Press and Firms Accountability: Evidence from Toxic Emission in the US

In this paper I investigate whether newspapers make rms accountable on their environmental standards, extending on the literature on the e ect of mass-media on economic and political outcomes. I compare plants that are located in the same county and operate in the same 3-digits industry and year, but have di erent distances from the closest newspaper. My results suggest that being 9 miles closer to a newspaper with respect to the average distance corresponds to having about 29% less toxic emissions, as measured within Environmental Protection Agency administered Toxic Release Inventory Program. This is consistent with a model in which there is asymmetric information between a rm and agents interested in its plants (communities, regulators, consumers, investors), and plants emit less toxic substances because they are worried about negative coverage that would reveal hidden information. Indeed, I show that in the sample of Top 20 polluters by State the probability that coverage of plant-level TRI statistics arise in the plant's closest newspaper is larger, the lower the distance from the closest newspaper itself. ∗IIES, Stockholm University. Email: [email protected]. I thank my supervisor David Stromberg for sharing with me his data on US newspapers, and for many useful comments and suggestions on this project.