Table 4

Additionality Estimates for the Effect of Nutrient Management Payments on Nutrient Management Practices

Adoption Rates
Nonpayment FieldsAdditionality
Optimal Kernel BandwidthPayment FieldsMatchedNonmatchedEstimate95% Confidence Interval
Practice(1)(2)(3)(4)(5)(6)(7)
Soil Test (% of fields)
Nitrogen0.060.440.240.240.190.040.34
Phosphorus0.010.270.190.160.08–0.070.23
Timing (% of nutrients applied in fall)
Nitrogen0.040.020.110.14–0.09–0.27–0.01
Phosphorus0.010.120.100.220.02–0.290.26
Method (% of nutrients broadcast without incorporation)
Nitrogen0.200.020.220.22–0.20–0.25–0.13
Phosphorus0.030.080.230.27–0.15–0.330.11
Application Rate (lbs. of nutrient)
Nitrogen
  Corn0.02100103122–3.2–37.736.7
  Wheat0.1079646515.32.024.4
  Barley0.205954534.9–13.725.1
  Soybeans0.208.18.68.7–0.5–4.54.3
Phosphorus
  Corn0.0120.836.946.8–16.1–30.52.5
  Wheat0.0522.022.723.3–0.7–8.76.7
  Barley0.2022.519.220.03.3–6.812.2
  Soybeans0.0631.331.233.70.1–9.114.4
  • Note: The table reports mean estimates of the average treatment-effect-on-the-treated from a Gaussian kernel-based propensity score matching estimator with a bandwidth optimally determined under leave-out-one cross-validation. Survey-weighted results are representative of the U.S. fields and are obtained by calculating the weighted average difference in outcomes between treatment and matched control observations with observation-specific survey weights. The 95% confidence interval in determined by bootstrapping the additionality estimates.