I investigate the role of demand shocks in the ready-mix concrete industry. Using Census data on more than 15,000 plants, I estimate a model of investment and entry in oligopolistic markets. These estimates are used to simulate the effect of eliminating short-term local demand changes. A policy of smoothing the volatility of demand has a market expansion effect: The model predicts a 39% increase in the number of plants in the industry. Since bigger markets have both more plants and larger plants, a demand-smoothing fiscal policy would increase the share of large plants by 20%. Finally, the policy of smoothing demand reduces entry and exit by 25%, but has no effect on the rate at which firms change their size.