This paper presents a new method for estimating treatment effects of regulations in the presence of inter-firm and intra-firm spillovers. Difference-in-differences (DD) estimation cannot identify effects when the regulated outcomes influence the unregulated, as when firms compete. To overcome this limitation, we posit a structural model of differentiated product markets that generalizes previous work and show how to estimate treatment effects for multiplant firms. We recover unbiased estimates of treatment effects in Monte Carlo experiments, while DD and other available methods do not. We apply our method to study the European carbon market's effects on French manufacturers. We find that this regulation reduced emissions of regulated plants, increased sales of regulated firms, and decreased emissions in aggregate, relative to an unregulated counterfactual. (co-written with Geoffrey Barrows, Raphael Calel, and Martin Jégard)