Abstract:We contribute to the empirical understanding of how private individuals and institutions respond to government spending on public goods by examining the effects of U.S. federal land programs on private land trust conservation . Our econometric analysis evaluates the crowding out (or crowding in) of land trust acres at the county level using a panel regression framework . Using data from The Nature Conservancy (TNC) and from Land Trust Alliance (LTA) censuses, we have identified or estimated the total acres of open space held by land trusts in each U.S. county during 1990 and 2000 . On the government side, we have constructed a county-level panel of Conservation Reserve (CRP) acres, Wetland Reserve (WRP) acres, and federal land holdings (e.g., national parks and forests) . With these data we estimate separate crowding out\crowding in effects for TNC and LTA trusts . We also provide the first analysis of other county-level determinants of growth in private land conservation including population and per capita income growth . With the LTA data, we find a small crowding-out effect from CRP that is largest in counties with significant endowments of natural amenities . With the TNC data, we find both crowding in and crowding out, depending on the government program in question . The effects of other factors such as population growth, per capita income growth, and a county's natural amenity endowment also differ between TNC and the LTA trusts due to institutional differences that we describe.The authors are grateful to John Shoaf and Keri Perez for able research assistance. 1 The conditions identified by Rose-Ackerman include the presence of increasing returns to scale in the production of public goods and the use of matching grants by governments. 2 The cited studies use field data. Other empirical studies use data from laboratory experiments to infer incomplete crowding out (see, e.g., Andreoni 1993, Bolton and Katok 1998, Eckel et al. 2005.