Agricultural production negatively impacts the environment. Nitrate and pesticide runoff contaminate drinking water. Animal waste causes bacterial contamination. Odor from livestock operations affects nearby residential areas. The list goes on. Despite these negative impacts, farmers continue the production practices causing these issues, while garnering the benefits but not bearing the costs – a textbook scenario of the tragedy of the commons. To correct this problem, governments intervene with a variety of policy tools. In particular, mandates are an increasingly popular corrective instrument. Using them, governments simply require that farms adopt production practices that are more environmentally conducive. For instance, Ohio and Vermont require manure to be applied as fertilizer during the spring rather than the winter to prevent nitrogen runoff that occurs when rain and snow wash the manure off into nearby bodies of water. Mandates like these are easy to implement and produce immediate, visible environmental benefits.
Unfortunately, they are also extremely blunt policy instruments that can be destructive if they are not adequately tailored to the farms they are applied to. Take the manure application example cited above. When farms are prohibited from applying manure during the winter, they must transport excess waste off the farm. Empirical studies have found that these transportation costs are highest for small-sized farms, lowest for medium-sized ones, and midway between for the largest operations because of the difference in the scale of their economies. By not taking these differences into account, the manure application mandate causes large farms to downscale their operations until the transportation costs become feasible and small farms to downscale or even exit the market entirely. This lost economic activity is an unintended cost that may well outweigh any environmental benefits.
Given these considerations, it is important to ensure that farms are not being adversely coerced into adopting sub-optimal practices. The first step toward accomplishing this is to measure how many farms are being coerced. For instance, would farms in Ohio and Vermont that were prohibited from applying manure during the winter have chosen to not apply the manure anyways? If so, then the mandate did not coerce them at all. However, if they would have spread manure during the winter, then it is possible that the mandate destroyed more economic benefits than the environmental benefits produced. This is because those farms would not have spread the manure unless it was economically beneficial for them to do so. In this way, simply measuring how many farms were coerced by the mandate is a first step toward gauging whether farms may have been adversely coerced.
Data and Methods
However, measuring this is difficult because it is impossible to know what the farms would have done absent the mandates because such a situation ceased to exist when the mandates were imposed. This unobserved situation can be simulated by comparing the farms in regulated states with farms of similar size and type in non-regulated states. For example, while it cannot be known whether farms in Ohio and Vermont would have applied manure during the winter, it can be extrapolated what they would have done by comparing them to farms of similar size and type in New York and Pennsylvania where farmers may apply manure year-round. Using the USDA’s Agricultural Resource Management Survey (ARMS) data and logistic regression, I was able to make a comparison.
The ARMS is a comprehensive survey including data on farm production practices and household characteristics for about 5,000 fields and 30,000 farms each year from the 48 contiguous states. The ARMS data is uniquely suited for measuring the effect of mandates for two reasons. First, it is the only nationally representative data set linking farm production practices and household characteristics. Since we want to measure the effect of mandates on farms operating under different state requirements, but still have the same household characteristics, a nationally representative survey is necessary and ARMS alone meets this need. Second, ARMS is the only data set that includes information on whether farms implemented their production practices to satisfy mandates or for other reasons. This uniquely enables comparisons between farms that implemented practices because of mandates and those that implemented them for reasons other than mandates. By training a logistic regression model on the latter group and then applying it to the former, we estimated how many of the farms that were compelled to adopt practices by mandates, would not have adopted those practices without being required.
The model found that between 42 and 61 percent of farms compelled by mandates to adopt certain practices would not have done so in the absence of the requirement. This implies that the mandated practices were most likely not economically beneficial for these farms. Thus, while the mandates may have produced environmental benefits, these benefits may well have been offset by economic costs. Given the substantial portion of farms for which this was the case, these findings demonstrate the need for further work analyzing these benefits and costs to ensure farms are not being compelled to adopt practices with insufficient environmental benefits to justify the coercion.