Figure 8.9 summarizes the ancillary benefits per tonne
of carbon from 15 studies, along with available confidence intervals around
the mid estimate. Multiple entries for a study on the Figure result from modelling
of multiple policy scenarios. Most of the studies focus solely on public health
impacts.
From Figure 8.9, it can be observed that:
- midpoint estimates are mostly less than US$100/tC, but range from less than
US$2 up to almost US$500/tC;
- US estimates are the lowest while estimates from one study for Chile and
several for Norway are the highest (the latter includes a broader range of
benefits);
- significant divergence in estimates occurs across studies for the same country;
and
- uncertainty bounds are quite large for most of the studies that report them.
Figure 8.10 provides ancillary benefits per tonne estimates
related to the size of the carbon tax (in 1996 US$/tC). Points on the diagonal
line AB = MC indicate that marginal private mitigationcosts (MC) equate to the
tax. Some points fall on this line; more appear above it than below, with the
Norwegian/Western Europe and the US studies split. If the damage (benefit) function
is linear, then average benefits equate marginal benefits. Thus, points on the
diagonal imply that the carbon tax is “quasi-optimal” (Dessus and
O’Connor, 1999), in that the tax is optimal without considering either
the direct climate mitigation benefits or any social costs over private costs
(such as deadweight losses from the tax interaction effect). Alternatively,
it can be assumed that the private mitigation cost function is quadratic (Total
Cost=b(X2)), where X is carbon reduction.
In this case, the tax rate equals marginal private mitigation
cost and average private mitigation cost is half marginal private mitigation
cost. The heavy diagonal line equates ancillary benefits to average private
mitigation cost. Points above this line imply there are net benefits to carbon
policy, with the same important caveats as above. More points appear above the
corresponding line (AB=AC) on the graph than above the AB=MC line.
In the general case, a larger carbon tax should lead to progressively smaller
carbon reductions (if the marginal abatement cost curve is upward sloping).
For all but one study (Abt Associates and Pechan-Avanti Group, 1999), the ratio
of ancillary benefits to the tax rate does fall. As for the change in ancillary
benefits per tonne of carbon, Burtraw et al. (1999) show this ratio falling
dramatically in percentage terms with higher carbon taxes. In contrast, Dessus
and O’Connor (1999) show it rising slightly, and in the Abt study the ratio
of benefits to the tax rate rises dramatically (Abt Associates and Pechan-Avanti
Group, 1999). This last result reflects that this analysis treated the SO2
cap as non-binding considerably below the higher tax-rate modelled. In addition,
this study treated the National Ambient Air Quality Standards as a cap, with
abatement below these “caps” treated as benefits, but reductions above
these caps treated as saving abatement costs.