Measures of capacity utilization and its determinants: a study of Indian manufacturing

We estimate Capacity Utilization (CU) rates in for selected industries in Indian manufacturing for the 20-year period 1976–1996. We estimate a generalized Leontief variable cost function, with capital as a quasi fixed input, to derive our CU measures, using error-component techniques. We note substantial variations in CU both across industries and over time. In general, we find that CU rates were higher in the earlier time-period, dropped in the mid-80s and started rising again in the early 90s. CU rates in our analysis are sensitive to input prices with the sole exception of the price of labor. We also confirm the standard result that variations in demand are a significant driving force for variations in CU. We find that CU is positively related to the magnitude of labor intensity in production. This holds for both between-industries and within-industries. Empirical results also indicate that traditional measures of CU such as minimum capital output ratio and peak-to-peak are not appropriate proxies for the short-run decision making of the firm regarding CU. As compared to the estimates derived from the choice-theoretic framework, we find that the traditional measures exhibit substantial bias.