Performance variation across industries: role of ownership, trade and market structure

Abstract

Performance of industry has been of concern to academicians as well as to practitioners and politicians. Industrial growth has been considered essential to improve the standard of living through higher productivity of Industry. It has been argued that industry in India has failed to perform because of poor public sector performance, poor export earnings. Low productivity and high cost. A natural question is: what has gone wrong? Is it the presence of the public sector, government policy, protection from import competition, lack of competition in the market? In most cases, a measure of performance is high profitability. Why are some industries more profitable than others? Are profitable industries also efficient? The questions addressed by the study are: Are there any variations in the performance of various Industries? Are these significant? What has happened to these variations over time? What are the factors which determine inter—industrial variations in performance? What is the relative role of ownership, international trade variables and market structure in explaining differences in performance? Three different performance indicators have been employed —physical, financial and economic. A number of measures for each of the performance indicators have been used. For example, growth in employment and output were tried for physical performance, price—cost margins (POM) and return on capital employed (ROCE) for financial performance, and partial and total factor productivity (TFP) for economic performance. A time series of cross section of 42 large manufacturing industries in India was selected to study inter-industry differences in performance. This is exhaustive list of all industries which produce no less than 0.51 of total output of manufacturing sector. A series of parametric and non-parametric tests were used to test hypotheses about inter-temporal and inter-sectoral differences in performance. A set of hypotheses pertaining to the likely determinants of inter-industry differences were formulated. Theoretical possibility of superior private sector performance was examined. The impact of protection on industrial performance was also studied, along with other structural variables of the market. Single and simultaneous equations model were formulated to test hypotheses about determinants of performance. The empirical work was carried out for two specific periods, one for seventies (1974-78) and the other for eighties (1962-66). Averages were taken to mitigate any biases due to choice of an year. Besides the cross section study of two periods, time series analysis for total industry for the period 1973-86 was also carried out. A large part of the data required for the study was not available in the desired form. Thus, data on Capital Series, Effective Protection Rates, Concentration Ratio, Exports and Wholesale price indices was thus constructed for a meaningful study. The findings of the study are summarized below. There are considerable inter—industry differences in performance for the two periods (1974—76 and 1982-86). It is found that the mean profitability was lower in 1960s than 1970s, though the reverse was found to be true for relative variations in it. Relative contributions of different factors to TFP have undergone significant changes between the two periods. For the manufacturing industry, growth in TFP, though positive, is negligible (0.3% p.a.). Non-parametric tests suggest that labour productivity is higher for 1980s than 1970s. However, there are no significant differences in other productivity measures between the two periods. Profitability has declined over the decade. The conclusion is valid in terms of both PCM and ROCE. Growth in physical indicators has been higher in 70s as compared to 80s. With regard to the growth in TFP, the industries which have shown better performance are; vanaspati oils, fertilizers and pesticides, grain mill products and cement during 1970s. And petroleum refineries, paper, grain mill products and turpentine resins, etc. during 80s (Divisia Index). The industries which have shown poor performance are; paper, tanning and finishing of leather, and tyres during the 70s, and other chemical products, other edible oils, and cement during the 80s. There is no consistency among indices over the best and the worst performers. Labour productivity, as measured by output per worker or output per employee, was higher in fertilizers, petroleum, drugs. Turpentine and resins and cosmetics, and lower in cotton ginning, grain mill products and ready—made garments for the period 1913-86. Output per rupee of wages was highest in fertilizers and lowest in railway wagons during this period. Similarly, managerial productivity was highest in other textiles and lowest in dairy products during this period. Employee productivity improved during the 80s over 70s in petroleum, tea, paper and deteriorated in coal tar, electric motors etc. The industries which reveal above average profitability are paper and organic and inorganic chemicals during the 70s, and cement and electric motors and generators during the 80s. Grain mill products, other edible oils, and coal tar performed poorly during 70s while jute textiles performed poorly during 80s. Return on capital employed (ROCE1) was found to be higher in tea and cosmetics for both 70s and 8Os, and chemical products for 80s only. Regression results suggest that PCMs are significantly related to concentration for both the periods. Ownership, growth and protection do not appear as significant determinants of PCM. Two variables which have significant impact on PCM are export intensity (—vely related) and skill levels (+vely related). There is evidence of an inverted U effect of concentration on margins for both the periods. ROCE was influenced by economies of scale and relative price movements in 1970s. During 80s, besides export intensity and skill levels, wage rates also exerted a powerful influence on ROCE. Concentration and protection do not appear to be significant. Ownership has affected ROCE negatively for both the periods. Labour productivity, as measured by output per employee is mainly determined by capital intensity, wage rate and skill levels. Concentration, exports and protection do not appear as significant. Ownership enters the equation with a negative sign and is also significant. In line with other studies, this study indicates that growth rate of output is a significant determinant of growth in labour and total factor productivities. Ownership affected growth in TFP positively, indicating improvement in efficiency. PCM measures are not significantly different across the public and private sectors, though ROCE is significantly higher for the private sector. Output per rupee of wages and output per rupee of ‘salaries paid are relatively low in public sector. A comparison of growth in four measures of labour productivity suggests that barring managerial productivity, there are no significant differences between the two sectors. Capital and fuel productivities are relatively higher in private and public sector, respectively. There are no significant differences in materials productivity. Growth of capital, fuel and materials productivity is not significantly different across the sectors. Growth of TFP is higher in 19 industries in the public sector, while it is the reverse in 12 industries; in the remaining 11 industries the comparison was not permitted by data considerations. Non-parametric tests suggest no significant differences between the two sectors. To summarize, the estimated function explaining the inter-industry performance variation was found to differ in the two periods, and the differences were not only in terms of significance but also in terms of the direction (sign) of the relationships. The assertion that performance of industry has improved in terms of efficiency is not vindicated. Market structure (concentration) has positive and significant effect on performance. The impact of state ownership on performance variables is not conclusive. The impact of protection is insignificant. These findings are significant to the extent that they point towards the possibility of improvement in efficiency within the framework of public ownership. Thus, an emphasis on change in the ownership to improve efficiency may be misplaced.