Intermediate Input Substitutability

We estimate long-run elasticities of substitution between intermediate inputs for Indian manufacturing plants. India's trade liberalization in the early 1990s provides an ideal natural policy experiment, with permanent and heterogeneous tariff reductions inducing changes in relative prices which we use for identification. We find a high degree of substitutability at the plant-level between 8 broad categories of material inputs, significantly above the Cobb-Douglas benchmark of 1. In contrast, when considering shorter time horizons or more transitory shocks to relative prices, we find evidence of much lower elasticities. Using our long-run elasticities in a quantitative model, we find that misallocation of intermediate inputs causes a reduction in aggregate output of 29% - 3.5 times larger relative to a standard calibration with complementary inputs.