paid according to their marginal productivities in con­ditions of perfect competition, w_K is equal to the share of profits and w_L is equal to the share of wages. Given the data on the growth rates of output, capital and labour and the income shares, the contribution of technical change to growth can be estimated.

Empirical studies by Solow (1957) and others showed the overwhelming importance of “technical progress” for growth (see Felipe 1999), but how is one to interpret the Solow residual? According to Solow, the shift factor A(t) - dubbed total factor productivity by John Kendrick (1956) - is a catch-all term for the unexplained and includes not only the effect of new technology, but also economies of scale, skill improvements, better health and so on. Moses Abramovitz (1956: 11) therefore spoke aptly of “some sort of measure of our ignorance about the causes of economic growth”.

Embodied technical change is explored in the so-called vintage capital approach, in which “capital” cannot be treated as a homogenous entity, since new machines are typically better than old ones. In contrast to disembodied technical change, which evenly affects the whole stock of capital, here the age-profile of the collection of different kinds of pieces of capital determines productivity. As a change in the age-profile of the capital stock is pivotal to increases in productivity, (gross) investment is the key vehicle to foster technical change. The idea that investment carries new technologies into the system has been explored inter alia in works of Leif Johansen, Robert E. Solow, Wilfred E.G. Salter and a paper by Nicholas Kaldor and James A. Mirrlees in the 1960s. Compared with the Solow model, “now the manna of technical progress falls only on the latest machines” (Hahn and Matthews 1964: 837-8).