Technical and allocative efficiency in Indian dairy industry: a comparative study across institutional structures

Abstract

The dairy sector in India has undergone rapid changes in last two decades in terms of milk production, market structure and Government policy orientation. India’s emergence as the largest milk producer in the world, coupled with the availability of liquid milk at competitive farm-gate prices and favourable Government policies since 1992 saw the entry of large number of private players to compete with the cooperatives. The consequent sector wide structural change resulted in the erosion of the cooperative monopoly. The structural changes coupled with the availability of milk in plenty resulted in expectations around its future potential as a major international dairy player and gave rise to speculations and debates on domestic competitiveness of the industry, in general and relative performance of institutional structures, in particular. The aforesaid speculation and debate affords opportunity for academic interventions which serves as the motivation for this study. The primary intent of the research was to study the performance of the private plants vis-a-vis the cooperatives in milk procurement and processing. Prima facia, one is inclined to hypothesize that the private dairy firms are efficient enough to compete with the cooperatives, as assumed in the basic premise to the Milk and Milk Products Order or MMPO (1992). The changes in Government policy through MMPO (1992), and its subsequent amendments, that lead to the de-licensing of the sector to facilitate entry of private dairy firms, were based on this hypothesis. Even some limited studies in terms of scale, scope and empirics have supported the same while assessing the performance across the institutional structures, namely the private plants and the cooperatives.

The alternative, but less apparent, hypothesis, cautions against this assumption, and claims that the cooperatives are more efficient than the private plants and hence better suited in terms of international competitiveness. The comparative efficiency across the institutional structures is, however, an empirical issue and empirical evidence in this regard is scarce. This study makes a contribution to an improved understanding of dairy market structure by rigorous empirical estimations of the efficiency across the institutions. Acute scarcity of systematic research on milk procurement and processing is the second motivation for this study. Keeping in mind the motivations discussed above, this study has two objectives. First is to compare the cost efficiency in procurement and processing across the institutional structures and secondly, to explore the determinants of variation in plant cost efficiency. Three main product lines are chosen for the study, namely, fluid milk, ghee and butter and milk powder. Together the product lines comprise over 95 percent of the total milk and milk products produced in India.

Two efficiency measurement methods are used: stochastic cost frontier analysis and data envelopment analysis, followed by censored regression analysis to explore the determinants of efficiency. In cost frontier, both Cobb-Douglas and the more flexible translog functions are applied for each product line as well as to determine the overall efficiency. The study utilizes an unbalanced panel data sample of 48 plants, comprising 27 cooperative plants and 21 private plants spread across Gujarat, Punjab and Haryana, observed between 2000-1 and 2005-6.

The results of stochastic cost frontier analysis and DEA suggest that the cooperatives are significantly more efficient than the private dairy plants (at 5% level). Average cost efficiency of both cooperatives and private plants are estimated to be 0.95 in the stochastic cost frontier analysis when controlled for institutional structure with the dummy variable representing institutional structure. The estimates also suggest significant efficiency advantage for the cooperatives. In the multi-output input oriented variable returns to scale DEA model, the cost, technical and allocative efficiency of cooperatives are estimated at 0.86, 0.91 and 0.95. For private plants the comparable figures are 0.75, 0.84 and 0.89 with significant differences across the institutional structures. The significant efficiency advantage for cooperatives is found to be consistent across procurement, processing of all the product lines, and in the overall procurement and processing of milk and milk products.

The results of the censored regressions show that capacity and capacity utilization, investments in plant infrastructure and milk-sheds, sophistication in manufacturing practices and diversification have significant positive influence on plant efficiency. The cooperatives are found to be significantly more efficient than the private plants even after discounting the factors related to plant efficiency. However, the level of significance decreased in the censored regression analysis as compared to the DEA and stochastic cost frontier analysis. The cooperative’s control over the supply chain in procuring milk with the desired quality, possibly explains their comparative efficiency over the private players.The private plants, on the other hand, procure milk through third party contractors resulting in lesser control. Furthermore, most of the private plants, procure milk from a few selected districts, characterised by a rich buffalo population, where they compete among themselves apart from competing with the cooperatives. The limited access constrains their supply of quality milk especially during the lean seasons. The cooperatives, on the other hand, are found to be scattered throughout the States, procuring from even the interior-most villages which enables them to access better quality raw milk. It is thus argued that the extensive spread of their milk supply source and the stringent control exercised over the chain by the cooperatives augment their relative efficiency.