CHOICES
A publication of AAEA
Conservation programs funded and managed by the USDA have two key goals: protecting natural resources and supporting farm profitability. The environmental benefits of these programs, such as reduced soil erosion and improved habitat for wildlife, are important for societal well-being. Their impacts on farm productivity and profitability are equally important. Conceptually, conservation programs could increase yields because some conservation efforts boost soil health on the fields where they are implemented or on nearby fields. They could also increase county-level yields through retirements of less-productive land. However, if there are tradeoffs between yield and ecological functions, or if some productive land is taken out of production, then conservation programs could decrease yields. This article provides evidence that USDA conservation programs can increase agricultural yields. Our findings make a compelling case for continued funding of USDA conservation programs that could simultaneously boost agricultural productivity and improve environmental outcomes.
To be specific, we study how payments from the Environmental Quality Incentives Program (EQIP) and Conservation Reserve Program (CRP) affect the county-level yields of three major crops: corn, soybeans, and winter wheat. Our analysis combines data on all EQIP contracts from 2005 to 2015 (as in Liu, Wang, and Zhang, 2023) with annual crop yield data from USDA-NASS (2024). Yield data are available for 1,700–2,000 counties, depending on the crop. We also use county-level data on CRP payments, weather, and soil characteristics.
Agronomists have evaluated the yield impacts of a small number of conservation practices, such as cover crops and conservation tillage (for example, Marcillo and Miguez, 2017; Peng et al., 2024). In contrast, our study examines the effects of the entire suite of EQIP practices using data on actual payments made to farmers. Sinceconservation practices have different units, using payments allows us to include all EQIP practices together in a single analysis.
Our analysis is based on a straightforward econometric model that allows us to understand the relationship between crop yields and EQIP payments received in a county in the past few years. To isolate the effect of EQIP, we control for several other factors that could also influence yields. The most important of these is CRP payments. CRP requires farmers to take fields out of production, and it is likely that farmers enroll their less-productive fields in CRP, which could increase average yields on the remaining farmland in the county. In addition, enrollment in EQIP and CRP could be correlated at the county level. It is therefore important to quantify the effects of EQIP and CRP simultaneously.
In addition, our analysis controls for broader economic and environmental conditions that affect all farms in a given county, such as quality of agricultural extension services, or all farms in a given year, such as an increase in fertilizer price. To control for these general conditions, our regressions allow for county-specific intercepts that capture time-invariant county-level characteristics. Similarly, we also use state-year dummy variables to control for state-specific annual unobservable variables.
We find that conservation under EQIP and CRP in the past few years is associated with small but measurable increases in crop yields at the county level. Specifically:
Because we use county-level data, our estimates represent the overall impact of conservation programs on the average yields of the county. We find stronger yield effects when accounting for payments over more than 1 year, which suggests that the benefits of the programs may take time to materialize. While these yield effects are small, they suggest that conservation efforts can enhance agricultural productivity. It is encouraging to know that conservation programs overall provide private benefits to farms by improving agricultural productivity along with generating societal environmental benefits.
The Soil Erosion Act of 1935, introduced in response to the Dust Bowl, marked the beginning of federal efforts to promote soil conservation (Baylis et al., 2022). Subsequent Farm Bills initially focused on retiring sensitive cropland under programs such as the Conservation Reserve Program (CRP), which was established in 1985. More recently, Farm Bills have allocated funding and technical support to farmers who implement conservation practices on actively working land through programs such as the Environmental Quality Incentives Program (EQIP), which was first authorized in 1996.
The Environmental Quality Incentives Program (EQIP), administered by the USDA’s Natural Resources Conservation Service (NRCS), covers up to 75% of thecosts of materials and labor associated with adopting conservation practices. NRCS headquarters allocates EQIP funds to state offices, which then distribute funds within their states using state-specific ranking tools to score applications. These ranking tools sometimes use
similar criteria but different weighting (US GAO, 2017). Payment rates and funding decisions are subject to each state’s priorities and available budget. During our study period, NRCS programs, in particular EQIP, were oversubscribed, and NRCS only funded about 35% of applications received due to budgetary constraints (USDA-NRCS, 2014). Federal law also imposes certain requirements on EQIP spending. For example, at least 50% of the financial assistance funds must support livestock operations. The 2018 Farm Bill authorized just over $2 billion in funding for EQIP in fiscal year 2023, and the Inflation Reduction Act added more than $8 billion for fiscal years 2023–2026.
Between 2005 and 2015, more than 200 unique conservation practices were adopted with support from EQIP, targeting improvements across soil, water, plant, animal, energy use, and air quality. EQIP practices were implemented in 98% of US counties during this period. Figure 1 maps average annual EQIP payments by county from 2005 to 2015. Regions with large counties—such as California, the Southwest, Northwest, and Northern Plains—and near precious water resources—such as the Chesapeake Bay, Lake Michigan, the Mississippi Delta, and the Everglades—are shown to have received more payments. These patterns illustrate the voluntary nature of participating in the EQIP program and the correlation between payments and agricultural production.
Under the Conservation Reserve Program (CRP) and Conservation Reserve Enhancement Program (CREP), the USDA’s Farm Service Agency (FSA) pays farmers to take environmentally sensitive or less-productive land out of production. Farmers convert this land to conservation uses—such as grasslands, forests, or wetlands—under contracts that typically last 10–15 years. These land retirements aim to reduce soil erosion, improve surface water quality, and create wildlife habitat (Babcock et al., 1996; Riley et al., 2019). From 1990 to 2010, about 9% of US cropland acreage was enrolled in CRP (Taylor et al., 2021). In 2024, producers received about $1.7 billion through the CRP and CRP Transition Incentive Program (USDA-FSA, 2024b).
To study the effect of EQIP on agricultural yields, we constructed a panel dataset from various sources. We spatially aggregated EQIP payments and practices from the watershed level to the county level because our other data, including data on crop yield, weather, and soil characteristics, are reported at the county level. This approach allows us to examine how EQIP payments within counties relate to changes in crop productivity over time.
We obtained watershed-level data on EQIP practices and payments directly from USDA-NRCS (Liu, Wang, and Zhang, 2023). From 2005 to 2015, EQIP payments were made in 3,071 of the 3,143 US counties. We adjusted all payment amounts to 2015 dollars to ensure comparability over time. We also know the unit of each practice, such as acreage or feet. We summed all practices with acreage as their unit to construct acres of EQIP practices at the county level for a robustness check.
We obtained county-level data on CRP rent payments made by FSA to farmers (USDA-FSA, 2024a). The dataset also includes county-level data on acres enrolled in CRP.
County-level yield data for corn, soybeans, and winter wheat were obtained from USDA-NASS (2024). Between 2005 and 2015, corn yield data is available for 2,021 counties, soybean yield for 1,699 counties, and winter wheat yield for 1,763 counties. Yields are reported in bushels per acre. Figure 2 illustrates the variation in average yields of the three crops, across counties.
We construct our weather variables using the dataset of Schlenker and Roberts (2009). The dataset provides daily grid-level weather data across the United States. We aggregate the grid-level data to the county level using crop area weights. The growing season is defined as April to September for corn and soybeans and September to May for winter wheat. Following the literature, we use daily minimum and maximum temperatures to construct crop-specific growing degree days (GDDs), specifically: GDD 10–19, GDD 20–28, and GDD 29+ for corn; GDD 10–19, GDD 20–29, and GDD 30+ for soybeans (Schlenker and Roberts, 2009); and GDD 0–10, GDD 11–16, and GDD 17+ for wheat (Tack, Barkley, and Nalley, 2015). Our analysis, based on annual data, uses the sum of growing season GDDs and growing season precipitation for each year.
We obtained data on baseline soil characteristics from Miller and White (2006), who processed the data from USDA-NRCS (2006). This dataset provides comprehensive information about US soil characteristics. To capture baseline soil fertility at the county level, we selected three key variables: pH, available water content, and bulk density. For these, we used measurements from the shallowest soil layers: 0–5 cm for pH and bulk density, and 0–100 cm for available water content. We overlay the soil grids onto county boundary maps and construct county-level averages for each soil characteristic.
Our empirical analysis helps us to explain how EQIP and CRP payments in a county in the past few years affect yields of corn, soybeans, and winter wheat. It is important to control for CRP payments when analyzing the effect of EQIP on yields because CRP is the largest land-retirement program, and less-productive land is more likely to be retired. As a result, CRP enrollment is likely to be positively correlated with yields in the same county, since the retired land no longer decreases average yields. We control for other variables that vary over time, including weather conditions during each crop’s growing season. We also control for baseline soil characteristics interacted with year dummy variables, to allow soil conditions affect yields flexibly over time.
We include county and state-year dummy variables to control for any unobservable features or shocks that are unique to a given county or state-year. Of course, some counties harvest significantly more acreage of corn, soybeans, and winter wheat. We weigh our regressions by cropland acreage to ensure that the estimated magnitudes of the effects reflect the economic size of the counties. Standard errors are clustered at the county level to account for correlation within counties over time.
Table 1 reports the estimated effects of EQIP and CRP payments from the past 5 years on yields of corn, soybeans, and winter wheat. Our sample of 10 years of data does not allow us to statistically meaningfully examine the impacts of conservation efforts from more than 5 years earlier. We demonstrate that EQIP payments improve corn yields. EQIP payments from the previous 5 years have positive and statistically significant effects on corn yields. As shown in column 1, a 10% increase in EQIP payments in the previous year is estimated to increase county-level corn yields by approximately 0.014 bushels per acre, and as shown in column 5, a 10% increase in EQIP payments in the previous 5 years is estimated to increase corn yields by approximately 0.085 bushels per acre. While this may appear minimal, consider that the average county harvests about 40,000 acres of corn. At a price of $5.00/bushel, a 10% increase in EQIP payments in theprevious 5 years would result in $17,000 in additional annual revenue for the average county.
We also find evidence that CRP payments are positively and statistically significantly correlated with soybean yields. Specifically, as shown in the second panel of Table 1, a 10% increase in CRP payments over theprevious 5 years increases soybean yields by at least 0.040 bushels per acre. The positive effect of CRP payments on yields likely reflects farmers’ decisions to retire less-productive land through CRP, thereby raising average yields on the remaining cropland. We have some weak evidence that CRP is associated with reductions in corn yield. As shown in column 2 of the top panel, a 10% increase in CRP payments in the previous 2 years decreases corn yields by about 0.13 bushels per acre. This could be driven by productive land being retired through CRP. Note that the negative impact of CRP payments on corn yields is not statistically significant in most cases. Winter wheat yield is not affected by either EQIP or CRP payments.Implementing conservation practices may enhance or reduce the yields of a farmer’s own field, depending on the practices implemented and how long the practices are in use. It is challenging to separately quantify the effect of a specific practice, considering that conservation practices tend to be implemented jointly.
Conservation practices could also benefit surrounding fields because of the external benefits of erosion reduction or flood control. For example, Karwowski (2023) finds that wetland easements decrease yields on farmers’ own fields but increase yields on surrounding fields. Since our analysis uses county-level data, our estimates capture the overall net effect of all practices on the average county yield.
We consider two robustness checks to assess whether our findings about the relationship between conservation efforts and yields are sensitive to changes to our modeling assumptions. First, we replace EQIP payments with acreage of conservation practices implemented under EQIP and CRP payments with acres enrolled inCRP. Note that not all practices are measured in acres, with examples including watering facilities or irrigation pipelines. Thus, acres of EQIP practices measure only the impact of a subset of EQIP practices. Table 2presents the results of these regressions. We find the same evidence that EQIP practices are associated with increases in corn yields and CRP-enrolled acres are associated with increases in soybean yields.
Because the USDA does not report yield data for every county in every year, we conducted another robustness test by restricting the analysis to counties with complete yield data for each year of the sample period. In other words, we created a balanced panel for each crop. This approach allows us to assess whether missing data affects our estimated relationships between conservation payments and crop yields. Table 3 reports the estimates using a balanced panel for each crop. These results closely align with Table 1, demonstrating that our main results are robust. With this balanced panel, CRP payments are not estimated to statistically significantly decrease corn yields. Thus, the negative impact of CRP payments on corn yields reported in Table 1 might also be driven by inconsistent reporting from counties with relatively little corn production.
In this article, we synthesize data from numerous sources to create a comprehensive panel dataset of county-level agricultural yields, conservation program payments, weather, and soil quality over 2005–2015. We find evidence that a 10% increase in EQIP payments made in the previous year increases county-level corn yields by 0.014 bushels per acre, while a 10% increase in EQIP payments made in the previous 5 years increases corn yields by 0.085 bushels per acre. In addition, we find robust evidence that a 10% increase in CRP payments made in the previous 5 years increases soybean yields by at least 0.040 bushels per acre. Robustness checks further confirm a positive and statistically significant correlation between acres of conservation practices under EQIP and corn yields and CRP enrolled acres and soybean yields.
Our results align with many findings in the existing literature on the effects of conservation practices on agricultural yields. We contribute to the literature by focusing on conservation programs and analyzing the effects of EQIP and CRP simultaneously on county-level yields. A limitation of this article is that we cannot disentangle the causal channels, that is, the impact on fields with conservation practices versus adjacent fields or the impact of one practice versus a combination of practices, in part because our analysis uses aggregated county-level data.
It is important that economists examine the impact of agricultural conservation programs on the environment, farms, and rural communities and communicate these findings with the public. Farmers’ perceptions about the impact of conservation programs affect their participation decisions (Cooper, 2003; Canales, Bergtold, and Williams, 2024). Moreover, conservation programs rely on state and federal public funding, and people might have reservations about government spending, especially on environmental programs. Additional work to evaluate the design and efficiency of agricultural conservation programs is important for agricultural sustainability in a changing climate.
Babcock, B.A., P.G. Lakshminarayan, J. Wu, and D. Zilberman. 1996. “The Economics of a Public Fund for Environmental Amenities: A Study of CRP Contracts.” American Journal of Agricultural Economics 78(4):961–971. https://doi.org/10.2307/1243852
Baylis, K., J. Coppess, B.M. Gramig, and P. Sachdeva. 2022. “Agri-Environmental Programs in the United States and Canada.” Review of Environmental Economics and Policy 16(1):83–104. https://doi.org/10.1086/718052
Canales, E., B.S. Bergtold, and J.R. Williams. 2024. “Conservation Intensification Under Risk: An Assessment of Adoption, Additionality, and Farmer Preferences.” American Journal of Agricultural Economics 106(1):45–75. https://doi.org/10.1111/ajae.12414
Cooper, J.C. 2003. “A Joint Framework for Analysis of Agri-Environmental Payment Programs.” American Journal of Agricultural Economics 85(4):976–987. https://doi.org/10.1111/1467-8276.00501
Karwowski, N. 2023. “The Direct and Indirect Effects of Wetland Easements: Evidence from Field-Level Yield Data in Wisconsin.” Unpublished manuscript. University of Wisconsin–Madison Department of Agricultural and Applied Economics. Available online: https://ageconsearch.umn.edu/record/335462/files/25725.pdf
Liu, P., Y. Wang, and W. Zhang. 2023. “The Influence of the Environmental Quality Incentives Program on Local Water Quality.” American Journal of Agricultural Economics 105(1):27–51. https://doi.org/10.1111/ajae.12316
Marcillo, G.S., and F.E. Miguez. 2017. “Corn Yield Response to Winter Cover Crops: An Updated Meta-Analysis.” Journal of Soil and Water Conservation 72(3):226–239. https://doi.org/10.2489/jswc.72.3.226
Miller, D.A., and R.A. White. 2006. Soil Information for Environmental Modeling and Ecosystem Management [database]. Pennsylvania State University. Available online: http://www.soilinfo.psu.edu/index.cgi?soil_data&conus [Accessed October 8, 2024]
Peng, Y., L. Wang, P.-A. Jacinthe, and W. Ren. 2024. “Global Synthesis of Cover Crop Impacts on Main Crop Yield.” Field Crops Research 310:109343. https://doi.org/10.1016/j.fcr.2024.109343
Riley, D., T. Mieno, K. Schoengold, and N. Brozovíc. 2019. “The Impact of Land Cover on Groundwater Recharge in the High Plains: An Application to the Conservation Reserve Program.” Science of The Total Environment 696:133871. https://doi.org/10.1016/j.scitotenv.2019.133871
Schlenker, W., and M.J. Roberts. 2009. “Nonlinear Temperature Effects Indicate Severe Damages to US Crop Yields Under Climate Change.” Proceedings of the National Academy of Sciences 106(37):15594–15598. https://doi.org/10.1073/pnas.0906865106
Tack, J., A. Barkley, and L.L. Nalley. 2015. “Effect of Warming Temperatures on US Wheat Yields.” Proceedings of the National Academy of Sciences 112(22):6931–6936. https://doi.org/10.1073/pnas.1415181112
Taylor, M.R., N.P. Hendricks, G.S. Sampson, and D. Garr. 2021. “The Opportunity Cost of the Conservation Reserve Program: A Kansas Land Example.” Applied Economic Perspectives and Policy 43(2):849–865. https://doi.org/10.1002/aepp.13040
US Government Accountability Office (US GAO). 2017. “USDA’s Environmental Quality Incentives Program Could Be Improved to Optimize Benefits.” GAO-17-225.
US Department of Agriculture Farm Service Agency (USDA-FSA). 2024a. “Conservation Reserve Program Statistics.” Available online: https://www.fsa.usda.gov/programs-and-services/conservation-programs/reports-and-statistics/conservation-reserve-program-statistics/index [Accessed April 24, 2024]
———. 2024b. “USDA to Begin Issuing $2.14 Billion to Agricultural Producers Through Key Conservation and Safety-Net Programs.” Available online: https://www.fsa.usda.gov/news-events/news/10-21-2024/usda-begin-issuing-214-billion-agricultural-producers-key-14
US Department of Agriculture National Agricultural Statistics Service (USDA-NASS). 2024. Quick Stats [database]. Available online: https://quickstats.nass.usda.gov/ [Accessed April 24, 2024]
US Department of Agriculture Natural Resources Conservation Service (USDA-NRCS). 2006. State Soil Geographic (STATSGO) Database. Available online: https://gdg.sc.egov.usda.gov/ [Accessed October 8, 2024].
———. 2014. “Regulatory Impact Analysis (RIA) for the Environmental Quality Incentives Program (EQIP), Amended by the Agricultural Act of 2014.”
