Hurdles and Steps:Estimating Demand for Solar PhotovoltaicsTsvetan Tsvetanov, Yale UniversityKenneth Gillingham, Yale UniversityNovember 12, 2014JOB MARKET PAPER: AbstractIn light of the steadily decreasing step schedule of financial incentives for solar photovoltaic (PV) installations in a number of states, understanding the demand for residential PV systems is important for both policymakers and firms. This paper estimates residential solar demand in Connecticut using a new approach to address three empirical challenges that can arise with count data in our setting: excess zeros, unobserved heterogeneity, and endogeneity of price. We develop a Poisson hurdle model that allows for both fixed effects and instrumental variable estimation. Our results imply a nearly unitary price elasticity of demand for solar PV systems of -1.03. Counterfactual policy simulations suggest that the number of new installations in Connecticut in 2013 would have been 35 percent less than observed in the absence of state financial incentives. Policies to eliminate permitting costs, such as those implemented in several states, would have increased the number of new installations by 1.5 percent.