CLIMATE POLICY BY BALANCING COSTS AND BENEFITS - The Climate Casino - page 206

COST-BENEFIT ANALYSIS APPLIED TO CLIMATE CHANGE
I now apply the cost-benefi t approach to evaluate different targets
for slowing climate change. I do this in a simplifi ed way that asks which
temperature objective minimizes the sum of abatement investments
and damages from climate change. This is the same as choosing a policy
that maximizes net benefi ts.
I do this by putting abatement costs and climate damages together
in a single graph, as shown in Figure 29 (as well as Figures 30 through
32). This important fi gure requires some explanation because it com-
bines several elements.
The idea is to examine the costs, damages, and net impact for differ-
ent climate objectives. I take targets of 2°C, 3°C, 4°C, and so forth. For
each objective, I calculate the abatement costs required to keep the
global temperature increase below the selected target; this cost curve is
the downward-sloping line. I calculate the damages from climate change
at that temperature target; this is the upward-sloping line. The costs
and damages are added together to arrive at the total costs, shown as
the U-shaped curve. Each curve is plotted as a ratio of the costs divided
by total global income on an annualized basis.
Each of these curves has been discussed in earlier chapters (12 and
15), so here I am simply putting them together. For example, the costs
of using emissions reductions to meet each of the temperature targets
was shown in Figure 26. Similarly, the damage curve uses the estimates
summarized in Figure 22. Note that these estimates do not include all
the costs of tipping points or of the less easily quantifi ed impacts, such
as ocean acidifi cation. 2 They also simplify by omitting the dynamics of
adjustment. Figures 29 through 32 all use the same graph to show the
damages and costs under four scenarios.
Begin with an economic analysis of effi cient policies without dis-
counting in Figure 29. This means that costs and benefi ts are calcu-
lated as if they occur in the same year. While this is a polar extreme of
discount rates (and an approach that I defi nitely do not recommend for
realistic analyses), it has the virtue of transparency. I further assume