What Stopped the Inflation? Unemployment or Commodity Prices?

Abstract

The main object of this paper is to argue that most of the deceleration of inflation in the OECD countries in general (including Britain) that took place between i980 and I982 should be attributed not to the direct impact of higher unemployment on the labour market but to the fall in 'commodity' (i.e. primary product) prices from I980 to I982, following their very sharp rise (accompanying the second 'oil shock') between I978 and I980. This led to a dramatic turn-round of about 50-60 % in the prices of primary products imported into the OECD countries from the rest of the world (and from other OECD countries), which, in turn, can account for the deceleration of about 20-25 % in the import prices for the average industrialised country. And our econometric estimates show that the pace of wage inflation is closely correlated with import prices. By contrast, our estimates for a variety of model specifications and choice of data do not show statistically significant correlations between aggregate unemployment and wage inflation.' We have not attempted to establish statistically that the turn-round in commodity prices was caused mainly by the strength and duration of the postI979 recession in the industrialised countries as a whole, but like most commentators we assume that this is the case. Given this assumption, therefore, our results imply that the recession in the industrialised world did not as is widely believed slow down the inflation by acting directly on the labour market (e.g. by crushing the unions or bringing a 'new breath of realism' into wage negotiations). Instead it worked by inducing a collapse of commodity prices (or, in the case of oil, inducing a slight fall by comparison with the preceding sharp rise). Insofar as our results are valid there are various obvious implications for some central topical economic debates. First, if there is no stable direct relationship between unemployment within any country and its own rate of wage inflation then there would be no 'NAIRU' and no clearly identifiable Phillips curve as some other investigators have found. Hence, debates as to whether and how far individual economies have unemployment rates above or below their NAIRUs and so on are irrelevant. In the Conclusions we summarise what seem to be the main policy implications for individual countries. But, secondly, insofar as levels of demand in the industrialised countries taken as a whole influence commodity prices which feed back on wage inflation via