The value of Earth’s agricultural and forested land has been estimated at 27.2 trillion dollars, more than the combined market capitalization of the World’s 100 largest companies. Local variation in the value of land is a critical determinant of land-cover and land-use change (LCLUC). Land prices underpin patterns of urban sprawl, the designation of protected areas, and the emergence of agricultural frontiers. Land prices also serve as one of the best ways to reveal the value society places on the environment. While land prices provide important information to understand and predict land-use change, representative, spatially explicit data on land values are rare, especially in rural portions of the Global South.
The overarching goal of our research project is to predict property values using remotely sensed data on agricultural practices, and highlight the novel social science made possible through the availability of this data. To achieve this goal, we will focus upon the following objectives:
Characterize field-level agricultural practices: Remotely sensed time-series data are crucial for understanding agricultural dynamics over large regions. We will use a spatial-temporal fusion of data from multiple satellite sensors to generate maps of crop species and agricultural practices at the field-scale;
Map rural property values: Although land is the most important financial asset for rural households throughout the world, spatially explicit data on its value is sparse. Using a machine-learning based hedonic model, we will combine government datasets on property transactions with detailed data on each property’s agricultural practices to predict rural property values;
Quantify the costs of conservation: Deforestation restrictions on private lands are a common conservation policy in South America, but their costs and distributional impacts are largely unknown. Using econometric, quasi-experimental methods applied to property value data, we will generate causal estimates of the costs such regulations impose upon private landowners.
Our research will focus on the Pampas, Chaco and Alto Paraná Atlantic Forest of Paraguay and Uruguay, critically understudied ecoregions with an urgent need for advances in land system science. By developing new methods to map agricultural practices and property values in this region, we will help fill a crucial gap in the study of global LCLUC processes. Land values are a critical determinant of both LCLUC and human livelihoods. As a result, we expect that our research will yield transformative impacts on the science and practice of many fields including agriculture, rural development and conservation planning.