Twenty-two years ago, in Whitman v. American Trucking Ass’ns,1 the U.S. Supreme Court held that the U.S. Environmental Protection Agency (EPA) could not consider costs when setting National Ambient Air Quality Standards (NAAQS) under the Clean Air Act. Eight years later, in Entergy Corp. v. Riverkeeper Inc.,2 the Court upheld the EPA’s use of cost–benefit analysis to set performance standards reflecting “the best technology available for minimizing adverse environmental impact”3 as a permissible interpretation of the Clean Water Act. Then, in Michigan v. EPA,4 the Court held that EPA was required to consider costs when deciding whether it was “appropriate” to regulate emissions of mercury and other hazardous air pollutants from coal-fired power plants under the Clean Air Act. While the statutory language was different in each case, the trend in these decisions signals a Court increasingly favorable to the use of cost–benefit analysis in rulemaking. Indeed, several commentators have concluded that, after Michigan, agencies must conduct cost–benefit analysis in a wide variety of statutory contexts.5

At the same time, courts have been restricting governments’ ability to take explicit account of race when attempting to redress disparities in the distribution of benefits or harms.6 For example, in 2021, four different district courts enjoined a federal loan forgiveness program for “socially disadvantaged farmers and ranchers”—a category defined in part by racial or ethnic identity. The courts concluded that plaintiffs were likely to prevail on the merits of their claims that the program violated the equal protection component of the Due Process Clause of the Fifth Amendment.7 In the current term, the Court heard arguments in November in two cases regarding race-conscious college admissions practices;8 after the arguments, most observers felt that the Court appeared poised to overrule Grutter v. Bollinger9 by holding that such practices are unconstitutional.10

Daniel Farber’s wide-ranging and timely article addresses, among other things, the consequences of these two trends colliding. If cost–benefit analysis is ascendant, how can agencies take into account distributional concerns? Moreover, how can agencies address environmental justice if they cannot explicitly consider race?11 In the last section of his article, Farber suggests a possible solution to these dilemmas: because both exposure and vulnerability to pollution vary geographically and are strongly correlated with both race and income, “a heightened focus on differences in exposure and vulnerability offers the most promising path forward for environmental justice.”12

This comment makes three observations regarding Farber’s proposal: (a) he is perhaps unduly pessimistic about the feasibility of implementing it, given recent advances and innovations in air quality monitoring; (b) although Farber focuses on how to incorporate equity into rulemaking, such approaches can go only so far in addressing environmental inequities, which are more often a matter of site-specific decision-making; and (c) Farber’s proposal addresses only one aspect environmental justice—distributive outcomes—while ignoring the expressive consequences of his proposal for procedural justice.

Farber suggests that one barrier to implementing his proposal is that the EPA lacks sufficiently fine-grained data on variations in pollution levels.13 As he points out, the impacts of pollution from sources like heavily traveled roadways are concentrated in neighborhoods very close to those sources, and the existing regulatory monitoring network cannot detect these variations.14

Farber is undoubtedly correct that the existing air monitoring network is not well-suited to detecting fine-grained differences in pollution exposure, either geographic or temporal. For example, only 651 of 3,100 counties in the country have PM2.5 monitors;15 as a result, roughly 120 million Americans live in counties where PM2.5 is not monitored by EPA.16 Furthermore, there is evidence that monitors in counties with higher percentages of minority and low-income residents are more likely to be located in such a way that they miss higher pollution levels in the region, thus leading to potential misclassification as being in attainment with air quality standards.17 In addition, many particulate-matter and lead monitors operate on a 1-in-3-days or 1-in-6-days sampling schedule,18 so they can miss transient pollution events.19 Intermittent sampling is also subject to manipulation; a 2021 study found that pollution levels were on average 1.6% lower during the on days of 1-in-6-days intermittent monitoring schedules compared to the off days.20

EPA has made some efforts to address these shortcomings, for example by introducing near-road monitoring requirements for NOx, CO, and PM2.5.21 But regulatory monitors are simply too expensive ever to provide comprehensive coverage of small-scale differences in pollution levels.22

Yet several promising technologies and practices provide opportunities for significant improvements. The first is satellite-derived data. Satellites can measure levels of nitrogen dioxide, sulfur dioxide, ammonia, carbon monoxide, certain volatile organic compounds, and (through a more complex process) PM2.5.23 They are also achieving ever-greater spatial and temporal resolution. For example, the NASA Tropospheric Emissions: Monitoring of Pollution (TEMPO) satellite, which launched on April 7, 2023, will take hourly measurements of ozone, nitrogen dioxide, and other pollutants at a high-resolution (sub-urban) scale.24

Satellite-derived data have strengths and limitations that are complementary to those of the existing ground monitoring network. On the one hand, because they offer comprehensive geographical coverage, satellites can fill in temporal and geographic gaps in the regulatory monitoring network. On the other hand, because satellites measure the overall density of molecules of a gas between the satellite and the Earth’s surface, they can be blocked by clouds and they cannot identify the vertical distribution of the pollution.25 Ground monitors can help address these limitations by providing a baseline for validating satellite-derived data.26 A 2022 study is an example of satellites’ power: it used satellite data and modeling to reveal significant geographic variations in PM2.5 exposure that were not captured by ground monitors in Los Angeles and Sacramento.27 In short, satellites can produce high-resolution data at hyperlocal scales, helping to identify pollution hotspots and disproportionately pollution-burdened communities.

Low-cost air sensors are another promising set of technologies.28 The term “low-cost sensors” is used to describe sensors that cost tens or hundreds of dollars, as opposed to thousands or tens of thousands of dollars for regulatory monitors.29 They are designed to take measurements at high frequencies and, because of their low cost, can potentially be deployed in large numbers.30 Many low-cost sensors can report data at one-second intervals,31 and some, such as the PurpleAir network, already provide real-time data for much of the United States.32

The accuracy and reliability of air sensors vary widely, and questions have been raised about their suitability for regulatory monitoring.33 However, the technologies are rapidly improving, and even the current versions can be important supplements to regulatory monitors. For example, a 2020 study found that a model for predicting PM2.5 levels at a specific location was more accurate if it was based on both monitor and sensor data than if it was based on monitor data alone.34 Both satellites and air sensors therefore provide the opportunity to carry out ambient air monitoring for at least some pollutants at the micro scale that Farber’s proposal requires.

In addition to these new technologies, a variety of monitoring systems—including community monitoring, mobile monitoring, and fenceline monitoring—can help to both improve ambient monitoring and determine the impacts of specific pollution sources. As mentioned above, Congress recently allocated funding to community air monitoring.35 In 2017, California began developing a community monitoring network through Assembly Bill 617.36 A few states have also started to use mobile units—vans or trailers outfitted with monitoring equipment.37 Mobile air monitoring can be used to track down sources of pollution and migrating plumes,38 respond to public air quality complaints,39 and fill in when stationary monitors are not operating. The Texas Commission on Environmental Quality (TCEQ) developed its van fleet primarily to improve air monitoring after emergencies during which the stationary monitoring network is out of service.40 Finally, fenceline monitoring can be an important supplement to smokestack monitoring for facilities with multiple potential sources of fugitive emissions, such as leaks.41 The EPA mandated fenceline monitoring for benzene at oil refineries in a 2015 rule.42 Four air districts in California have expanded on this requirement by implementing multi-pollutant fenceline monitoring at oil refineries.43 In addition, the EPA recently proposed requiring oil and gas production facilities to adopt a similar approach to detect methane leaks and to allow community monitoring to prompt corrective measures at these facilities.44

Although the EPA has broad authority to require monitoring of air quality and emissions from specific sources,45 integrating the technologies and practices described above into some aspects of EPA’s regulatory programs is nevertheless challenging. To take just one example, EPA regulations governing Federal Equivalent Methods (FEM)—which are supposed to provide some flexibility in adopting new attainment monitoring approaches that are less costly than the Federal Reference Method monitors—assume that a FEM is a physical object whose functioning can be tested inside a chamber.46 This requirement is not compatible with satellites, collections of multiple air sensors, or hybrid models. The use of these technologies for assessing an area’s compliance with the NAAQS would therefore require regulatory, and potentially statutory, change.47

In sum, a variety of technologies and practices show great promise for fulfilling a factual prerequisite for Farber’s proposed solution: fine-grained data of local variations in pollution concentrations. Given the pace of technological innovation in this area and the amount of funding the federal government has recently directed toward it, it seems feasible for agencies to incorporate hyperlocal exposure data into their decision-making in at least some contexts. In particular, it could be incorporated into individual siting decisions or when determining emissions standards for individual facilities. Nevertheless, as described above, there remain challenges with integrating these data sources into at least some parts of the Clean Air Act regulatory regime.

The focus of Farber’s article is on how to reconcile the consideration of equity with a rulemaking process dominated by cost–benefit analysis. He says less about what it means in practice for the design of an environmental regulation to incorporate equity considerations.48 This is not necessarily a straightforward matter. When setting an ambient air quality requirement (such as the NAAQS) or establishing performance standards for a category of sources (such as Maximum Achievable Control Technology standards for hazardous air pollutants or New Source Performance Standards for other regulated pollutants), one option is simply to set a more stringent standard that will lower pollution exposure for everybody. But this is not a particularly targeted way to reduce inequities. In addition, if—as some have suggested—pollution control regulations are regressive,49 such an approach would do more harm than good to disadvantaged communities.

A more promising approach would be to design regulations that focus on source-specific emissions. For example, a recent paper modeled the effectiveness of three types of regulatory strategies in reducing racial and ethnic inequalities in PM2.5 exposures.50 It concluded that a “location-specific” approach was more effective than either an ambient-air-quality or sector-specific approach. The authors found that because the most exposed racial–ethnic groups and many emissions sources are clustered in specific locations, the location-specific approach can eliminate exposure disparities at much lower cost than the other approaches while still achieving equal or better reductions in overall pollution levels. This conclusion is consistent with the experience of environmental justice advocates, who have traditionally focused their advocacy more on specific siting decisions than on the adoption of regulations.

This finding in turn suggests that although integrating equity considerations into rulemaking is certainly important, greater gains in reducing environmental inequities might result from reforms to permitting processes and other site-specific decision-making processes. Several states have in recent years taken important steps in that direction. For example, New Jersey enacted a groundbreaking environmental justice law in 2020, under which the New Jersey Department of Environmental Protection must reject a permit application for new facilities that would “cause or contribute to adverse cumulative environmental or public health stressors in the overburdened community that are higher than those borne by other communities,” unless the “facility will serve a compelling public interest in the community where it is to be located.”51 More recently, New York enacted a law providing, “No permit shall be approved or renewed by the [Department of Environmental Conservation] if it may cause or contribute to, either directly or indirectly, a disproportionate or inequitable or both disproportionate and inequitable pollution burden on a disadvantaged community.”52 Of course, to the extent that such laws are premised on race-based distinctions, they are potentially subject to the same legal risks that Farber ascribes to race-conscious rulemaking more generally. If, along the lines of Farber’s proposal, such laws were to define “overburdened communities” in terms of exposure and vulnerability, they would be less susceptible to such legal challenges.

Another consideration is the expressive function of the approach Farber suggests. By evaluating environmental policies using the neutral terms of exposure and vulnerability, Farber’s proposal can achieve changes in material circumstances that promote environmental justice. However, its expressive function is very different from an analysis explicitly framed in terms of race or income. This difference has both advantages and disadvantages. On the positive side of the ledger, it can, as Dan Kahan has suggested with respect to deterrence arguments in criminal law, “cool … expressive disputes.”53 The explicit consideration of racial disparities in rulemaking is currently a topic of significant legal and political disputes in the United States. Using the neutral vocabulary of geography, exposure, and vulnerability, regulators can avoid inflaming those disputes.54

For many environmental justice advocates, however, this strength of Farber’s proposal could also be its greatest weakness. The mission of the environmental justice movement is to not only eliminate material environmental disparities but also to call attention to the decades or centuries of discriminatory practices that have produced those disparities. From this perspective, masking the true nature of discriminatory actions and their consequences through the use of value-neutral language means that agencies are not addressing the full range of harms—both physical and emotional—caused by environmental discrimination. As Robert Bullard, the father of environmental justice, put it, “[w]hen we see environmental racism, we will call it what it is.”55

One way of conceptualizing this gap in Farber’s proposal is through the common description of environmental justice as encompassing both distributive and political aspects.56 Distributive justice addresses whether environmental “burdens are ‘equitably’ distributed among communities.”57 Political justice, by contrast, “concerns the fairness of the decisionmaking process.”58 Farber’s proposal addresses the former component of environmental justice but not the latter.

In fact, some environmental justice advocates further divide it into three components: “equity in the distribution of environmental risk, recognition of the diversity of the participants and experiences in affected communities, and participation in the political processes which create and manage environmental policy.”59 The third aspect identified in this formulation—recognition—is explicitly nullified by the approach Farber suggests.

Farber’s article thus represents an important starting point. It identifies a promising method of addressing inequities in rulemaking while not running afoul of the doctrinal trends of the current Supreme Court majority. Congress and the EPA can and should do more to support the continuing development of promising technologies that will make Farber’s approach increasingly feasible. Finally, because Farber’s proposal does not address the participatory or recognition aspects of environmental justice, agency decision-makers who adopt this approach to analyzing the costs and benefits of regulations should be careful to take those aspects of environmental justice into account in other aspects of their rulemaking processes.


Whitman v. Am. Trucking Ass’n, 531 U.S. 457 (2001).


Entergy Corp. v. Riverkeeper, Inc., 556 U.S. 208 (2009).


33 U.S.C. § 1326(b).


Michigan v. EPA, 576 U.S. 743 (2015).


See, e.g., Jonathan S. Masur & Eric A. Posner, Cost-Benefit Analysis and the Judicial Role, 85 U. Chi. L. Rev. 935, 976 (2018) (describing Justice Kagan’s dissent in Michigan—which was less favorable to cost–benefit analysis than the majority—as “suggest[ing] a default rule: agencies must weigh costs and benefits, at least in some fashion, absent an explicit statement to the contrary”); Cass R. Sunstein, Cost–Benefit Analysis and Arbitrariness Review, 41 Harv. Envt L. Rev. 1, 15 (2017) (“The [Michigan] dissenters clearly adopted a background principle that would require agencies to consider costs unless Congress prohibited them from doing so. There is every reason to think that the majority—which did, after all, invalidate EPA’s regulation—would embrace that principle as well.”).


See, e.g., Ricci v. DeStefano, 557 U.S. 557 (2009), which involved the New Haven fire department’s rejection of a list of candidates eligible for promotion—a list that resulted from a testing process—because all the firefighters on the list were white. The city explained that it feared it would be subject to disparate-impact liability if it moved forward with the list. The Supreme Court, however, in a 5–4 decision, held that the city’s rejection of the list itself violated Title VII because it was explicitly based on race, and the city had not made a sufficient showing that it would have been subject to liability for using the list to justify that decision. Id. at 586–87.


See Wynn v. Vilsack, 545 F. Supp. 3d 1271, 1294–95 (M.D. Fla. 2021) (enjoining the loan forgiveness program); Holman v. Vilsack, No. 21-1085, 2021 WL 2877915, at *13–14 (W.D. Tenn. July 8, 2021) (same); Miller v. Vilsack, No. 4:21-CV-0595, slip op. at 23 (N.D. Tex. July 1, 2021) (same); Faust v. Vilsack, 519 F. Supp. 3d 470, 478 (E.D. Wis. 2021) (imposing a temporary restraining order).


The cases are Students for Fair Admissions, Inc. v. President and Fellows of Harvard College (Docket No. 20-1199) and Students for Fair Admissions, Inc. v. University of North Carolina (Docket No. 21-707).


Grutter v. Bollinger, 539 U.S. 306 (2003).


See, e.g., Adam Liptak, Supreme Court Seems Ready to Throw Out Race-Based College Admissions, N.Y. Times, Oct. 31, 2022,; Amy Howe, Affirmative Action Appears in Jeopardy After Marathon Arguments, SCOTUSblog (Oct. 31, 2022, 7:44 PM), [Editor’s note: As predicted by observers, the Supreme Court rejected two universities’ admissions practices that asked applicants to identify their racial or ethnic identities as violative of the Constitution. Students for Fair Admissions Inc. v. Harvard University, 600 U.S. ____ (2023); Students for Fair Admissions, Inc. v. University of North Carolina, 600 U.S. ___ (2023).]


As Farber notes, however, the use of race as a factor in the regulatory context may be treated differently by the Court than when the action under review involves choices between identifiable individuals. Daniel A. Farber, Inequality and Regulation: Designing Rules to Address Race, Poverty, and Environmental Justice, 3 Am. J. L. & Equal. 52, 37–38 (2023).


Id. at 44.


Farber’s proposal also depends on identifying differences in vulnerability. There is strong evidence of differences in vulnerability to air pollution—along lines of age, gender, and race, see, e.g., Mercedes Medina-Ramón & Joel Schwartz, Who Is More Vulnerable to Die from Ozone Air Pollution?, 19 Epidemiology 672 (2008)—and, as Farber observes, along socioeconomic lines, Farber, supra note 11 at 43–44. This comment addresses only the risk aspect of Farber’s proposal, but even as to vulnerability, Farber himself acknowledges that “[i]n a world of burgeoning data mining, Big Data analysis, and artificial intelligence—a world in which advertising can already be tailored to individuals based on minute knowledge of their behavior—we may not be as far from completely disaggregated models of risk and vulnerability as one might think.” Id. at 50–51.


Farber, supra note 11 at 47.


Daniel M. Sullivan & Alan Krupnick, Using Satellite Data to Fill the Gaps in the US Air Pollution Monitoring Network 2 (Res. for the Future, Working Paper No. 18-21, 2018),


David Coursen, Environmental Justice Requires Adequate Air Quality Monitoring System, Bloomberg L. (Mar. 9, 2021),


Corbett Grainger & Andrew Schreiber, Discrimination in Ambient Air Pollution Monitoring?, 109 AEA Papers & Proc. 277 (2019).


40 C.F.R. § 58.12(c)-(e) (2023). For example, as of 2018, there were 624 Federal Reference Method PM2.5 monitors: 70 providing daily data, 422 providing data every third day, and 132 providing data every sixth day. EPA Off. of Air Quality Plan. & Standards, Policy Assessment for the Review of the National Ambient Air Quality Standards for Particulate Matter, at 2-18 (2020),


A 2018 Reuters investigation found that the air monitoring network failed to “identif[y] risks from 10 of the biggest refinery explosions over the past decade.” Tim McLaughlin et al., Special Report: U.S. Air Monitors Routinely Miss Pollution—Even Refinery Explosions, Reuters (Apr. 26, 2018),


Eric Yongchen Zou, Unwatched Pollution: The Effect of Intermittent Monitoring on Air Quality, 111 Am. Econ. Rev. 2101 (2021). The study found no comparable variation in areas with daily monitoring, and when 1-in-6-day monitors were retired, the differences in those areas disappeared. Id. at 2102.


40 C.F.R. Part 58 app. D (2023) (requiring near-road monitoring of NO2 and at least one collocated CO monitor, and “[f]or CBSAs with a population of 1,000,000 or more persons, at least one PM[2.5] monitor is to be collocated at a near-road NO[2] station”).


The EPA estimates that regulatory monitors cost $15,000 to $40,000 and can be operated only by highly trained individuals. Andrea Clements & Robert Vanderpool, FRMs/FEMs and Sensors: Complementary Approaches for Determining Ambient Air Quality (Dec. 18, 2019),


Bryan N. Duncan et al., Satellite Data of Atmospheric Pollution for U.S. Air Quality Applications: Examples of Applications, Summary of Data End-User Resources, Answers to FAQs, and Common Mistakes to Avoid, 94 Atmospheric Envt 647, 648, 656 (2014); Laura Gladson et al., Evaluating the Utility of High-Resolution Spatiotemporal Air Pollution Data in Estimating Local PM2.5 Exposures in California from 2015–2018, 13 Atmosphere 2022, 85, 86 (2022); Aaron van Donkelaar et al., Global Estimates of Fine Particulate Matter Using A Combined Geophysical-Statistical Method With Information from Satellites, Models, and Monitors, 50 Envt Sci. & Tech. 3762, 3762–72 (2016).


Liftoff! TEMPO Instrument Soars into Space, Harv. & Smithsonian Ctr. for Astrophysics (Apr. 7, 2023),


Duncan et al., supra note 23, at 648–49.


Gladson et al., supra note 23, at 85, 86.


Id. at 90, 96–97. Other studies demonstrating the ability of satellite data to identify high-pollution areas undetected by the ground monitoring network include Oliver Schneising et al., Severe Californian Wildfires in November 2018 Observed from Space: The Carbon Monoxide Perspective, 20 Atmospheric Chemistry & Physics 3317 (2020); Meredith Fowlie et al., Bringing Satellite-Based Air Quality Estimates Down to Earth (Nat’l Bureau of Econ. Rsch., Working Paper No. 25560, 2019); Qian Di et al., An Ensemble-Based Model of PM2.5 Concentration Across the Contiguous United States with High Spatiotemporal Resolution, 130 Envt Intl 1, 11 (2019); Sullivan & Krupnick, supra note 15; see generally Nicolas Gendron-Carrier et al., Subways and Urban Air Pollution (Nat’l Bureau of Econ. Rsch., Working Paper No. 24183, 2018).


Congress recently recognized this promise by appropriating $3 million “to deploy, integrate, and operate air quality sensors in low-income and disadvantaged communities.” Inflation Reduction Act of 2022, Pub. L. No. 117-169, § 60105(c).


Presentation, Ron Williams, Air & Energy Research Program, EPA, New Paradigm for Air Pollution Monitoring: 2014–2018 Progress Report, 9 (Apr. 30, 2018),


Id. at 4.


Id. at 7.


See PurpleAir, (real-time map) (last visited May 8, 2023).


See, e.g., Lidia Morawska et al., Applications of Low-Cost Sensing Technologies for Air Quality Monitoring and Exposure Assessment: How Far Have They Gone?, 116 Envt Intl 286 (2018); Nuria Castell et al., Can Commercial Low-Cost Sensor Platforms Contribute to Air Quality Monitoring and Exposure Estimates?, 99 Envt Intl 293 (2017).


Jianzhao Bi et al., Contribution of Low-Cost Sensor Measurements to the Prediction of PM2.5 Levels: A Case Study in Imperial County, California, USA, 180 Envt Rsch. 108810 (2020).


See text accompanying note 24, supra.


2017 Cal. Stat. ch. 136.


See, e.g., Erin Douglas, Texas Rolls Out Mobile Pollution Monitoring Capabilities Following Failures During Hurricane Harvey, Tex. Trib. (Nov. 10, 2020),; Bruce Finley, Catching the Air: Colorado’s Mobile Monitoring Lab Goes to the Source in Fight Against Pollution, Denver Post (Feb. 2, 2020),


Thomas O. McGarity, Hazardous Air Pollutants, Migrating Hot Spots, and the Prospect of Data-Driven Regulation of Complex Industrial Complexes, 86 Tex. L. Rev. 1445, 1457 (2008).


See Finley, supra note 37.


See Douglas, supra note 37. According to the TCEQ, stationary monitors are taken offline and secured before a major storm to protect the equipment. Gary Rasp, Air Monitoring During Hurricanes, We Were Wondering: A TCEQ Blog (June 2, 2021), The TCEQ first deployed its new Strategic Mobile Air Reconnaissance Technology rapid assessment survey van (“SMART van”) in the wake of Hurricane Laura in 2020 to drive to the fencelines of facilities that had shut down during the storm and monitor emissions throughout the subsequent startups. Kelsey Johnson & Kierra Sam, TCEQ Unveils New Mobile Technology to Help with Air Monitoring During Emergencies, 12 News (Nov. 10, 2020),


EPA regulations define “fugitive emissions” as emissions that “could not reasonably pass through a stack, chimney, vent, or … equivalent opening.” 40 C.F.R. § 52.21(b)(20) (2023). The agency has explained that “a typical refinery or chemical plant can emit 600–700 tons per year of [volatile organic compounds] from leaking equipment, such as valves, connectors, pumps, sampling connections, compressors, pressure-relief devices, and open-ended lines.” EPA, EPA-305-D-07-001, Leak Detection and Repair—A Best Practices Guide 3 (2007).


Petroleum Refinery Sector Risk and Technology Review and New Source Performance Standards, 80 Fed. Reg. 75,178 (Dec. 1, 2015). The fenceline monitoring requirements were codified at 40 C.F.R. § 63.658.


See Bay Area Air Quality Mgmt. Dist., Refining Emissions Tracking, Regulation 12, Rule 15 (Apr. 20, 2016),; San Joaquin Valley Air Pollution Control Dist., Petroleum Refinery Fence-line Air Monitoring, Rule 4460 (Dec. 19, 2019),; Santa Barbara Air Pollution Control Dist., Refinery Fenceline and Community Air Monitoring, Rule 364 (May 21, 2020),; South Coast Air Quality Mgmt. Dist., Refinery Fenceline and Community Air Monitoring, Rule 1180 (Dec. 1, 2017),


Standards of Performance for New, Reconstructed, and Modified Sources and Emissions Guidelines for Existing Sources: Oil and Natural Gas Sector Climate Review, 87 Fed. Reg. 74,702, 74,744 (Dec. 6, 2022).


See, e.g., 42 U.S.C. §§ 7410(a)(2)(F), 7414(a)(1), (3), 7619(a) (2023).


Throughout 40 C.F.R. Part 53, a FEM is discussed as being either an “analyzer” or a “sampler”—in both cases incorporating the assumption that it is a single object. In addition, for example, 40 C.F.R. § 53.20(d) recommends the creation of “a temperature-controlled environmental test chamber” in which analyzers for SO2, CO, O3, or NO2 monitoring are to be placed for the duration of the test.


Conversely, EPA’s current weight-of-evidence approach to determining whether one region contributes to another region’s violation of the NAAQS would appear to be more easily adapted to the integration of satellite and air sensor data. See Catawba Cnty. v. EPA, 571 F.3d 20, 28 (D.C. Cir. 2009); Presentation, EPA, Air Quality Designations 101: Initial Area Designations for the National Ambient Air Quality Standards 6 (July 25, 2017),


His only discussion of this issue is one paragraph in the introduction to his article. He states: “When an agency discovers [that disadvantaged communities are exposed to more pollution and suffer greater harms from exposure], it has several options. It can make the regulation as a whole more rigorous, providing these communities and others with greater protection. It can add features to the regulation providing special restrictions on sources that impact areas with elevated pollution levels. It can also address this impact in various other ways: by requiring more rigorous monitoring of those sources, making enforcement actions against them a priority, or launching (or working with other agencies to launch) additional rulemakings aimed at pollution hotspots.” Farber, supra note 11 at 4.


See, e.g., id. at 34; Lucas W. Davis & Christopher R. Knittel, Are Fuel Economy Standards Regressive?, 6 J. Assn Envt Res. Econ. S37 (2019). But see Antonio Bento et al., Who Benefits from Environmental Regulation? Evidence from the Clean Air Act Amendments, 97 Rev. Econ. Stat. 610 (2015) (finding that the benefits of the 1990 Clean Air Act Amendments were progressive).


Yuzhou Wang et al., Location-Specific Strategies for Eliminating US National Racial-Ethnic PM2.5 Exposure Inequality, 119 Proc. Nat’l Acad. Scis. e22005548119 (2022).


N.J. Rev. Stat § 13:1D-160(c) (2023). In March 2023, the New Jersey Department of Environmental Protection adopted regulations to implement this law. N.J. Admin. Code § 7:1C (2023).


N.Y. Envt Conserv. Law § 70-0118 (2022),


Dan M. Kahan, The Secret Ambition of Deterrence, 113 Harv. L. Rev. 413, 417 (1999); see also Note, The CITES Fort Lauderdale Criteria: The Uses and Limits of Science in International Conservation Decisionmaking, 114 Harv. L. Rev. 1769, 1782 (2001) (suggesting that “by guiding discussions away from contentious disputes about attitudes toward nature and reasons to value species and toward neutral dialogues about numerical estimates of ‘decline’ and ‘fragmentation,’” a change in the decision-making criteria of the Convention on the International Trade in Endangered Species could “reduce dissension” and “improve the discourse” among member states); Louis Murray, Note, Reconceptualizing Pretext’s Role in Administrative Law, 57 Harv. J. on Legis. 481, 497 (2020) (“In light of the country’s growing political divide and vitriol for members of the opposing political party, it is not unreasonable to believe that complete honesty about the reasons for Executive Branch officials’ actions may not always be the best policy for promoting national cohesiveness.”).


See Dan M. Kahan & Donald Braman, More Statistics, Less Persuasion: A Cultural Theory of Gun-Risk Perceptions, 151 U. Pa. L. Rev. 1291, 1319 (2003) (“Because they elide contestable judgments of value, instrumental arguments are the ‘don’t ask, don’t tell’ solution to cultural disputes in the law—not just over gun control, but over policies like the death penalty, hate crimes, welfare reform, environmental regulation, and a host of other controversial policies.”).


Maria Paula Rubiano, The Event That Changed the Environmental Justice Movement Forever, Grist (Nov. 1, 2021),; see also U.S. Commn on C.R., Environmental Justice: Examining the Environmental Protection Agencys Compliance and Enforcement of Title VI and Executive Order 12,898, at 95 (2016) (Statement of Chairman Martin R. Castro: “Environmental ‘justice’ should be the end result of a complaint of environmental ‘discrimination’ or environmental ‘racism.’ I know the last two terms make some people feel uncomfortable, however, we must call it what it is and treat this kind of discrimination and racism in the same manner we treat traditional discrimination and racism in other settings—whether it is in the form of disparate treatment of communities of color or disparate impact on communities of color.”).


See, e.g., Alice Kaswan, Environmental Justice: Bridging the Gap Between Environmental Laws and “Justice,” 47 Am. U. L. Rev. 221, 230 (1997).


Id. at 231.


Id. at 233.


David Schlosberg, Reconceiving Environmental Justice: Global Movements and Political Theories, 13 Envt Pol. 517, 517 (2004); see also Margot Hurlbert & Jeremy Rayner, Reconciling Power, Relations, and Processes: The Role of Recognition in the Achievement of Energy Justice for Aboriginal People, 228 Applied Energy 1320 (2018).

This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits you to copy and redistribute in any medium or format, for non-commercial use only, provided that the original work is not remixed, transformed, or built upon, and that appropriate credit to the original source is given. For a full description of the license, please visit