Governments in EMDEs routinely intervene in agriculture markets to stabilize food prices in the wake of adverse shocks. Such interventions involve a large increase in the procurement and redistribution of agriculture output, which we refer to as a redistributive policy shock. What is the impact of a redistributive policy shock on inflation and the distribution of consumption amongst rich and poor households? We build a two-sector-two-agent NK-DSGE model (2S-TANK) to address these questions. Using Indian data, we estimate the model using a Bayesian approach. We characterize optimal monetary policy. We show that the welfare costs of redistributive policy shocks are substantially higher when non-optimized rules are used to set monetary policy in response to such shocks.