Cover Crops and Interactions with Corn and Soybean Yields: Evidence from Satellite data in Indiana

• Agricultural Economics Workshop

Date and Location

Thursday, March 10, 2022, 4:10 PM - 5:30 PM
Online Meeting, Zoom

Abstract

To mitigate the increasing negative environmental impacts of crop production (e.g., chemical runoff, soil erosion) adopting conservation practices in farming has become a very important policy and research focus. However, the interest of farmers in adopting conservation practices is highly dependent on the potential on-farm benefits that these practices bring (Lee et al 2018). Planting cover crops is one such practice which can minimize negative externalities of farming on the environment, and at the same time contribute to factors that promote crop resilience and productivity (Bergtold et al, 2012; Myers and Watts 2015; Wittwer 2017). With the potential for increased weather variability and frequency of extreme weather events, cover crops have been gaining more attention, as their on-farm benefits can provide an effective cushion against unfavorable weather events such as drought and excess moisture.
In this study we investigate the effect of cover crops on the yield of corn and soybeans. We also explore whether cover crops can help reduce the yield loss of these crops during excess dryness or moisture events. While some studies have looked at the economic benefits of farms adopting cover crops, such as changes in farm profits or NPV risk of a cover crop investment (e.g. Boyer et al 2017), this study is the first to our knowledge that investigates yield loss abatement effect of cover crops on practicing farms in the US.
For this study, we utilize county level cover crop acreage data for corn and soybeans in the state of Indiana from 2006-2015 obtained from the Operational Tillage Information System (OPTIS) dataset. Crop yield data were obtained from the RMA Information Reporting System. Additionally, weather variables used in our analysis were produced from data collected from Oregon State PRISM weather data bank. Taking advantage of the panel nature of our data, we utilize county/crop linear fixed effects specifications, and several sensitivity tests in our primary analysis to address endogeneity issues related to time invariant, unobserved factors such as farm management and land quality. We also include year fixed effects to account for year specific, particularly weather induced, factors such as annual fluctuations in pest pressures.
Despite the use of fixed effects modeling, concerns about endogeneity in our primary explanatory variable (i.e., cover crop acreage) still exist in our analysis. To validate the strength of our fixed effects model results and address other endogeneity issues we implement the IV free regression which allows us to assess the behavior of our coefficient estimates based on specific assumptions about the remaining endogeneity in our model. In this method, we exploit non-orthogonality conditions in the form of bounds on the admissible extent of endogeneity which enables us to build feasible test procedures that do not require instrumental variables.
Findings from this study show that planting cover crops do not have any significantly beneficial effect on the mean yield of corn and soybeans at the county level. However, our results suggest that there is a significant reduction in yield loss for crops planted to a cover during extreme weather conditions. Parameter estimates suggest that crops planted to a cover can have as much as a 5 percent reduction in damages due to drought and 7 percent reduction in damages due to excess moisture. On the other hand, results from our IV free regressions give us a range of parameter estimates for the base effect of cover crops on mean yield of corn and soybeans and damage reduction under extreme weather conditions. Our results suggest that the sign of the effect of cover crops on mean county level crop yields depends strongly on the sign of the bias correlation in cover crop observations. For damage abating effect of cover crops however, we find a significant and positive effect of cover crops on corn and soybean yields for a large range of endogeneity assumptions. Specifically, for bias correlation of cover crop observations greater than -0.4, the damage abating effect on crops planted to cover crops is positive, statistically significant and robust to alternate specifications.
Our study has important implications for the crop insurance program and the potential for rating counties with high adoption rates of cover crops. Additionally, as cover crop adoption remains low, much work has gone into understanding the benefits on farms with which may drive adoption. With sparse findings in the way of increased profitability or overall risk protection (reduced variance) our work shows that cover crops may offer protection from downside risk specifically, and that such a focus on downside risk protection may be a useful view that could be used to drive adoption of the practice. Using the IV free analysis, our work can also lead to discussions on endogeneity in studies such as ours and lend understanding for improvements in future related work.