Capital-output ratios and their behaviour

Abstract

The purpose of the present investigation is to examine empirically whether or not the capital-output ratio is a technological coefficient and a constant, within the basic framework of the macro-dynamic model of Kaldor and the micro-dynamic model of Barrre. The empirical investigation is confined to a cross-section analysis of aggregate ratio with reference to the ECAFE countries, covering the period 1950, and a time-series analysis with reference to eighteen manufacturing industries in India over the period 1950 to 1965, using linear regression technique. In the aggregate study pertaining to ECAFE countries, an attempt was made to examine the role of investment in economic development whether there is a unique (proportional) relationship between investment and growth rate and to throw light on the determinants of investment-output relationship such as the composition of investment by type of assets, sectoral distribution of investment, sectoral growth of output, and sectoral contribution to domestic product. The main findings of this investigation are:

There is no unique relationship between growth rate and investment ratio. Growth rate is positively related with investment in machinery and equipment and negatively with investment in buildings; therefore, it is determined by the composition of investment by type of assets. There is a close relationship between growth rate and sectoral distribution of investment; growth rate varies positively with increase in investment in secondary and tertiary sectors and negatively with increase in investment in the primary sector. There is a close relationship between overall growth rate and sectoral performance; again, growth rate is positively related with growth in secondary and tertiary sectors and increase in the contribution of these sectors to domestic product, and negatively with that of the primary sector.

By way of further exploration, a disaggregated study with particular reference to India was made with a view to fill in the empirical lacunae梑y testing a few theories which have not been put to empirical verification like Kaldor's model, which tries to justify constancy of capital coefficient assumption, and Barrere model which explains capital intensity as being determined by economic factors, namely, disposable funds, as against technological factors. The time-series study pertaining to select industries in India examines the nature of the behaviour of average and incremental capital-output ratios and the relationship between the behaviour of average capital-output ratio and investment behaviour. The main findings of this investigation are:

The ratio, whether defined as average or incremental, has not remained constant in any of the industries, as reflected in annual and period-wise comparison of the ratios, industry-wise. However, a tendency towards stability (that is, the tendency for the average coefficient to reach a certain magnitude in spite of intermediate fluctuations) has been observed in twelve out of eighteen industries. While in some industries, the ratio has moved in an upward direction, in others, a declining trend has been observed. In those industries where tendency towards stability has been observed, R in the case of investment function is 0.40 and above in the case of five industries; one of the regression coefficients of the function has the expected sign in such industries, though not significant in all the cases, the only exception being Cement; also, the Technical Progress Function curve has taken the assumed shape in these industries except in two industries pottery, China and Earthenware, and Rubber and Rubber Manufactures. The regression results relating capital-output ratio and profits after depreciation and tax have shown that R is 0.40 and above in only seven out of thirty-six cases. Introduction of lags has resulted in an improvement in ten out of eighteen cases. When capital-output ratio and profits are related with one-year lag, the regression coefficient has a negative sign in fourteen out of eighteen industries and is significant in five cases. Relationship assuming two-year lag has changed this significantly; in eleven out of eighteen industries, the relationship is positive although the regression coefficient is not significant in most of the cases. In short, if we take the sign of the regression coefficient as the satisfactory criterion for testing the hypothesis, Barrere's hypothesis is supported by about 40% of the observations.

These findings have a bearing on the assumptions involved in the use of a errtatic ratio. As pointed out in the introductory chapter, the basic assumptions on which the erratatic ratio (as inspired by Harrod-Domar model) rests are: (i) the capital-output ratio is a technological coefficient and a constant; (ii) that there is a proportional relationship between capital and output梐ssumptions on which predictions are based; and (iii) investment is the factor in output generation. Our empirical investigation the aggregate study with reference to ECAFE countries questions the validity of the view held in the aggregative approach of Keynes, Harrod and Domar, namely, that the rate of growth is a function of the rate of investment and that there is a unique relationship between investment and growth rate. Further, our study has shown that aggregate investment-output relationship depends on the technical structure of investment and on sectoral performance. Our time-series study has denied the assumption field both in theory and in practice of constant coefficient. However, the investigation has shown that a tendency towards stability can be observed if Kaldor investment function and Technical Progress Function hold good; but in a good number of industries, it is these rigid conditions that are not satisfied, thus pointing to the weak foundation on which the constancy assumption rests. Regarding the question whether capital-output ratio is a technological coefficient, the influence of only one economic factor profits in the ratio has been studied. Only in a few industries has profit exerted influence on the capital coefficient and not in all, and the possibilities of both direct and inverse relationship have also been observed. However, on the basis of this restricted study, without analysing the influence of other economic, more specifically behavioural factors, it is not possible to pass a judgment on whether the capital coefficient is a purely technological coefficient or a behavioural coefficient, especially in view of the weak relationship observed. For the same reason, no attempt has been made to discuss the policy implications of the investigation.