Optimal emission reduction in a cap-and-trade regime estimation of a finite horizon dynamic game of imperfect competition

This paper estimates a finite horizon dynamic game to study how firms make strategic decisions on emission reduction in a cap-and-trade regime, and to examine the application and implication of the appropriate government policies. The environment consists of power plants in a competitive electricity market--an industry that is the biggest emitter in the economy. Plants also participate in the cap-and-trade market of NOx emission allowances--one of the longer running, most geographically extensive and most successful programs. In the model, after observing the current adoption state of all plants and forming an expectation of all plants' production and adoption decisions, each plant decides when to adopt a pollution abatement equipment with a substantial sunk cost. The adoption will be realized in the next period. After weighing its adoption decision if it has not adopted, a plant then decides how much electricity to produce, taking allowance prices as exogenous, since the permit market covers a much larger area geographically. Estimates of the adoption cost over time, the evolution of every plant's production and adoption decisions are covered. Results from our counterfactual experiments indicate that (1) every dollar spent on subsidizing the adoption cost in a permit trading system most likely produces a positive return on health benefits; (2) a high tax alone could reduce more emissions than the current permit system, but a low tax would be ineffective; (3) subsidies in a tax system mostly likely does not produce a positive return.