Event Recap: A Latent Class Analysis of Farmer Preferences Regarding Filter Strip Programs

Nov. 25, 2012

On Monday, November 19, 2012 AEDE welcome Gregory Howard, AEDE Postdoctoral Researcher, to present his research titled, “A Latent Class Analysis of Farmer Preferences Regarding Filter Strip Programs” as part of the AEDE Seminar Series.

Howard’s research, which was co-authored with AEDE Professor Brian Roe, focuses on Ohio farmers’ preferences for filter strip programs on their farms. Filter strips help to eliminate agricultural run-off into waterways. Agricultural runoff has left many Ohio lakes, rivers and streams with increased levels of nutrients, resulting in unbalanced ecosystems in these water systems.

Howard’s research aims to assess farmer knowledge of the effects of agricultural runoff, as well as gauge their support of filer strip programs. He aims to help policymakers to answer questions about the use and adoption of filter strip technology, including: how do you encourage people to have better adoption rates of filter strip programs? How do we do it at a lower cost? How can we make payments for ecosystem services better? Howard has presented this research at a number of locations in the past few months, including a recent presentation to policymakers at the Water Management Association of Ohio (WMAO) Conference.

By utilizing the responses from a survey of Ohio corn and soybean farmers, which was conducted as part of a National Science Foundation Impact Grant awarded to AEDE faculty members Brian Roe and Elena Irwin, Howard analyzed responses to examine to what extent farmer preferences for filter strip programs were influenced by the filter strip program attributes. In looking at the survey data he also aimed to gauge farmer perceptions of the efficacy of filter strips in preventing agricultural runoff.

In conducting his analysis, Howard adopted a two-stage estimation method to control for endogeneity of the efficacy measure using exogenous field attributes and exogenously assigned filter strip size as instruments. He utilized latent class analysis in both the first and second stage, allowing for discrete preference heterogeneity among farmers. In the second stage, he used a conditional logistic regression for which the dependent variable is an indicator variable for whether the program was considered the best.

Howard found that farmers do exhibit preference heterogeneity regarding filter strip program selection and that increases in farmers’ perceptions of program effectiveness corresponded to greater likelihood of filter strip program enrollment. He organizes the farmers in his research findings into one of two latent classes: a larger class for which increased program efficacy and decreased filter strip width increased the probability of filter strip program enrollment, and a smaller class for which filter strip width and perceived efficacy were not statistically significant for increasing the probability of their adoption of filter strip programs. He notes that the smaller class also exhibited a strong preference for the status quo, suggesting that members are generally less likely to enroll in conservation programs. Additionally, he found that while the majority of farmers believe that various field and filter strip attributes can impact the ability of filter strips to reduce the probability of runoff, approximately 20 percent of his sample communicated that they believe that filter strips have no effect on runoff, regardless of field or filter strip attributes.

This research is the first analysis to allow for farmer preference heterogeneity through latent class analysis. Previous work in the field has indicated that, in addition to larger payments, farmers are more likely to engage in conservation programs when they have higher incomes, larger farms, more education, lower quality soil, and when they express greater levels of concern for the environment. However, no studies have been able to isolate the degree to which farmers believe these practices will reduce runoff, which Howard’s research is one of the first to do. The research is also the first to examine how farmer perceptions of program efficacy in reducing runoff influence their preferences towards filter strip programs.

To see Howard’s full presentation at the AEDE seminar, please click here.

November 26, 2012