Oscillation et croissance

Abstract

Oscillation and growth. The world economy exhibits both a trend and oscillations. These phenomena have traditionally been dissociated and studied separately. The trend-oscillation dichotomy has given rise to an exogenous-endogenous dichotomy, a long-term — short-term dichotomy and a strudural-conjundural dichotomy. The cycle dissociated from its trend is, to use Kaldor's expression, a static cycle. It derives from physics and has given rise to many models with either an implicit oscillator (Aftalion, Hawtrey) or an explicit one (Kaldor, Samuelson, Metzler). All show that a trendless cycle is possible. The trend without cycles is the older approach : Smith, Ricardo, Malthus considered secular movements fundamental, oscillations negligible. The trend may, first, merely be expressed as an annual rale of growth (Harrod’s growth coefficient) based on observable data. It is, secondly, a phenomenon to be explained : growth is governed by increasing population and productivity, and institutional rigidities. The 'Kondratieff wave is however a problem case in the trend-oscillation dichotomy : is it due to exogenous (or structural ) causes, or endogenous (or confunclural) ones ? The dissociation of trend and cycle is, in fact, dangerous ; they must be studied jointly. Two approaches are possible. According to the first, growth is a succession of open cycles, i. e. is generated by cyclical forces. According to the second, cycles are a by-product of growth (Harrod's and Domar’s view) ; it is well-known that a change in the rale of growth of one series can change the direction of another : a crisis is thus equated to a change in the rale of growth. (An intermediate case is Hicks's model : cycles are sandwiched between a rising floor and ceiling. But these limits, as Goodwin has shown, remain unexplained). The first approach (cycles determining growth) is by far the most frequently adopted. The second approach should now be developed. This would entail, in particular, the study of economic structures — a vague term which needs expliciting — the importance of which has been stressed by Akerman.