A large share of American land conservation happens on privately owned land protected through conservation easements, where a landowner receives tax incentives in exchange for permanently donating the right to develop their land. In Virginia, measuring environmental value with a state-constructed ranking of conservation priority shows that the quality distribution of easements is wide, despite the state's large tax incentives and strong checks on easement fraud. The quality distribution of easements approximates the distribution of unconserved undeveloped land statewide and is much lower than that of publicly owned conservation land. A difference-in-difference analysis around a 2002 tax reform finds that increasing tax incentives attract donations of lower quality, particularly in regards to agricultural land. I use these results to build a model of conservation easement supply. Compared to a universal increase in conservation subsidies, offering increased tax incentives only to land with high environmental value could substantially increase the amount of high-quality land conserved at a cost of 1.18 acres of lower quality land per acre of high quality land. Comparisons to a model with constant marginal quality show that ignoring the changes in marginal quality can lead to overestimates of the environmental benefits of any tax subsidy increase. The no-selection model overestimates the very high quality acres conserved by a nontargeted subsidy increase by 75%, and by a targeted subsidy increase by 35%.
