Oral Abstract Details

Author(s):
DAVID A. FLEMING

Affiliation:
PENN STATE UNIVERSITY

Presentation Type:
Oral

Topic Area:
Spatial/scale aspects of land-use change

Abstract Text:
As cities grow, land conservation has arisen as main concern for many planners. However, few researches have examined how these two phenomena can affect final land use decisions done by owners of non-protected areas. To address this point, in this work I use cross-sectional models to account for the marginal effect that conservation programs and urban growth have over agricultural expansion in counties of the Midwestern U.S. Following empirical models used to address land use change in the literature, I use as dependent variable the change of non-urbanized land to agriculture (between 1992 and 2001) as well as disaggregated changes from forest, wet and grass lands to agriculture. Main explanatory variables include land under the Conservation Reserve Program (CRP) and urban sprawl, among others. In order to avoid endogenenity problems from these land use covariates, alternative models consider the use of instrumental variables (IV) defined by the weighted average of the variables (CRP and sprawl) in counties that surround the county under scrutiny. In other words, to predict agricultural expansion in county j, IV models consider as independent variables the average values of conservation land and urban sprawl of its neighboring counties i’s (with i ≠ j). As source of land cover change data, this work analyzes satellite imagery from the NLCD Retrofit Change Product. This product, released in 2008 by the USGS, provides the most accurate (to date) remote sensing data of land cover changes in the U.S. (for the period 1992-2001). In particular, with this resource it is possible observe changes in land covers at a pixel precision of 30x30 meters. Estimation results show that, in average, the percentage of land under the CRP has a larger effect than urban sprawl on the rate of agricultural expansion of a particular county. Both variables present different magnitude and significance depending on the original land cover to be converted to agriculture. The analysis finds that, for example, the impact of the CRP on land conversion from wetlands to agriculture has a lower magnitude and significance than changes from grassland to agriculture.