matter, PM), air toxics (e.g., mercury), and other environmental concerns that could require
substantial additional capital investment at coal-fired power plants. These ancillary regulations
are likely to affect many inefficient coal-fired plants that will also be a target of regulations
to reduce GHG emissions. The substantial investments required to comply with these
rules, coupled with EPA efforts to regulate carbon emissions, leave the industry with a difficult
choice: risk the ‘‘stranding’’ of investments by future carbon regulations or retire a substantial
amount of coal-fired capacity. Thus, these other regulations could achieve substantial collateral
reductions in GHG emissions from stationary sources even without specific EPA GHG
rulemaking under the CAA.
Comparing CAA Regulation with Alternatives
At the Copenhagen climate meetings in December 2009, PresidentObama pledged the United
States to emissions reductions ‘‘in the range of ’’ 17 percent below 2005 levels by 2020. At the
time, the president rested hopes for meeting this goal on comprehensive cap-and-trade
legislation. TheWaxman–Markey legislation that passed the House in June 2009 set a domestic
target of 17 percent reductions by 2020, with an additional 3 percent anticipated from
international forestry projects. The legislation proposed by Kerry–Boxer in the Senate set
a domestic target of 20 percent reductions. Before comparing the level of reductions that
might be possible under the CAA with what was expected to occur under these bills, it is
useful to examine the nature of reductions that would have been achieved through legislation.
The Legislative Baseline
We use the Waxman–Markey bill as the legislative baseline because it is the most studied
example of climate legislation and because it has passed one chamber of Congress. As shown
in Figure 4, which is based on EIA (2009) estimates, the expected emissions reductions under
the bill approach 33 percent from 2005 levels by 2020. However, these reductions include
offsets, which are emissions reductions occurring outside the regulated sector. Of the 33 percent
indicated in Figure 4, 5 percent was expected to come from domestic offsets and more
than 15 percent was expected to come from international offsets—a policy tool unlikely to be
available under the CAA (Richardson, 2010). Only 10 percent of the reductions was expected
to come from regulated domestic sources.12
It is also important to note that banking of allowances clouds the assessment of emissions
reductions under the Waxman–Markey legislation. Banked allowances can be used to cover
emissions in excess of the cap at a later time; hence, while banking reduces short-term emissions,
it has no effect on long-term emissions unless emitters keep a perpetual reserve or
allowances are later devalued or confiscated by regulators. The EIA (2009) model predicted
that allowance banking would result in large emissions reductions beyond the 17 percent cap
level in 2020, as illustrated in the figure. Thus, if one wishes to ignore contributions of emissions
allowances to the bank, then one would scale back the total emissions, including the
12The model also did not account for the endogenous opportunities for efficiency improvements listed in
Table 1. Hence it may overstate the expected cost or understate the magnitude of emissions reductions that
could be achieved under cap and trade.
Greenhouse Gas Regulation under the Clean Air Act 13
D