Voluntary Pollution Abatement as an Indirect Source of Emission Spillovers

• In Event: Environmental Policies at Different Scales, With Overlaps and Confounds: Analytic Solutions?

Abstract

We examine the potential for intra-firm pollution leakage when some but not all facilities in a firm choose to lower their emissions by participating in a voluntary pollution abatement program.  From the perspective of a non-participating facility, a sister facility’s voluntary pollution abatement may result in firm-wide reputational spillovers leading to reduced regulatory and/or public pressure to curb emissions, thus lowering its marginal emissions cost (MEC); on the other hand, information and technology spillovers within a firm decrease its marginal abatement cost. (MAC). Ultimately, the direction and magnitude of the change in the non-participating facility’s emissions depends on the relative strengths of these two countervailing spillovers.  

 We focus on the EPA’s “Pollution Prevention” (P2) program, analyzing emission spillovers in over 8,100 treated TRI (Toxics Release Inventory) facilities out of 23,800 facilities in total that appear in our analysis sample between 2000-2022.  Following Rijal and Khanna (2020), we define a treated facility as having at least one sister facility participating in P2, and a control facility as one that has never had any sister facilities participate in P2.  To account for the endogenous assignment of treatment, we employ a counterfactual/imputation estimator (Athey et al., 2021; Liu et al., 2022) that allows for treatment status to be switched on and off with limited carryover effects, while avoiding the negative weights problem and biases caused by staggered treatment and treatment effect heterogeneity (Goodman-Bacon, 2021). The estimator treats observations under the treatment condition as missing and uses data under the control condition to impute counterfactuals of treated observations, which are then used to estimate the average treatment effect on the treated. To accommodate variations in regulatory scrutiny, we control for each facility’s inspection history, total penalties paid, number of informal actions, the number of significant violations in the last 3 years, as well as other factors such as facility age, state LCV score and county NAAQS non-attainment status.

 We find evidence of pollution leakage under the P2 program: on average, non-P2 participating facilities increase their total TRI emissions by about 11% when at least one of their sister facilities was concurrently participating in P2.  The bulk of the increase occurs in the first year of treatment and, on average, the effect disappears over 3-4 years. Moreover, facilities that remain in treatment for 4+ years eventually decrease their emissions by more than 16% as compared to their pre-treatment levels suggesting that the decrease in a treated facility’s MEC from firm-wide reputational spillovers is slowly outweighed by the decrease in its MAC via within-firm information spillovers. Overall, our results reveal a nuanced picture of the short and long term environmental benefits of voluntary pollution abatement programs.

 In on-going work we are exploring heterogeneities in the magnitude of pollution leakage by source reduction type and scale of program participation.

 Keywords: pollution leakage, Pollution Prevention, self-regulation

 JEL codes:  Q52, Q53, Q59

Author

• Binish Rijal, University of Richmond
  Presenting Author