Voluntary environmental programs (VEPs) are designed based on a win–win approach to environmental protection that reconciles environmental protection and economic performance. Despite the claims about VEPs, there has been an ongoing debate over their efficacy with regard to whether environmental goals are balanced by economic interests on both theoretical and empirical grounds. To resolve this controversy, this paper empirically investigates a public VEP by the US Environmental Protection Agency: Green Lights (GL). For this, the paper constructs a treatment effects regression model to account for the effects of non‐random assignment for GL participants and non‐participants. The proposed model can simultaneously estimate probit models that predict corporate participation in the GL program and linear models that test the extent to which this participation contributes to economic performance. The results indicate significant positive effects of corporate participation in the GL program on economic performance, providing support for the win–win perspective. Copyright © 2013 John Wiley & Sons, Ltd and ERP Environment