Measuring the Effects of Working-Land Conservation Programs on Adoption of Soil-Erosion Reducing Practices and Permanent Vegetative Cover

We use Maryland farm-level data to study the overall effect of voluntary conservation programs on permanent vegetative cover and the level of adoption of three soil-erosion reducing practices. In order to control for self-selectivity in participation we use a multivariate switching regression model where censored response equations correspond to the levels of adoption of the different practices under analysis. Full information maximum likelihood estimation is made feasible by a Monte Carlo Expectation Maximization algorithm. We find that participation increases the levels of adoption of the three erosion-reducing practices but it reduces permanent vegetative cover. Additionally, the magnitudes of program effects change with farm size; reduction in vegetative cover is more intense among smaller participant farms, while the greatest increase in adoption of minimum or no tillage is observed on large farms.