A Design Flaw in the Clean Air Act
Why have technology-based standards if you have air quality standards?
The Clean Air Act has two kinds of standards. It sounds like having two kinds of standards should improve air quality more than a single standard. But in reality, one type of standard can result in canceling out the benefits of the other type. If you understand the statute, this is actually pretty obvious once you stop to think about. I have to admit, however, that I hadn’t stopped to think about it until today, even though I’ve taught this stuff dozens of time.
What are the two types of standards? First, the Act tells EPA to issue national air quality standards for major air pollutants and requires states (or as a back-up, the Feds) to create plans to meet those standards by certain deadlines. Second, it sets national pollution controls standards for new cars, factories, and power plants. The air quality standards are based on public health, while the national requirements for new sources are based on the best available technology for controlling emissions.
There are some significant pollutants that aren’t covered by air quality standards, most notably toxic chemicals and carbon dioxide. My argument doesn’t apply to them.
What’s the problem? The problem is that, if the air quality requirements are working, they cancel out the air pollution benefits of the technology requirements.
Here’s why: Suppose a state has to cut emissions by 10%. A new factory is being built. If it weren’t for the new source standards, it would add 2% to the state’s pollution. So now the state would have to cut 12% of its existing pollution. Instead, the new source standards cut the emissions from the plant in half, so it now only adds 1%. It sounds like that’s good for air quality, right? Well, not really. Now the state only needs to cut existing emissions by 11%, not 12%. In other words, the decreased pollution from the new source allows the state to ease up on controlling existing sources by exactly the same amount. The net effect or the new source standards on air quality is zero. On the other hand, the state might have been able to use lower costs reductions instead of the technology-based standards, so that’s a disadvantage.
So why even have the technology based standards for new plants? I can think of two types of justifications. First, you might have the national air quality standards for reasons unrelated to air quality. The reason for having national standards for new cars is just that it would be very inefficient for every state to set its own standards, which would drive carmakers crazy. The federal new source standards for factories seems to be intended to keep them from leaving industrial eastern cities for elsewhere. Congress was also afraid that other place might cut their standards to get the business. Little did they know that the manufacturing would simply move to China.
The other possibility is simply that you might not have much confidence in the air quality standards and the accompanying state implementation plans as drivers of emission reduction. If the air quality standards are soft and flexible, states may not feel any need to compensate for pollution reductions in one sector (new sources) by easing up on emission reductions in another (existing sources). In the extreme case where the air quality standards and state plans are completely ineffective, technology-based standards are really the only way to get pollution reductions. That may be true of cars and trucks, for instance. States have little appetite for cutting use of existing cars, so pollutions standards for new cars may be the only way car emissions ever go down.
The drafters of the Clean Air Act may well have lacked confidence in the air quality standards and state plans as a mechanism, so they may have been hedging their bets.
You might also use technology standards in two specific situations where air quality standards won’t do the job. First, technology standards could be useful when current air quality exceeds national standards. If you want to preserve pristine air, technology standards will help limit the impact of new plants in those areas. Second, you could use technology standards as a punishment for states whose plans fail to work, making it hard for them to attract new industry unless they get serious about the national standards. Later amendments to the Clean Air Act do adopt these strategies. The amendments reflect gaps and deficiencies in the “air standards plus state plans” mechanism. That doesn’t detract from the basic point that if the air quality standards were really rigorous and binding, technology standards would either be ineffective or pointless.
In short, if the national air quality standards and the state implementation plans actually do what they were supposed — reduce air pollution to a harmless level — the technology based standards for new sources would contribute nothing. Technology based standards are useful only to the extent air quality standards don’t work. Otherwise, they merely add to the expense of attaining the standards.
Reader Comments
10 Replies to “A Design Flaw in the Clean Air Act”
Comments are closed.
About Dan
Dan Farber has written and taught on environmental and constitutional law as well as about contracts, jurisprudence and legislation. Currently at Berkeley Law, he has al…
READ more
This doesn’t seem at all mysterious. Plans are plans — they need to be translated into specific actions to matter, at specific places. New source permitting (and existing source rules) are those concrete actions that ensure actual sources reduce. Moreover, source-level innovation makes lower planning targets possible if it creates technological change (and the Act has a very good record of that). For the same reason an article outline does not make a well-crafted paragraph superfluous, the plan and the permits reinforce each other.
Hi Craig – If the state is actually committed to lowering pollution as much as it can, of course, your argument is right. But think about Texas, for instance. If a replacement refinery lowers pollution because it uses better pollution control technology, Texas can ease up on other pollution sources correspondingly and still satisfy the air quality requirement.
Its an old very tired polluter lobbyist’s pretense that the Clean Air Act has many ‘burdensome redundant overlapping requirements’. Is the air clean today?
I dispute your premise, Dan. During the fights of the 80’s and 90’s to get meaningful SIPs, over control was never the problem – the Act’s authors wisely used belt and suspenders knowing a singular approach wouldn’t be effective. FIPs were routinely threatened across the Country as the last and only tool to cudgel states into making hard decisions. Looking forward, there are no safe levels of ozone or fine PM exposure, so there will always be a need for better controls. All forms of transportation, including EVs, generate unhealthful PM near human populations, then add wildfire-related PM and its another problem too big to solve. As someone that’s litigated the adequacy of Texas SIPs and interstate air pollution, and witnessed the unimaginable conditions that refineries knowingly impose on disadvantaged communities such as Beaumont-Port Arthur, I can tell you that the Act’s public health goals need every possible tool to try to make for livable conditions in parts of Texas. Happy talk of how to ratchet back the Act ignores both present conditions and the ominous future humanity faces. Did you read the recent news about the dramatic shortening of average lifespan – most of these people are dying from respiratory disease that is either caused or exacerbated by air pollution. I don’t know where you are going with this post but I must respectfully object.
I applaud the work you’re doing in Texas. The question I’m raising isn’t about over-control. It’s more that the technology standards are like pushing on a water bed: they may cut pollution in one place but that just allows Texas to slack off elsewhere while still purporting to have a valid SIP. I think there are arguments for using the technology standards anyway, partly for technology forcing and partly because the SIP process hasn’t proved to be as effective as initially hoped.
well there is a different mechanism for NSR for non-attainment (NNSR)as opposed to NSR/PSD. And yes, Congress did allow for more “lenient” treatment of existing sources as part of the economic tradeoff. And the amount of reduction for existing sources in NNSR depends on the extent of the non-attainment. If the standards were the same, existing sources would be incentivized to move to another Non-attainment area with less stringent requirements.
Congress did not repeat this particularly state-federal process in the Clean Water Act so perhaps it learned this was an awkward dance.
and depending on the severity of the non-attainment, the NNSR facility would have to obtain offsets from other factories or impose more stringent controls on its own sources. The offsets could be used to retire older, less-efficient pollution sources.
and then there is the “bubble” approach which of course is what led to the Chevron doctrine…
Ambient standards don’t have much bite, while technology and emissions standards do.
For historical takes on the issue, see these excerpts from Victor Flatt and Craig Oren:
https://environmentlawhistory.blogspot.com/2016/04/how-technological-solutions-changed.html
https://environmentlawhistory.blogspot.com/2016/10/a-historical-understanding-of-emissions.html
The standards tell states HOW to meet air quality standards in a common way. Perhaps it could have been left to diverse means but as you point out, manufacturers and other businesses in interstate trade don’t want diverse solutions. The targets for the standards provide more economic certainty for manufacturers to make investments. I’m not sure I see a problem with one standard undermining another. The same thing happens with energy efficiency investments that undercut savings from other energy efficiency investments. This is the nature of isolating regulations rather than constructing them as a portfolio. That’s a different problem.
Technology-based standards can only have a waterbed effect in areas that reach attainment. Otherwise, they just form a baseline from which to go further. And even states in attainment must have maintenance plans, which tend to limit the waterbed effect.