Eye on the EPA: Transparency Request #4, Snapshot Approach Toward Economic Analysis Doesn’t Work
April 25, 2013
Lifting the Veil of Secrecy
Following the nomination hearing for Gina McCarthy to lead the U.S. Environmental Protection Agency (EPA), the EPW Republican Senators reiterated five concerns they have with the Agency and the expected responses, in order to ensure that transparency is more than a buzzword. The Senators hope to restore public confidence that EPA will stop undermining public trust behind a veil of secrecy. Below is a detailed explanation of request number four of five. Click here to read all five initial requests.
Transparency Request #4: That written assurances be given the committee that the EPA will conduct cost/benefit analyses as required under various executive orders and as required by the Clean Air Act (CAA), Section 321(a), specifically through issuance of new guidance mandating "whole economy" modeling on major rules.
According to EPA, "Whole economy models are part of the range of tools in an economist's toolkit to examine public policy interventions" and "we have used them at the EPA in selected instances." EPA crafts Regulatory Impact Analyses (RIAs) to support the development of national air pollution regulations. RIAs contain descriptions of the potential social benefits and social costs of a regulation, including those that cannot be monetarily quantified and a determination of the potential net benefits of the rule including an evaluation of the effects that are not monetarily quantified. The RIAs are submitted to the Office of Management and Budget (OMB) for all regulations estimated to cost $100 million a year or more. These regulations are considered "economically significant" and the RIA is to include estimates of the cost and benefits of the rule, and sometimes also report estimates of "employment impacts."
Despite certain Executive Orders (E.O.) and a Congressional mandate requiring the EPA to conduct continuing evaluations of potential job loss and shifts in employment resulting from implementation and enforcement of the CAA, the Agency often conducts only a partial economy analysis of the impact of the rules versus an economy-wide analysis when proposing and finalizing individual rules. A recent review of RIAs issued since 1997 found only two RIAs that contained employment impact estimates utilizing an economy-wide analysis (computable general equilibrium analysis or CGE): the Clean Air Interstate Rule (CAIR)(2005) and Clean Air Visibility Rule (CAVR-BART)(2005).
In 2011, E.O. 13563 called out "job creation" as something to be promoted under the U.S. regulatory system and, during the Obama Administration, EPA has claimed that its new, major, economically significant regulations create jobs. However, often in response to newly issued regulations, industries frequently announce job layoffs.
Where is the disconnect?
In response to E.O. 13563, EPA's RIAs for recent major air rules try to provide employment impact estimates, but these are calculated using a formula designed to guarantee estimates that each new regulation will result in an increase in jobs. Frequently, the formula goes so far as to generate the result that the higher the cost of the regulation, the greater the projected job increase.
Instead of using an economy-wide modeling method called "computable general equilibrium" (CGE), which makes it possible to assess a regulation's effect on the price of energy and therefore the subsequent effect in other sectors not directly subject to the regulation's compliance requirements, EPA utilizes a basic job multiplier formula that boils down to 1.55 jobs per million (formula accounts for what the dollar was worth in 1987) of compliance spending increase. In each RIA, the estimate of the direct compliance cost to the sector being regulated is determined, then restated in 1987 dollars, and multiplied by 1.55. This becomes EPA's job impact of the regulation. This method always finds that the new regulation will create jobs. Taking it a step further, the higher the compliance cost, the more jobs created. Sounds like lazy, unfounded math, right?
This is not an economy-wide analysis
Is this an economy-wide analysis? Is it a partial economy analysis? Is it an economic model? It is an averaging of the weighing of impacts on four sectors based on each sector's compliance costs from 1979-1991. Where does this multiplier come from that EPA applies to the cost of new regulations that it is analyzing in the recent air RIAs? It comes from an econometric study reported in a 2002 paper, Jobs Versus the Environment: An Industry-level Perspective, by Dr. Richard Morgenstern, William Pizer, and Dr. Jhih-Shyang Shih. It considers changes in the total payments to labor in four different industries (pulp and paper, plastics, petroleum, and steel). No measure of the actual numbers of jobs is provided. It is an estimate of how total labor payments change as a result of past environmental spending. It fails to provide changes in the number of employees as EPA's RIA's may suggest. Changes in worker wages/labor payments are summarized by EPA in terms of number of jobs, when these figures are actually job equivalents.
EPA's simple multiplier formula fails to capture the scenario of a regulation's cost being passed to the ultimate end user: the customer. EPA's use of this oversimplified multiplier fails to capture indirect effects - the piece where a majority of the negative impacts occur. An economy-wide model would be a much better, more common sense option.
Utilizing an economy-wide model would take into account a regulation's potential for increasing energy prices, including electricity and gasoline. Many of the recent rules put forward by EPA and those in the pipeline to be proposed are economically significant and affect the energy sector, therefore requiring economy-wide analysis. EPA still has the CGE model used in 2005 and 2006 and, while it may need to be updated, it still exists.
For example, EPA utilized the CGE model when crafting the recent Section 812 report, "Benefits and Cost of the Clean Air Act from 1990 to 2020." Although, EPA chose not to report the labor impact changes, they were among the set of unreported results. EPA conducted two model simulations: one for compliance cost only and one for compliance cost with labor force adjustments and medical cost reductions. As the methods associated with the second model simulation are still only experimental, going forward when employing CGE, a cost-only simulation should always be provided and if alternative cases with adjustments are also completed, EPA should provide two separate runs with labor force adjustments and medical cost reductions run at one time, as well as combined, so that it is possible to understand which adjustment has the greater effect. EPA should also fully document the assumptions made for those adjustments.
Cost of EPA's Lazy Math
A recent joint National Economic Research Associates and U.S. Chamber of Commerce study discovered that an economy-wide analysis of the Utility Mercury and Air Toxics Standard (MATS) rule resulted in a negative impact on worker incomes equivalent to 180,000 to 215, 000 lost jobs in 2015, as well as a negative worker impact persisting at the level of 50,000 to 85,000 job equivalents annually thereafter. Conversely, EPA's simple multiplier formula showed that the regulation would create 46,000 temporary construction jobs and net 8,000 new permanent jobs. This disconnect by the EPA is unacceptable.
The NERA/U.S. Chamber study also found that a full economy-wide analysis consistently found negative impacts on worker incomes from each of the following regulations: MATS; CSAPR; Boiler MACT; and projected Ozone NAAQS at 65ppb. A majority of these negative impacts are associated with sectors outside those directly regulated. Investing in compliance takes away from the ability to invest in growth or expansion projects.
Getting EPA to consistently perform a full economy-wide analysis is the first step towards a truly comprehensive jobs estimate. The next step is to also ask and answer: will the impacts of the regulations result in layoffs or fewer new positions; long term or short term job loss; fewer jobs available overall; lower wages; fewer hours; or outright job elimination.