22
Despite our main focus on the discussion of (28) one needs to keep in mind that according to
(29) the efficient tax scheme A requires to levy a tax on all polluting residuals. It doesn't fol-
low from setting teA ≠ Ye as prescribed by (28) that no emissions tax other than te is needed to
achieve efficiency in case that the production residuals are not the only pollutants. For that
scenario our analysis provides an important lesson: If all polluting residuals are taxed in an
effort to correct for the allocative distortion, it is not efficient, in general, to set the tax rate of
the production residuals equal to their marginal abatement cost, conventionally defined.
Turning to the interpretation of tax scheme B we observe that B is also capable of restoring
efficiency in the market economy and does so without any tax on abatement residuals. This
scheme is a particularly interesting option for efficient pollution control if abatement residuals
are difficult and costly to monitor and therefore cannot readily be used as a tax base. How-
ever, avoiding taxes on abatement residuals comes at the price of taxing labor and material (in
addition to post-consumption and production residuals).16 Since the signs of the derivatives
RI and Rjm for j = 1, 2 are ambiguous, it is not clear, whether tm and 11 are subsidies or
taxes (proper). At any rate, securing efficiency by means of tax scheme B requires to drive tax
wedges between demand prices and supply prices on all three markets. Taking a closer look at
te under tax scheme B shows that
Ui
-∑1 ~T ( Xe +Σ,XR ) = tBB = Ye - 'Y . (32)
iUi ejraj eeece
To the left of the first equality sign in (32) we have the sum of the direct (positive) and indi-
rect (negative) marginal benefits of a small reduction in the emission of production residuals.
The indirect benefit is, in fact, the marginal environmental damage from the emission of
abatement residuals caused by stepping up abatement. Correspondingly, the far right side of
(32) represents the social marginal abatement cost which is the same as in tax scheme A for
the case that Xra1=Xra2=0. Since teB < Ye, it is optimal to abate more production residuals
than under the conventional Pigouvian tax rule.
5. Concluding remarks
16 One may wonder why there isn't a third efficiency restoring tax scheme, that taxes also labor and material
(like B) but in which te captures the impact of all kinds of residuals. After all, due to rc = y, post-consumption
residuals are generated uno actu with all other outputs. However, when we set ta1= ta2=tc=0 in (26) there is
no way to find values for ti, tm and te such that market equilibria turn out to be efficient.