Discussion Paper

Alternative Climate Policies and Intertemporal Emissions Leakage: Quantifying the Green Paradox


Efforts to limit cumulative emissions over the next century may be partially thwarted by the responses of fossil fuel suppliers. Current price-cost margins for major reserves are ample, leaving scope for significant price reductions if climate policies reduce demand for fossil fuels through conservation or substitution to clean alternatives. Most models simulating the consequences of climate policies completely disregard these supply responses. As for theoretical models, under standard assumptions they predict such strong supplier responses that climate policies may have no effect on cumulative emissions and may even leave society worse off, suffering damages from global warming sooner and with less time to adapt (the “green paradox”).We contribute to this literature by developing a richer theoretical model that takes account of the different extraction costs and emissons rates of different fossil reserves. We use this model to compare the qualitative effects of four policy options—accelerating cost reductions in the clean backstop technologies, taxing emissions, improving energy efficiency, and a clean fuel blend mandate. We also discuss the consequences of mandating carbon capture and sequestration. All policies can reduce cumulative emissions, but the backstop policy accelerates emissions while conservation policies (energy efficiency or blend mandates) delay emissions. We then calibrate the model using data on costs, reserves, and emissions factors for five major categories of oil. Using this calibrated model, we estimate the interemporal leakage rate—the percentage error in cumulative emissions reductions that would arise if no account is taken of the supply responses of oil producers. We find that conservation policies can have higher intertemporal leakage rates and backstop policies can have lower leakage than an emissions tax. Leakage rates generally decline as the policies become more stringent.