【正文】 uded. First there is consensus amongst macroeconomists on the interpretation of the 197374 and the 197879 oil price increases as 39。supply shocks39。 to the global economy. Rasche and Tatom (1977) explored the possible adverse effects of high energy prices on economic growth by specifying a new type production function. This function includes an energy variable in addition to the conventional variables identified above. They were able to show with US data that not only had energy price increases exerted declining trends on potential GNP, but that actual output has been much closer to capacity than at any time in American economic history. Okun (1974, 1975), on the other hand, refuted this claim on the grounds that energy, pared with other production inputs, constitutes only a relatively 39。small cost share39。 in total output. This being the case, energy price changes will have a relatively small impact on the economy, especially taking into account substitution possibilities provided by other factors of production. Implicitly, perfect factor substitution has been assumed between energy, labour and capital. The size assumed for the substitution coefficient is an important consideration to this debate. It appears that the coefficient assumed by Okun (1974, 1975) is much higher than that adopted by RascheTatom (1977). Thus, an increase in energy prices will result in the substitution of labour for capital. Others have supported this view. Perry (1975, 1977) has suggested that 39。it is hard to believe that high energy prices can affect productivity and output growth39。 since it is just one of many production ponents. Evidently, the debate on the direct impact of high energy prices or other energy supply constraints on economic growth may have been exaggerated and 39。unwarranted39。. Im plicitly, there is a presupposition that existing theoretical economic approaches to productivity trends should deal adequately with the energy crisis. Given this assertion, it would seem the RascheTatom type production function is unnecessary and superfluous. Indeed, similar deductions have been made by Giersch and Wolter (I 983). They observed that this debate rests on how the OPEC 197374 and 197980 price increases are perceived. They contend that when this incidence is treated as domestic tax increases, it bees difficult to see why the economic impacts should necessarily differ from those of deflationary fiscal policies. While there is a consensus between Giersch and Wolter (1983) and Perry (1975, 1977) on the ability of conventional economic theories to explain productivity trends, they disagreed on the economic impact of energy prices. While Perry (1975, 1977) argued for a minimal impact, Giersch and Woiter (1983) believed the impact to be 39。indirect39。, whose size or magnitude depended on the effectiveness of government policies. Evidently, these results have been influenced by the methods of analysis employed as shown in Table !. Bemdt and Wood (1975) argued that while energy and labour may be substitutable, the plementary relationship between energy and capital raises its importance far more than its cost share. Implicitly, the dynamic impact of this relationship for output and productivity underline the important influence that energy has on economic growth. Hudson and Jorgenson (1974) similarly echoed this view when they found that the output position of the USA did change with energy price movements. In Table 1, it is evident that energy and labour, on the one hand and on the other, capital and labour are substitutable as argued by Okun (1974, 1975). Thus, it is possible to pensate for an energy induced decline in national ine by substituting labour or capital. However, there is no unanimous agreement on the relationship between energy and capital。 rather a divergence of findings are reported. Berndt and Wood (1975), Hudson and Jorgenson (1975) and Matsui et al, 1978 (as cited in Susuki and Takenaka (1981)) found energy and capital to be ple mentary while Griffin (1979) found energy and capital sub stitutable. However, looking at the type of data used (time series versus pooled data), the differences in results can be explained. Usually, time series analysis infer shortrun relationships and pooled data point to longrun situations. It can be argued therefore that energy and capital are plementary in the short run but bee substitutable in the long run. Developing countries In the case of the developing countries, a priori observation may suggest that the analysis of Okun, (1974, 1975) and that of(Perry 1975, 1977) apply, given the scale and size of the 39。modern39。 sector in total productive activities。 the substitution possibilities between labour which is 39。surplus39。, and technology that is intermediate and rudimentary. On a critical note, these characteristics are likely to intensify energy consumption in the event of these countries experiencing growth in economic activities. First, the perfect substitution assumed between labour and energy is doubtful. Even if we were to make an extreme assumption that the opposite is the case, in the absence of any defined optimum limit to profit, the demand for higher wages and better working conditions will hinder any substitution possibilities. Second, the rarity of skilled labour in underdeveloped economies cast serious doubts on any substitution possibilities. Given such supply constraints, capitallabour substitution bees inevitable. Furthermore, if the nature of technologies employed are considered, the fact that these are largely intermediate technologies that are energy intensive highlights the energy intensity of their use in production activities. Thus, it is not just the doubtful substitution possibility between labour and capital that is the problem here, but rather that the energy implication of the absence of such a possibility demands some attention. It is this fact that underlines the plemen