The magic of the EPA's benefit/cost analysis
© The Hill

Benefit/cost analysis: It sounds so scientific, so rational, so impartial. So sound as a tool with which to resolve conflicting assertions about the wisdom of regulatory proposals. So divorced from partisanship or ideological influence.


Oh, please. Democracy is the art of wealth redistribution, the bureaucracy is an interest group and benefit/cost analyses can be shaped to support virtually any outcome preferred politically. Consider, for example, the Environmental Protection Agency (EPA) efficiency rule for medium and heavy trucks, part of the administration's climate action plan; the benefit/cost "analysis" used to justify it is hugely revealing. The EPA has published estimates of the effects of the rule, as follows:

The results of the analysis, summarized in Table VII-37, demonstrate that relative to the reference case, by 2100 ... global mean temperature is estimated to be reduced by 0.0026 to 0.0065 °C, and sea-level rise is projected to be reduced by approximately 0.023 to 0.057 cm.

The EPA then states that "the projected reductions in atmospheric CO2, global mean temperature, sea level rise, and ocean pH are meaningful in the context of this action." And so we arrive at the benefit/cost conclusion, given in all seriousness:

[We] estimate that the proposed standards would result in net economic benefits exceeding $100 billion, making this a highly beneficial rule.

Can anyone believe that a temperature effect by 2100 measured in ten-thousandths of a degree, or sea-level effects measured in thousandths of a centimeter, could yield over $100 billion in net economic benefits? How is that possible?

Welcome to the fascinating world of EPA benefit/cost analysis. Before this truck rule and the other components of the Obama climate policy were promulgated, the administration conducted an "analysis" of the "social cost of carbon" (SCC), in order to generate an estimate of the marginal externality cost of greenhouse gas emissions (GHG). The problems with that analysis are legion, but the central ones are the use of global (rather than national) benefits to drive the benefit/cost comparison; the failure to apply a 7 percent discount rate to the streams of benefits and costs, despite clear direction from the Office of Management and Budget; and — most important — the use of ozone and particulate reductions as "co-benefits" of climate policies. The administration's estimate is about $36 per ton in 2015 ($31 per ton in 2010).

And that is how a regulation yielding future changes in temperatures and sea levels approaching zero can be claimed to yield net benefits "exceeding $100 billion, making this a highly beneficial rule." In the EPA's benefit/cost framework, the actual effects of the policies literally are irrelevant; just compute the assumed reduction in GHG emissions, multiply by $36, and voila!

Should you respond that any journey begins with a tiny step, consider the following. The Obama climate policy calls for a 17 percent reduction below 2005 levels in U.S. GHG emissions by 2020. If we apply the EPA's own climate model, that would reduce temperatures by the year 2100 by fifteen one-thousandths of a degree. In addition, the U.S.-China Joint Announcement on Climate Change calls for an additional 10 percent reduction by the U.S. by 2025. (The Chinese commitment — it's unenforceable — is for a "peak" in their emissions by 2030. Total emissions at that peak and the emissions path after the peak are not addressed.) The additional 10 percent U.S. reduction yields another one one-hundredth of a degree. The standard deviation of the temperature record is about 0.1 degrees, so that these effects would be too small even to be measured, let alone to affect sea levels and cyclones and all the rest.

If we add an additional 20 percent emissions cut by China by 2030, that adds 0.2 degrees; and another 0.2 degrees if we assume a 30 percent emissions cut by the rest of the industrialized world, by 2030. If we assume also a 20 percent reduction by the less-developed world by 2030, temperatures would be reduced by another one-tenth of a degree.

The grand total: a bit more than 0.5 degrees.

And these model predictions use underlying parameters highly favorable to the policies under examination; that is, assumptions that increase the predicted effects of the policies. The most important is a "climate sensitivity" (the temperature effect in 2100 of a doubling of GHG concentrations) assumption of 4.5 degrees, a number 50 percent greater than the median reported by the Intergovernmental Panel on Climate Change in its latest assessment report. And even the latter is about 40 percent higher than the median of the estimates published in the recent peer-reviewed literature.

For obvious reasons, these trivial temperature benefits of "climate" policies have not been publicized extensively. It is the delegation of legislative powers to the regulatory agencies that has allowed such game-playing in pursuit of an ideological agenda. The only means with which to restore political accountability to the regulatory process is a requirement that all regulations be approved by Congress.

Zycher is the John G. Searle scholar at the American Enterprise Institute.