Clean Air Act (CAA)

Key Concepts

  • Different types of pollutants are regulated differently by the Clean Air Act,1 depending primarily on their health effects.
  • Different polluters are regulated differently, depending on several factors. Most notably, newly constructed sources are subject to stricter controls, as are sources located in high-pollution areas.
  • Some regulatory standards are based on the desired health outcome, whereas others are based on what is technologically feasible. Standards under the Clean Air Act may be “technology-forcing.”


In 1948, a temperature inversion trapped steel mill emissions and created a killer smog in the small town of Donora, Pennsylvania. Over the course of four days, twenty people died and at least a third of the population of 14,000 was sickened. Four years later in London, a killer fog imbued with coal smoke killed thousands of residents in less than a week. These catastrophes were wake-up calls. Until then, most people reacted to rising levels of air pollution with complacency—or even welcomed it as a hallmark of industrial productivity. In 1962, Rachel Carson published Silent Spring, which documented the environmental harm from DDT and other pesticides and criticized the performance of chemical companies and public officials. Increased public concern in the 1960s led to a grassroots demand for environmental protection, which prodded governmental action. Congress enacted the original Clean Air Act (CAA) in 1963.

Initially the federal government played a relatively passive role, providing funding and research, but deferring to the states to regulate pollution. This did not work well. Some states were conscientious about environmental protection, but many were not. The reason is not hard to understand. It is expensive for industry to comply with antipollution laws, which gives companies an incentive to build new factories where regulation is less strict. States with lax regulations were therefore perceived to have a competitive edge in attracting new industry and new jobs. This potential advantage of looser environmental regulation resulted in what is often called the race to the bottom.

race to the bottom

The tendency of some states to be lax in environmental protection, in the absence of national standards, in order to attract new industry and jobs

Congress responded to the lack of state action in 1970—the same year the Environmental Protection Agency was formed—by significantly revamping the Clean Air Act. A major feature of the revised act is uniform national standards that states are required to meet and maintain. Several other federal environmental laws were enacted by Congress over the next ten years, which is sometimes called the “environmental decade.”

uniform national standards

Federal standards that apply uniformly across the country


In typical fashion, Congress begins the Clean Air Act with findings and a declaration of purpose. In its findings, Congress defines the problem thus: “[T]he growth in the amount and complexity of air pollution brought about by urbanization, industrial development, and the increasing use of motor vehicles, has resulted in mounting dangers to the public health and welfare.”2

The explicit purpose of the act is “to protect and enhance the quality of the Nation’s air resources so as to promote the public health and welfare.”3 The meaning of “public health” is clear enough. The term “welfare” is used broadly here, encompassing such things as effects on soils, water, wildlife, visibility, and climate as well as property damage, transportation hazards, and effects on economic values and personal well-being.4

The Environmental Protection Agency (EPA), in conjunction with the states, implements the Clean Air Act. (For convenience, we will usually refer in this chapter and elsewhere just to the EPA, but many of the actions—such as issuance of permits—are often done by state authorities.) The act is composed of multiple programs attacking various aspects of air pollution. Several of the most important programs will be discussed here. Congress made additional major revisions to the act in 1977 and 1990 to strengthen and improve its protection of our air.

Sometimes new regulations issued by the EPA to implement the act cannot be achieved with existing technology. This forces industry to develop new pollution-reducing techniques and technologies, which otherwise it would have no incentive to do. Although these “technology-forcing” regulations are often decried by industry, they have helped protect our air quality.


The National Ambient Air Quality Standards (NAAQS) Program is the centerpiece of the Clean Air Act. The NAAQS program focuses on outdoor air pollutants that endanger public health, welfare, or both. Principle targets are ozone and airborne particles such as soot that create what we commonly call smog. This part of the CAA mandates that the EPA identify the relevant pollutants and determine what levels are safe in the air we breathe. Later sections of this chapter deal with regulatory efforts to achieve those safe levels.

National Ambient Air Quality Standards (NAAQS)

A program of the CAA that sets allowable concentrations of criteria pollutants in the ambient air

Criteria Pollutants

The EPA’s first task in the NAAQS program is to identify pollutants that (1) come from many sources and are pervasive in our outdoor air, and (2) pose a danger to public health or welfare at ambient (outdoor) levels.5 The pollutants meeting this definition are called criteria pollutants. It’s not a very descriptive term, but everyone uses it. The list of the EPA-designated criteria pollutants is short:

criteria pollutants

A short list of pollutants that are pervasive in the ambient air and harmful at ambient levels

  • Sulfur dioxide (SO2)
  • Nitrogen oxides (NOx)
  • Particulate matter (PM10 and PM2.5)
  • Carbon monoxide (CO)
  • Ozone
  • Lead

These are not the only dangerous air pollutants, but they are the only ones that the EPA recognizes as pervasive and for which there is clear evidence of adverse effects at ambient levels—such as the sulfur dioxide and particulates responsible for the Donora and London killer smog episodes. Many noncriteria pollutants pose the risk of serious hazards, including cancer. But the evidence is not derived from observations of people at ambient exposure levels; rather, it is extrapolated from effects observed in people exposed to much higher levels (usually in workplaces) or from laboratory animal studies. Such pollutants, called hazardous air pollutants, are discussed later in this chapter.

There is intense controversy at present about how to regulate carbon dioxide (CO2) and other greenhouse gases, including controversy over whether the EPA should designate them as criteria pollutants and add them to this list.

Ambient Air Standards

The EPA’s second task under the NAAQS program is to set ambient air standards, specifying how much of each criteria pollutant is allowed in the air we breathe. These must be uniform national standards. This innovation in 1970 was a significant change from the previous piece-meal state control, and it effectively counteracted the harmful “race to the bottom” referred to earlier. The EPA may make distinctions for certain sensitive areas, such as national parks. But distinctions are not allowed among the states.

The act requires two types of standards: primary standards related to public health and secondary standards related to the public welfare. Congress’s instructions here are quite cryptic. The EPA is told to adopt primary standards “requisite to protect the public health” allowing an “adequate margin of safety” and secondary standards “requisite to protect the public welfare.”6 It is the EPA’s job to interpret what Congress means by these cryptic phrases and then to translate Congress’s intent into standards—mostly numeric standards—based on the best science available. We will focus on primary standards in discussing how the EPA proceeds.

primary standards

Standards intended to protect human health

secondary standards

Standards intended to protect the public welfare

Interpreting the Legislative Goal

What did Congress mean when it told the EPA to adopt standards “requisite to protect the public health”? Interpreting the intent of these words, the EPA had to make some important policy decisions:

  • Who is to be protected? The EPA decided that protecting “public health” means standards adequate to protect the general population, including sensitive subpopulations such as children, pregnant women, the elderly, and people with chronic illnesses, such as asthma. This interpretation of the act’s language is not inevitable. The EPA could conceivably have interpreted public to refer, for example, to the average healthy individual or to the most vulnerable individual.

  • How much protection? For every primary standard it sets, the EPA conducts an extensive and detailed study, intended to estimate, based on the best scientific evidence available, what impacts would be sustained by what sensitive groups of people from various ambient levels.7 Estimates are generally based on conservative assumptions, so as to err on the side of safety. The EPA does not interpret the act as demanding zero risk. As a goal, the EPA seeks to identify ambient levels that will not cause any excess incidence of disease (that is, over and above the occurrences of disease from other causes, called the background level).

background level

The incidence of a disease (or other measurable condition) in the absence of a particular factor of concern

In practice, what constitutes protection of public health with an adequate margin of safety cannot be objectively pinpointed. The EPA must exercise discretion, and it is usually upheld by the courts. In general, the courts pay great deference to an agency’s interpretation of a statute it is charged with implementing, as well as factual determinations within the agency’s expertise. Even if there are other conclusions the EPA could reasonably have reached, and even if the court thinks a different decision would have been better, the court will uphold the EPA’s discretion in setting a standard so long as there is some reasonable basis for it.8

Setting Numerical Standards

Having interpreted the congressional intent, the EPA’s next task is to translate it into numeric ambient standards sufficient to protect public health with an adequate margin of safety. The act directs the EPA to use the best available science. In contrast to many standards it sets, the EPA does not consider cost at this point (although it is considered at a later stage when control options are developed). The CAA does not explicitly bar the EPA from taking cost into account in setting ambient standards, but both the EPA and the courts have interpreted it that way.9

Primary standards for criteria pollutants are stated in terms of allowable concentration in the ambient air. Ambient concentrations vary over time, so standards are stated as a maximum concentration averaged over a specified period. The EPA sets each standard based upon the toxicology of the individual pollutant and on the likely scenarios of exposure. With respect to four of the criteria pollutants (CO, NO2, PM2.5, and SO2), the EPA has established two primary standards averaged over different time periods. For example, carbon monoxide (CO) has a dual standard: a maximum of 9 ppm (parts per million) averaged over an eight-hour period, and 35 ppm averaged over a one-hour period. Why? Because CO attaches to blood hemoglobin (red blood molecule), displacing the oxygen it should be carrying. Scientific studies have determined the CO blood level that impairs oxygen-carrying capacity enough to increase the risk of angina attack among people with preexisting heart disease (a sensitive subpopulation). That level is reached when people breathe an average of 9 ppm for eight hours, or 35 ppm for one hour.

The CAA requires the EPA to review national ambient air quality standards every five years.10 While the EPA has often missed deadlines, it has made several revisions to the standards over the years based on advances in scientific knowledge and understanding of exposure. For example, the ozone standard was originally set as a one-hour average, but was later changed to an eight-hour average. That decision illustrates the importance of science in the EPA’s standard-setting. Ozone results when hydrocarbons and nitrogen oxides from vehicle exhaust bake for a few hours in sunlight. Not surprisingly, ozone levels in Los Angeles peaked a few hours after the morning rush hour—back when the one-hour standard was first set. But by the 1990s, rush hour in Los Angeles lasted virtually all day, keeping ozone levels high throughout daylight hours. That change in exposure scenario, plus toxicological evidence that the lung damage caused by ozone accumulates over this multihour period, meant that the one-hour standard was not adequate protection, especially for children. Children are a sensitive subpopulation, because their size and immature development make them more vulnerable to toxic effects, such as that of ozone. Children tend to play outdoors all day during summer, so an eight-hour exposure is a realistic scenario for them. Moreover, running and playing means they breathe more heavily, thus taking in even more ozone. Based on consideration of all these factors, the EPA revised the ozone standard to set a maximum concentration averaged over an eight-hour period.

The EPA has in-house scientists, but it has also established several committees of independent scientific experts to assist in setting science-based standards. One of the earliest advisory bodies was created not by the EPA itself, but by congressional mandate in the Clean Air Act. This is the Clean Air Science Advisory Committee (CASAC). Its mission is to advise the EPA on air pollution issues, based on the best science available. In particular, CASAC advises on how much of the criteria pollutants can be allowed in the ambient air and still meet Congress’s goal—in essence, the no-excess-effects threshold.

It isn’t possible to pinpoint this important threshold exactly. The committee reviews and discusses all the relevant scientific data, and each member forms his or her own judgment of the most likely threshold. These individual judgments vary, but they are usually clustered fairly closely together. In essence, that cluster reflects the range of reasonable scientific opinion as to where the actual (but unknowable) threshold lies. CASAC recommends to the EPA—as its consensus opinion—that the standard for the relevant pollutant be set within that range, usually pointing out that the upper end of the range does not provide any margin of safety.

The EPA administrator is the individual authorized to set the numerical standard (and to make other determinations under the act). He or she is a political appointee, whose decisions are based on both scientific and policy considerations. Traditionally, the administrator sets the standard somewhere within the range recommended by CASAC, although this is not mandatory. Setting the standard toward the lower end of CASAC’s recommended range is one way the EPA often provides a margin of safety.

Although other scientific advisory boards are created by the EPA itself rather than by statute, they tend to operate in similar fashion, using a consensus approach to establish a reasonable range of estimates for some number that is impossible to pinpoint exactly. On particularly contentious issues, the EPA—either on its own initiative or at the direction of Congress—often seeks advice from the National Academy of Sciences.


To have clean air, you need to worry not only about pollutant levels in the ambient air, but also about what is coming out of industrial smokestacks (stationary sources) and vehicle tailpipes (mobile sources). You can’t achieve NAAQS unless you control pollutant emissions—that is, pollution measured at the point of discharge. The following sections relate to stationary sources. Mobile sources will be discussed later in the chapter.

The Clean Air Act distinguishes between stationary sources based on several factors. The way polluters are regulated depends in large part on whether they are new or existing sources, and whether they are located in an attainment or nonattainment area. Other distinctions that affect regulation include the amount of pollution a source is capable of emitting (major versus minor source) and industrial category. What we have is a framework of intersecting regulations for stationary sources, not a single one-size-fits-all. Each emission standard is commonly referred to by an acronym, adding even more confusion (see table 3.1).

industrial category

The Standard Industrial Classification Manual published by the federal Bureau of the Budget identifies and assigns code numbers to different industries and subindustries

emission standard

A restriction on how much of a pollutant an industrial source may emit into the air; the particular standard that applies depends on the type of pollutant and other circumstances

Table 3.1 Alphabet Soup: Emission Standards for Stationary Sources


Defining New Sources

As is often true with legislation, ordinary-sounding words don’t mean quite what they seem to mean. A new source under the Clean Air Act means a facility that was constructed or modified after an applicable standard was first announced. A source is deemed modified if physical or operational changes result in emitting either more of a pollutant or a new pollutant. Stricter regulation of new sources makes sense, because a newly constructed facility can be designed to include the latest pollution-prevention techniques. Treating a modified source like a new source also makes sense. If you remodel a facility enough to increase pollution, good planning should enable you to build in antipollution improvements.

new source

A new or modified source—generally one that was constructed or modified after an applicable standard was first proposed or announced

New Source Performance Standards (NSPS)

The Clean Air Act distinguishes between new and existing stationary sources, imposing stricter standards on new sources. For new sources, the Clean Air Act directs the EPA to establish New Source Performance Standards (NSPS).

New Source Performance Standards (NSPS)

Uniform emission standards for emissions of criteria pollutants by new sources under the CAA

The act requires uniform national NSPS emission standards which reflect the degree of emission limitation achievable through the best “adequately demonstrated” technology, taking into consideration cost, energy requirements, and other (nonair) environmental impacts.11 For existing sources, the job of regulating criteria pollutants is left to the individual states (so long as the area is in compliance with NAAQS, which is addressed later). The EPA offers guidance, and it must approve a state’s implementation plan, but the EPA does not directly set emission limits.

Although NSPS rules apply nationwide, actual emission limits based on those rules are not the same for everyone. Industries differ in the types and amounts of pollutants they emit and how much they can feasibly reduce emissions. Therefore, for each designated industry category, the EPA develops NSPS standards with respect to the pollutants it emits. For example, standards for oil refineries are different than standards for cement manufacturers.

New Source Review (NSR)

New Source Review (NSR) is a program that helps with implementation and enforcement of regulations pertaining to new and modified sources, by requiring that projects be reviewed and a permit issued before construction can begin. The requirement of a permit enables the EPA or state environmental authorities to monitor, control, and, if circumstances warrant, bar development of a pollution-producing facility. These NSR permits are also commonly referred to as construction permits or preconstruction permits.12

New Source Review (NSR)

A CAA program requiring review and issuance of a permit before construction of a new emission source is begun

technology-based standards

Standards based on technological feasibility as opposed to desired health outcome

To set emission standards, the EPA has to interpret Congress’s intent and translate it into enforceable—commonly numeric—standards. The term best adequately demonstrated technology (BADT) implies technology that is successfully in use by industry, as opposed to technology that is cutting edge or on the drawing board. The EPA’s general approach is to review the emission reduction systems in existence in the industry and their effectiveness, compare their respective cost, energy use, and non–air side effects, as required by the act. Examples of “nonair quality health and environmental impacts” that could occur when pollutants are removed from air emissions include increased water pollution or increased production of hazardous waste. With regard to cost, the EPA does not look at whether an emission standard would be too expensive for an individual company, but rather whether it would significantly harm the industry as a whole.

best adequately demonstrated technology (BADT)

A uniform technology-based emission standard; the standard applicable to emissions of criteria pollutants by new major sources

The EPA does not prescribe the technology or process that an industrial emitter must use to control emissions. Rather, the EPA determines the level of emissions from the best performers in the industry that use adequately demonstrated technology and extrapolates reasonable emission limits from these role models. Essentially the EPA says, “If they can keep emissions down to this level, you can too.” Companies then have flexibility to decide how to meet the required standards—for example, through pollutant reduction equipment, cleaner energy sources, alternative raw materials, or some combination of factors.

Notice the difference between NSPS and NAAQS. NAAQS is a risk-based standard. NSPS is a technology-based standard, so called because the EPA primarily assesses technology (as well as other factors) to determine what degree of pollution control is feasible.

risk-based standard

A standard based on the desired outcome, as opposed to a technology-based standard that is based on feasibility

The NSR program is aimed at major sources. It is common for regulatory programs to focus on large stationary sources that produce the most pollution. Whether a source is major depends on the quantity of a regulated NSR pollutant it emits, or has the potential to emit. The threshold for a new or modified source to be deemed a major source may vary, depending on the pollutant, the type of source, and whether located in an attainment or nonattainment area. Generally, though, thresholds are set at either 100 or 250 tons per year (TPY) of emissions.

major sources

Under the CAA, a source whose emissions meet a threshold volume that varies depending on various factors but most commonly are set at either 100 or 250 tons per year

Congress originally anticipated that existing sources would gradually disappear through attrition. But compliance with standards for new sources is costly, and thus industry has gone to great lengths to preserve existing sources. Industry has nursed along old facilities, aggressively fought new source rules in court, and designed modifications to fall just short of triggering new source rules. As a result, many old facilities are still in existence, operating under the laxer existing source rules. Determination of whether changes in a facility rise to the level of a modification, thus triggering stricter standards applicable to new sources, is one of the most frequently disputed issues under the Clean Air Act.

An NSR permit specifies what construction is allowed and what emission limits and other requirements must be met. The emission limits and other requirements will depend, to a large extent, on where the facility is located—specifically whether the area is or is not in compliance with the National Ambient Air Quality Standards. An area not in compliance—called a nonattainment area—is subject to much stricter Nonattainment New Source Review (NNSR) environmental controls (see below), and these are reflected in the preconstruction permit. In particular, NNSR requires the proposed facility to employ emission reduction measures adequate to satisfy the tough Lowest Achievable Emission Rate (LAER) standard, as well as securing emission reductions from another area source as offsets. An area in compliance with NAAQS is subject to less stringent requirements set forth in the Prevention of Significant Deterioration (PSD) program (discussed later in this chapter), and which are reflected in preconstruction permits. Less stringent does not mean lax. To get the permit, a facility must employ emission reduction measures sufficient to satisfy the best available control technology (BACT) standard. All preconstruction permits include requirements for monitoring, recording, and reporting emissions to assure the source complies with its permit.

nonattainment area

An area not in compliance with national ambient air quality standards for one or more criteria pollutant

Nonattainment New Source Review (NNSR)

A stricter version of New Source Review applicable to a proposed new source that will emit a pollutant for which the area is not in attainment