Recent literature has investigated whether the welfare gains from environmental taxation are larger or smaller in a second-best setting than in a first-best setting. This question has mainly been addressed indirectly, by asking whether the second-best optimal environmental tax is higher or lower than the first-best Pigouvian rate. Even this indirect question, though, has itself been approached indirectly, comparing the second-best optimal environmental tax to a proxy for its first-best value, an expression for marginal social damage (MSD). On closer examination, however, MSD becomes ambiguously defined and variable in a second-best setting, making it an unreliable proxy for the first-best Pigouvian rate. With these concerns in mind, the current analysis reevaluates the central welfare question both directly and indirectly and finds that when compared directly to its first-best Pigouvian value, the second-best optimal environmental tax generally rises with increased revenue requirements. Even in cases where the second-best optimal environmental tax is lower than its first-best value, the welfare gains may be greater than in a first-best setting. These results suggest that the marginal fiscal benefit (revenue recycling effect) exceeds the marginal fiscal cost (tax base effect) over a range of environmental tax rates that, for benchmark models, extends above the first-best Pigouvian rate. Results in the tax interaction literature are fully consistent with these interpretations once the effects of normalizations and numeraires are fully recognized. These findings reinforce the intuition that environmental policy complements rather than competes with the provision of other public goods.
