PIPE Workshop: Hannah Druckenmiller (Caltech)

Working at the intersection of economics and machine learning, I study the economic tradeoffs involved with land use regulation. I begin by demonstrating how we can use deep learning to predict the scope of Clean Water Act regulation. Then I show how these predictions can be used to quantify the economic costs of regulation in terms of foregone development opportunities. Four findings emerge from this analysis. First, recent rule changes to the Clean Water Act greatly alter regulatory stringency. Second, regulation decreases development activity, as measured from permitting and satellite data. Third, Clean Water Act regulation substantially decreases the values of non-residential properties. Fourth, we compute the economic costs of the Clean Water Act's land use regulation. Finally, I introduce a new, data-driven framework for estimating the economic benefits of regulation in terms of environmental services that can be compared to the costs of regulation.

Host or Publisher PIPE Collaborative