This paper develops an analytical framework for estimating the marginal excess burden (MEB) of taxes on labor, gasoline, cigarettes and alcohol, allowing for externalities and interactions between the different taxes. The formulas are estimated using plausible parameter values for the United Kingdom. Given the uncertainty over various elasticities and external damages, we obtain wide ranges of possible outcomes for the MEBs. By performing Monte Carlo simulations, however, we can assess the likelihood that the MEB of one tax exceeds that of other taxes.
We find that the MEB of labor taxes lies between 0.18 and 0.34 with 80% probability for tax increases used to finance transfer spending. The MEB for the gasoline tax is much larger: it is more than double that of the labor tax in 75% of our simulations and more than treble in 51%. Similar results apply for the cigarette tax. Even though these goods are relatively weak leisure substitutes, this is more than offset by large incremental welfare losses in the commodity markets, because the commodity tax rates are substantially higher than estimated marginal external damages in most of our scenarios. In contrast, our central estimate for the MEB of alcohol taxes is similar to that for labor taxes, because the alcohol tax is much closer to our assumed values for marginal external costs. But the MEB is still positive, even in scenarios when the alcohol tax is below marginal external damages, due to the impact of the tax on exacerbating the labor market distortion. When additional government spending is on public goods rather than transfers, the MEB is significantly lower for the labor tax but less so for commodity taxes.
In the United Kingdom context, our results suggest the possibility of significant social welfare gains from tax reforms that shift some of the burden of taxation off gasoline and cigarettes and onto labor. The methodology could be readily extended and applied to tax systems in other countries.