We have seen in Part 1 of this book that one of MM’s strengths is its interpretable ‘steady state’; namely, a balanced growth path. The economics of balanced growth was worked out in the ‘fifties and ‘sixties in terms of single-sector growth models (Solow, 1956; Swan, 1956; Meade, 1961). One conundrum which had to be solved in this literature was how to treat technological change. It turns out that only relatively specialized treatments of technological change are simultaneously consistent with both the existence of a balanced growth path and with the stylized facts characterizing modern economies.
Unable to display preview. Download preview PDF.
- 1.The notation used in sections 5.2 and 5.3 is self-contained. Readers familiar with elementary ideas about technology in production economics may prefer to skip section 5.2.Google Scholar
- 3.Hicks-neutral and Harrod-neutral technical change are defined below.Google Scholar
- 4.There is an exceptional case (the Cobb-Douglas production function, (5.2.1)) in which Harrod neutrality and Hicks neutrality are equivalent — see Exercise 5.3 below. In the exceptional Cobb-Douglas case, the factor shares remain stationary over time irrespective of the time rates of improvement in the marginal productivities of the factors.Google Scholar
- 10.Thus we have established the third item listed under (ii) above, again conditional on the constancy of the capital-output ratio.Google Scholar
- 11.Of course, we have argued (somewhat loosely) from a closed-economy one sector model. Nevertheless, the results are valid in MM.Google Scholar
- 12.Because export demand schedules in MM slope downwards, the terms of trade vary endogenously. It follows that to keep K/Y fixed it is necessary to sterilize endogenous changes in the terms of trade by making the foreign demand curves shift outwards during the steady state at the domestic economy’s natural rate of growth γ.Google Scholar
- 13.This very low estimate of (3 was regarded as requiring further research (Murphy 1992a, p. II-26 (04/05/92). (In comparing this estimate with others, it should be kept in mind, however, that MM measures labour input in persons, not in person-hours.) Indeed, in more recent versions of MM, the estimated value of ß is higher, as it was in earlier versions of the model — 0.81 per cent per annum is quoted in Murphy (1992b, p. 190).Google Scholar