When I was a kid, I liked historic towns where guides would explain how coopers, blacksmiths, and cooks used now-forgotten tools to provide the town’s basic needs. As technology advanced, many of those tools faded from use. Today, they decorate the walls of antique stores, B&Bs, and restaurants. Such will be the fate of agency benefit-cost analysis which unfortunately may soon be hung on the walls of B&Bs in Cambridge and in other university towns if current trends continue. This obsolescence is occurring not due to technical advancement, but due to the tool’s lack of useful information.

Take the U.S. Environmental Protection Agency’s (EPA) recent final National Ambient Air Quality Standard (NAAQS) for fine particulate matter.[1] Putting aside the relative merits of EPA’s selected standard, let’s examine whether the benefit-cost analysis informs policy officials and the public about the trade-offs for the different policy options. What does EPA’s benefit-cost analysis consider and doesn’t it consider? EPA’s scope of analysis aims to measure the social benefits and social costs to bring the parts of the country out of attainment with the new standard into attainment. That’s good. However, the way EPA estimates this scope is pure fantasy.

Consider EPA’s baseline assumptions for southern California. First, EPA assumes that the region is in compliance with the most recent 2012 fine particulate matter NAAQS. Southern California counties have never been in compliance with any EPA fine particulate matter NAAQS since EPA issued the first one in 1997. In EPA’s previous, 2012 NAAQS update, EPA’s RIA could not find enough available technologies for southern California to come into compliance with the 2012 NAAQS. It is not a reasonable or a realistic projection to assume that suddenly, in a few years, the region will come into attainment. As discussed more below, human civilization seems to preclude these areas ever from being in compliance.

EPA then assumes that all its recent and planned rules that reduce particle matter emissions are in place and effective. In reality, at the time EPA issued the final NAAQS, several of these key rulemakings assumed will be implemented have been enjoined by courts.[2] Moreover, regardless of the Administration, courts have struck down or remanded many key criteria air pollution regulations and permitting procedures – often multiple times – delaying and reducing emission reductions.[3] It is not a realistic or a reasonable projection of the future to assume all EPA rules will come into force as promulgated.

Even with this baseline that assumes huge and unrealistic emission reductions have already happened before the new NAAQS mandates, the agency cannot even imagine how to reach attainment. The agency considers the additional costs California regions could take to reduce emissions. The agency caps the cost of possible emission reduction measures at $160,000 per ton. The RIA helpfully states: “We selected the $160,000/ton marginal cost threshold because it is around that cost level when…(ii) opportunities for additional emission reductions diminish.”[4]  Unfortunately for California, with this cap, EPA only finds 26 percent of the required emission reductions for attainment. Further reductions would fundamentally alter society and conflict with many other consumer and public policy needs – such as home heating, reliable electricity service, and many others.

Imagine if the RIA took a comparable unrealistic approach to compliance costs as it does for the other assumptions. What would it cost for compliance? Spot trades reported by SCAQMD for the right to emit one ton per year of particulate matter last traded eight years ago for $190 million to $205 million.[5] With those values as marginal costs, attainment costs for just California would range from $2.8 trillion to $2.9 trillion, or about 10 percent of annual US output. Of course, the actual marginal costs will be larger than the 2013-2016 reported prices as emission reduction potentials become scarcer. Recall also that tons required start from a baseline and are much closer to attainment that actual levels – in other words, the marginal costs to reach the final NAAQS from today’s actual levels is much greater. Would society realistically spend $50 trillion (or $100 trillion to be less conservative) on one public policy goal at the expense of so many others?

EPA’s cost benefit analysis therefore does not analyze reality or a realistic future. Policy officials and the public do not get the true choices and trade-offs we face to reduce this pollutant. More troubling, EPA does not analyze the more likely and immediate social benefits and social costs from the new NAAQS.

Specifically, the analysis failed to consider the immediate and the real costs in all other parts of the country. Proposed projects in attainment areas must conduct an analysis to ensure that the project will not jeopardize the area’s attainment status with the new NAAQS. Mathematically, the more stringent the standard, the less “room” available to allow new emission sources and to stay in attainment. It is reasonable to estimate additional compliance costs for new projects throughout the United States, especially in those areas with particulate matter levels close to the standard. Philadelphia, Seattle, parts of the Atlanta and other metro areas are close to the final NAAQS and would have difficulty granting permits for new emission sources.

Plenty of research finds we should expect firms to shift production due to Clean Air Act permitting costs and uncertainty.[6] Suppose Clean Air Act permitting uncertainty drives firms to shift $1 billion in planned, near-term capital investment to Canada, Mexico, or overseas. This amount could equal one electric battery component plant, some clean energy manufacturing, or a bunch of smaller projects when firms with plants in the Philadelphia or Seattle areas decide to shift their operations to sister plants in other countries. Assuming a rate of return on capital of 22 percent, the U.S. loses a stream of return of $220 million per year. This representative opportunity cost does not include the additional direct costs that all permittees throughout the country will spend in the next few years on modeling and other analyses as part of the Clean Air Act permitting process. These expected costs (and associated avoided pollution benefit) have strong methodological foundation and can realistically estimated from today’s data. Many projects have been announced in the 117 counties close to the final NAAQS value. EPA could estimate whether a $1 billion shift is likely too high or too low. However, the RIA does not even consider these social benefits or social costs.

In sum, EPA’s analysis considered scenarios that will never happen and failed to consider scenarios very likely to happen. The analysis not only is a waste of public resources, but also is a failure to support our democratic processes by educating the public with the best available economic methodologies. If they ever build a Cracker Barrell restaurant in Cambridge, this useless analytic methodology deserves a place on the wall next to the washboard and the wagon wheel.

[1] U.S. Environmental Protection Agency, “Reconsideration of the National Ambient Air Quality Standards for Particulate Matter,” 89 FR 16202, March 2024.

[2] Ohio, et al., Applicants v. Environmental Protection Agency, et al.

[3] West Virginia, et al. v. Environmental Protection Agency, et al., June 2022.

[4] U.S. EPA, Final Regulatory Impact Analysis for the Reconsideration of the National Ambient Air Quality Standard for Particulate Matter, EPA-452/R-24-006, January 2024, p.11.

[5] See: https://www.aqmd.gov/home/permits/emission-reduction-credits/historical-active-erc-and-sterc-lists

[6] See, for example, GREENSTONE, M. (2002). The Impacts of Environmental Regulations on Industrial Activity: Evidence from the 1970 and 1977 Clean Air Act Amendments and the Census of Manufactures. Journal of Political Economy, 110 (6), 1175–1219