| This dissertation consists of three essays on the relationship between firm incentives and energy and environmental policy outcomes.;Chapters 1 and 2 study the impact of the 1990 Clean Air Act Amendments on the United States oil refining industry. This legislation imposed extensive restrictions on refined petroleum product markets, requiring select end users to purchase new cleaner versions of gasoline and diesel. In Chapter 2, I estimate the static impact of this intervention on refining costs, product prices and consumer welfare. Isolating these effects is complicated by several challenges likely to appear in other regulatory settings, including overlap between regulated and non-regulated markets and deviations from perfect competition. Using a rich database of refinery operations, I estimate a structural model that incorporates each of these dimensions, and then use this cost structure to simulate policy counterfactuals. I find that the policies increased gasoline production costs by 7 cents per gallon and diesel costs by 3 cents per gallon on average, although these costs varied considerably across refineries. As a result of these restrictions, consumers in regulated markets experienced welfare losses on the order of ;Chapter 2 estimates the sunk costs incurred by United States oil refineries as a result of the low sulfur diesel program. The complex, regionally integrated nature of the industry poses many challenges for estimating these costs. I overcome them by placing the decision to invest in sulfur removal technology within the framework of a two period model and estimate the model using moment inequalities. I find that the regulation induced between ;Chapter 3, coauthored with Hunt Allcott, turns the to retail market for water heaters to study the topic of energy efficiency. We run a natural field experiment at a large nationwide retailer to measure the effects of energy use information disclosure, customer rebates, and sales agent incentives on demand for energy efficient durable goods. We find that while a combination of large rebates plus sales incentives substantially increases market share, information and sales incentives alone each have zero statistical effect and explain at most a small fraction of the low baseline market share. Sales agents strategically comply only partially with the experiment, targeting information at more interested consumers but not discussing energy efficiency with the disinterested majority. These results suggest that at current prices in this context, seller-provided information is not a major barrier to energy efficiency investments. We theoretically and empirically explore the novel policy option of combining customer subsidies with government-provided sales incentives. |
