California’s Cap-and-Trade Program and Emission Leakage in the Western Interconnection: Comparing Econometric and Partial Equilibrium Model Estimates

We compare the effects of California's AB 32 cap-and-trade program on leakage in the electricity sector using two methods: a simulation-based partial equilibrium model that accounts for details of policy implementation and is parameterized using market data and an econometric model applying a quasi-experimental design with matching and a robust inference method that does not require parallel trends to hold exactly. Based on the estimated shifts in electricity generation, we infer CO2 emission leakage predictions in 2013 and 2016. The comparison allows us to identify critical assumptions driving the simulation results and to benchmark the empirical results in a complex policy setting where threats to identification undermine attempts at statistical inference. Over the study period, we find significant leakage potential ex ante and empirical evidence that is consistent with some resource shuffling ex post. Limiting the ability of electricity importers to claim the default emission factor may reduce leakage risks.