One set of simulations tests alternative levels of distribution, finding that cutting distribution to 1 million tons would have minimal effects on the level of rice price stability. Another set of simulations tested different import tariff levels, including the baseline rate of 25%1. We find that lower tariffs result in both lower rice prices and less price instability, as world rice prices tend to be more stable than local prices. Simulating a buffer stock with different price bands shows that a narrow band can achieve high price stability but at a high fiscal cost. A 20 T/kg (USD 0.26/kg) price band generates similar price stabilization at a lower cost compared to current policy. However, it is difficult to set the “right” purchase and sale price, and many simulations result in exhausting reserves or reaching warehouse capacity. An adaptive buffer stock, in which the price is adjusted as the stock runs too low or too high, solves some of these problems. In general, the study finds that current procurement and distribution patterns do not match well with the regional and monthly patterns of surplus and deficit, possibly reflecting multiple and conflicting goals of the public food distribution system.