In a parsimonious two-sector general equilibrium model, we challenge the widely-held tenet that within a cap-and-trade system renewable energy policies have no effect on carbon emissions. If the cap does not capture all sectors, we demonstrate that variations of a renewable energy subsidy change aggregate carbon emissions through an inter-industry leakage effect. We decompose this effect into intuitively intelligible components that depend in natural ways on measurable elasticity parameters. Raising the subsidy always reduces emissions if funded by a lump-sum tax, reinforcing recent findings that tightening environmental regulation can cause negative leakage. However, if the subsidy is funded by a levy on electricity, it can increase emissions. These results provide a valuable basis for an informed design of renewable energy policies and an accurate assessment of their effectiveness. We highlight how a state-of-the-art statistic used by governments to gauge such effectiveness, “virtual emission reductions”, is biased, because inter-industrial leakage effects are not captured.
In a parsimonious two-sector general equilibrium model, we challenge the widely-held tenet that within a cap-and-trade system renewable energy policies have no effect on carbon emissions. If the cap does not capture all sectors, we demonstrate that variations of a renewable energy subsidy change aggregate carbon emissions through an inter-industry leakage effect. We decompose this effect into intuitively intelligible components that depend in natural ways on measurable elasticity parameters. Raising the subsidy always reduces emissions if funded by a lump-sum tax, reinforcing recent findings that tightening environmental regulation can cause negative leakage. However, if the subsidy is funded by a levy on electricity, it can increase emissions. These results provide a valuable basis for an informed design of renewable energy policies and an accurate assessment of their effectiveness. We highlight how a state-of-the-art statistic used by governments to gauge such effectiveness, “virtual emission reductions”, is biased, because inter-industrial leakage effects are not captured.