Environmental Law

From wikilawschool.net. Wiki Law School does not provide legal advice. For educational purposes only.
Environmental Law
Relevant texts Image of Cases and Materials on Environmental Law (American Casebook Series)
Cases and Materials on Environmental Law (American Casebook Series)


Image of Environmental Law and Policy (Concepts and Insights)
Environmental Law and Policy (Concepts and Insights)
Image of Environmental Law in a Nutshell (Nutshells)
Environmental Law in a Nutshell (Nutshells)

Related course(s)

Economic and Non-Economic Perspectives[edit | edit source]

  1. The economic perspective
    1. Normative goal of the economic perspective: maximize social welfare. Environmental regulations justified only if they increase welfare of benefitted parties more than they decrease welfare of burdened parties.
      1. Goal isn’t to reduce pollution to the lowest possible level, eliminate adverse consequences of pollution, etc. but rather to maximize social welfare. A reduction in pollution is socially advantageous only if it increases the welfare of the benefited parties more than it decreases the welfare of burdened parties.
    2. Postive/descriptive aspect: divergence btwn polluter’s private costs and the social costs imposed by his activity (externalities) – e.g. steel producer’s costs (labor, electricity, raw materials, etc.) compared to pollution. If a polluter is not required to “purchase” goods like air, clean water, etc., then others will bear the costs of his activities.
      1. What are private costs? Inputs (labor, iron, energy, etc.).
      2. What are social costs? The pollution you emit into the air (essentially, you buy clean air). This doesn’t cost the polluter, but it does cost someone. Rational polluter will not self-regulate: will produce through mix of inputs that maximizes profit.
        1. External costs (externalities)
        2. Social object of regulation often described as internalizing this externality.
  • Attitudinal aspect: Does not view pollution as morally wrong, but rather as a natural result of the pursuit of self-interest by rational actors. No moral valuation of pollution. Not necessarily that economic perspective values free market over regulation. Quarrel here is with the regulatory scheme that does not monetize this activity and internalize the externalities upon rational actors.
  1. Why regulate?
    1. Distributional adjustments
    2. Workplace context: bargaining is possible w/perfect information. Correction of informational asymmetries.
    3. Internalization of externalities.
  2. Key concepts
    1. Tragedy of the Commons (Garrett Hardin, p.7):'
      1. Rational actors seek to maximize individual gain. Increasing exploitation of a common resource has a concentrated benefit to the individual, but only a diffuse cost. Thus, a rational actor will increase his exploitation.
      2. Why don’t they self-limit?
        • (i) Risk of cheating
        • (ii) Difficulty of enforcement
          • (1) How to enforce against participants in the agreement?
          • (2) How to enforce against non-participants/new-comers?
        • (iii) Problems of communication
        • (iv) Three problems
          • (1) Free rider problem
          • (2) Strategic behavior
          • (3) Transaction costs
        • (v) Transaction costs may be higher than benefit one gets from agreement – e.g.,
          • (1) identifying interested parties,
          • (2) negotiating with all involved parties, and
          • (3) enforcing the agreement.
        • (vi) When would agreements be made (thus, where paradigmatic tragedy of the commons is less likely to occur)?
          • (1) When you can identify parties
          • (2) When you have repeat players
          • (3) When the set of parties is limited
          • (4) When the parties have social/community relationships and have relatively-homogenous interests
          • (5) When you can price the good/access to information
        • (vii) When are resources not amenable to these types of agreements?
          • (1) Ocean better example than Hardin’s example – how to enforce limits?
        • (viii) Solutions (p16) and their problems:
          • (1) Taxation: How to valuate the tax? How to enforce? Where does the revenue go? If unknown, why would people support?
          • (2) Privatization: But there are resources that can’t actually be privatized; privatization incurs costs, and not every resource may be sufficiently productive to justify that cost. Also, how is it allocated? Will it be the most efficient distribution? What if the animals are grazing animals that need the whole land?
          • (3) Unitization (cooperative approach): problems include lots of infrastructure; shirking; free riding, etc.
          • (4) Government efforts to encourage cooperation – e.g. enforcement mechanism.
          • (5) Numerical constraints on the total number of animals allowed in the commons – allocated by auction, lottery, queue (first-come-first-serve), merit or need, or grandfathering.
  3. Prisoner’s Dilemma'
    1. Two guilty prisoners – Row and Column
      • (i) If neither confesses, they will each get convicted of a minor crime w/ 3 yr sentence
      • (ii) If both confess, each will get convicted of a more serious crime w/ 5 yr sentence
      • (iii) If one confesses and the other does not, the confessor will get a 1 yr sentence and the other will get a 6 yr sentence.
      • (iv) Sentence Length:
No confession (Column) Confession (Column)
No confession (Row) 3/3 6/1
Confession (Row) 1/6 5/5
  • (v) If the prisoners act rationally – both will confess, even though each would be better off if neither confessed. If Column assumes that Row has confessed, Column will confess b/c he will get 1 less Even if he assumes that Row has not confessed, Column will still confess b/c he will get 2 less years. So, either way, he individually is better off.
  1. Application to Commons:
    • (i) L = maximum capacity re: animals
    • (ii) Cooperative strategy for 2 herders is to each place L/2 cattle on the land, which will net 10u profit for each. Non-cooperative over-grazing will result in 0u profit for each. Non-cooperative under-grazing (F1 grazes L/2 and F2 grazes as much as he wants) will net 1u loss for F1 and 11u profit for F2
    • (iii) Profit
L/2 (Column) unlimited (Column)
L/2 (Row) 10/10
unlimited (Row) 11/-1 0/0
  • (iv) If the prisoners act rationally – neither will limit, even though each would be better off if they both limited. If Column assumes that Row will not limit, Column will not limit b/c he will lose 1u. Even if he assumes that Row will limit, Column will still not limit b/c he will get 1 more unit of profit. So, either way, he individually is better off.
  1. Logic of Collective Action'
  2. Public Goods'
    1. Characteristics:'
      • (i) non-rival (cannot be supplied to 1 individual w/o being supplied to a large number of others at the same time) &'
      • (ii) non-exclusive'
    2. Examples on non-public goods: A public park may be non-exclusive but may not be non-rival (may be rival) b/c others in the park lessens my enjoyment. A private country club is exclusive but not particularly rival.'
    3. Because individuals don’t have incentives to create public goods, such goods are most likely to suffer the tragedy of the commons and must be created by external interventions, such as by the government.'
  3. Externalities'
  4. Pollution under Coasian' bargaining'
    1. Reciprocity problem: Coase views pollution not as harm caused by the factory to the laundry or harm caused by the confectioner to the doctor, but as a reciprocal burden each imposes on the other.
    2. Invariance claim: It doesn’t matter who the gov’t gives the entitlement to, or, conversely, who is liable; the result will be the same.
      1. Only invariance as to the level of pollution produced by the bargaining, not who makes the payment.
    3. Efficiency claim: The result will be the efficient place – maximizing social welfare.
    4. Role of transaction costs: These assumptions depend on no (or sufficiently-low) transaction costs. Such costs are related to:
      1. determining who needs information,
      2. negotiating with all involved parties,
      3. enforcing the agreement,
      4. etc.
    5. When transaction costs are considered, a rearrangement of rights will occur (theoretically) when
      1. the increase in the value of production consequent upon the rearrangement (the benefit)
      2. is greater than the costs which would be involved in bringing it about.
      3. i.e.: net gain > $0.
    6. Example (p18):
      1. If parties can bargain costlessly:
Units of emissions Factory’s Pollution-Control Costs Laundry’s Costs from Pollution (increments of 4)
0 25 0
1 16 4
2 9 8
3 4 ($21 to get here from 0) 12
4 1 16
5 0 20
  1. The lowest combined cost is at 3 emissions ($16).
  2. If Laundry entitled to 0 emissions, the Factory would bargain up to 3 emissions by paying for damages. The payment would be more than $12, the Laundry’s cost to go up to 3 emissions, but less than $21, which is $25 minus $4 already incurred by the Factory at 3 emissions, in return for the Laundry’s agreement not to pursue its legal entitlement.
  3. If Factory entitled to 5 emissions, the Laundry would bargain down to 3 emissions (more than $4 but less than $8 = $20 minus $12 incurred at 3).
  4. Distributional justice analysis: If Laundry represents poor breathers, social justice may favor the Factory. But if the Laundry represents rich country clubbers and the Factory represents poor workers, social justice could favor the Laundry.
  5. Take-away:
    1. Government needs to clearly define who is entitled, needs to reduce transaction costs, and needs to provide for enforcement. Coase did not want the government to directly determine the optimum level – via regulation, tax, etc. Coase was anti-Peguvian (which would impose a $4 tax on the factor).
    2. What was wrong with the Peguvian approach? The gov’t could get it wrong (not knowing the costs—affected parties have incentive to exaggerate—or direct players and being biased by public choice problems or incentive to generate revenue by taxation). Setting the tax at a nonoptimal level destroys the efficiency of bargaining.
  6. Transition – Economic perspectives are “human-centered” – they only value animals and plants to the extent that such are valuable to humans.
  7. Non-economic (normative) perspectives''''
    1. Mark Sagoff – Human-centered (p22)
      1. Government Interaction
        1. Economic regulation
          • (i) “federal programs that set prices, performance standards, entry requirements, schedules, and so on, in the railroad, trucking, securities, telecommunications, and other industries thought to be ‘affected with a public interest.’”
          • (ii) involves the correction of market failures/inefficiencies
          • (iii) by independent agencies (heads can only be removed for cause) and
          • (iv) focuses upon specific industries.
          • (v) Economic regulation may correctly focus upon individuals as consumers.
        2. Social regulation
          • (i) does not involve the correction of market failures,
          • (ii) involves executive agencies (heads can be removed at the will of the President), and
          • (iii) focuses more broadly across the economy to achieve humane goals (“broad ethical and social objectives”).
          • (iv) Social regulation should focus on individuals as citizens.
        3. Problems
          • (i) Consider that different eras might view regulation as economic or social:
            • (1) anti-trust laws were used in the 30s to keep small business around (a social goal) but now are used to keep prices low (an economic goal).
            • (2) price supports (social: “We’re a nation of farmers;” but it fits underneath the definition of economic.)
          • (ii) How do we know that environmental regulation is social and not economic?
            • (1) Historically, the FCC statute was vague – the FCC was charged to regulate “in the public interest.”
            • (2) An economist would argue that environmental regulation seeks to reduce externalities.
  8. Individual Roles
    1. As consumers, people seek to maximize their self interests (personal interests)
    2. As citizens, people seek to further other, abstract goals (public interests)
    3. Example
      • (i) Sagoff’s students, as consumers, would want a ski resort development in Mineral King Valley. As citizens, though, the students would favor not developing the area.
      • (ii) Problems:
        • (1) Sagoff is focusing only upon (present) use value of the land. Option value involves the potential to use the resource later. Existence value relates to resources that a person never uses but still assign some kind of value to. In this case, the students were placing a higher existence value upon the park than the use value of the ski resort. Thus, an economist would expect the result by looking at a broader range of values.
        • (2) Sagoff’s example of driving a dirty car but supporting government action against dirty cars relates to the “free rider” problem. But this is an economically-rational decision. His individual action will incur concentrated costs and minimal benefits, but government action will incur distributed costs and greater benefits.
        • (3) What level of pollution does Sagoff want to achieve? Does he have a substantive standard that is more or less than the economic perspective? Seemingly more protective, but a deliberative political process could, for example, value jobs and corresponding pollution over pristine air.
      • (iii) Takeaway: Deliberation is good.
    4. Paul Taylor – Nature-centered (“biocentric”) (p32)
      1. Principles
        1. Self-defense – can use deadly force if life is threatened and no alternative
        2. Proportionality – basic interests trump non-basic interests
        3. Minimal harm – if neither life nor basic interests at stake but still choose to harm another, do so with goal of minimal harm
        4. Distributive justice – if basic interests in conflict, split the cost
        5. Restitutive justice – if harm done, make restitution
      2. Chart:
Non-human life at stake Non-human basic interests at stake Non-human non-basic interests at stake
Human life at stake H: (a) self-defense H: (a) H: (a)
Human basic interests at stake ? (H+nH)/2: (d) H: (b) proportionality
Human non-basic interests at stake nH: (b) proportionality nH: (b) proportionality or at least (c) minimal harm ?
  1. Why minimal harm? Cynical response – political hedging because people will not be prepared to abide by proportionality. Non-cynical response – On a sliding scale, higher moral agents’ desires trump lesser moral agents’ needs.
  • Singer – Capacity to suffer (p47n23)
    1. Concerned about eating. He would probably not support building a library over a wetland.
    2. Not a sliding scale – a being is either on one side of the line or not. Not as protective of plants as Taylor seems to be.
  1. Life/Non-life paradox - Christopher Stone (p48n24)
    1. Humans sometimes value non-life-bearing landscapes over life-bearing lands. Where does that preference come from? It seems to contradict the life/non-life dividing line that is often argued for.
  2. Generally, the non-economic perspective do not articulate substantive standards, while the economic perspectives do, which is why they are privileged over non-economic perspectives. Such standards are necessary to affect practical application.

'

Risk Assessment and Risk Management[edit | edit source]

  1. Policy Issues''''
    1. Should RA be separate from RM? Per William Ruckelshaus (p53), this would decrease the political pressure placed upon non-political actors, including scientists, who engage in risk assessment.
      1. Political decisions should not affect the risk assessment stage. In order to effect this, generic policies should be adopted that constrain individual assessor’s discretion.
      2. Problem: '
        1. Since scientists report to leaders who are appointed by politicians, how will the scientists be insulated from political influence?'
        2. If Congress or another agency determines that a “carcinogen” must be regulated in a certain way, then interest groups will seek to influence the risk assessment process.'
        3. Consider where one study indicates a substance is a carcinogen, but another does not. Who should decide whether to label it a carcinogen?'
        4. With a generic policy, assessors may not be able to indicate differing weights of different studies, and good decision making may be hampered.'
  • How to resolve uncertainties in the risk assessment process? '
    1. (see c. and d. above and (ii)4 below)'
  • Lay perception vs. expert perception
    1. Stephen Breyer (p83) points out cognitive issues that affect lay perceptions of risk.
      1. Perceived risks: newer technologies, flying in an airplane compared to riding in a car, catastrophic events
    2. Should regulation address expert perceptions or lay perceptions?
      1. Consider that the (mis)perceptions themselves may cause adverse health effects.
      2. Studies have shown that the gov’t tends to address areas of public concern more so than expert concern.
    3. Risk Assessment (re: cancer) (p69)''''
      1. Hazard identification: determination of whether an agent has adverse health effects.
        1. Epidemiological (population) studies:
          1. Attempt to establish associations between human exposure to a particular agent and the frequency of cancer among the human population.
          2. Expensive and difficult method because sample sizes and data sets are inadequate, such studies lack adequate control groups. Cancer has a long latency period, and exposure data is often complicated, uncertain, and poorly documented.
          3. Most reliable studies involve signature carcinogens (which cause unique and unusual types of tumors) or very potent carcinogens, as with catastrophic events.
        2. Toxicological (animal) studies:
          1. Animals are used because of their anatomical comparability to humans.
          2. Exposed to high doses in order to obtain results before the animals die of natural causes. Controversy as to applicability to low dose exposure in humans.
          3. Start off with a maximum tolerated dose (MTD), and subsequent doses are fractions of that - MTD/2, MTD/4.
        3. Dose-response evaluation – determination of how the likelihood of adverse health consequences is related to the level of agent exposure.
          1. A response threshold is the dose below which no adverse effect occurs. The population threshold represents the individual threshold of the most sensitive person in the population.
          2. Non-carcinogens: assumption of threshold level. Goal of studies is often to find what that level is. Work in bands, so you start with a high concentration and work your way down.
            1. LOAEL: Lowest observable adverse effect level
            2. NOAEL: No observable adverse effect level
            3. Threshold is somewhere between these. NOAEL used as the threshold; if you want a no-risk standard you add a margin of safety.
          3. Carcinogens: No-threshold model (conservative/protective decision): assumption that any exposure = risk. Although low dosages are not tested (due to drawbacks of animal research), their corresponding level of health effects are extrapolated from the curve at high doses.
            1. Least-squares regression to extend the line left. Assumption that this is linear b/c a linear curve is likely more protective than the actual curve, though opinion is not unanimous.
            2. How do we deal with margins of safety in these contexts?
              • (i) 95% confidence level as upper bound.
              • (ii) So the linear line is the MLE (maximum likely estimate), but you don’t use this to regulate. You use the upper confidence limit (UCL), which is typically 95% but could be higher.
            3. Policy decisions': whether or not something will be treated as a threshold contaminant, whether or not to use linear modeling, how conservative to be with confidence levels and margins of safety.'
              1. Problems:
                • (i) Risk assessors are inconsistent
                • (ii) Risk managers must make decisions about how best to expend resources, and multiplication of conservative assumptions makes the calculation of risk very imprecise.
              2. Exposure assessment: who is actually subject to what concentrations?
                1. Aggregate population risk (incidence)
                  1. For different exposure levels, how many people are going to be affected?
                2. Maximum Individual Risk (MIR)
                  1. The person most affected by an increase in level of this compound, whether due to their exposure level or their sensitivity.
                  2. If sensitivity is uniform, which the EPA assumes due to lack of data, then MIR = Maximum Individual Exposure (MIE).
  • Risk characterization: final determination of the 1/xx risk of death (incremental risk – incremental life risk of cancer due to a particular agent at a particular level of exposure).'
  • Case: 'The Benzene Case [Industrial Un. Dept., AFL-CIO v. Am. Petroleum Inst., U.S. (1980), p59]: '
    1. Facts: OSHA lowered allowed limit of benzene in the workplace from 10 ppm to 1 ppm. There was some evidence that risk existed between 1 and 10 ppm, though it wasn’t systematic. OSHA shifted burden to industry and required them to prove that anything about 1 was safe.
      1. OSHA has a generic policy where “carcinogens” are considered no-threshold substances where no safe exposure level can be determined and the substances must be regulated at the lowest technologically-feasible level that will not cripple the industry.
      2. Went to 1 ppm because they believed that was the lowest feasible solution that would not destroy the industry.
      3. Statute: OSH Act:
        • (i) 3(8) – “’occupational safety and health standard’ means a standard which requires conditions ... reasonably necessary or appropriate to provide safe or healthful employment and places of employment.”
        • (ii) 6(b)(5) – “standards dealing with toxic materials or harmful physical agents ... shall set the standard which most adequately assures, to the extent feasible, on the basis of the best available evidence, that no employee will suffer material impairment of health. . . .”
      4. Disposition: Remanded back to OSHA.
      5. Reasoning [Stevens’ plurality]:
        1. OSHA provided no indication that it made a finding, based upon scientific evidence, that an exposure level above 1ppm posed a “significant risk.” (no dose-response evaluation).
        2. Finding a significant risk is a threshold determination which needs to be made in order to proceed to application of § 6(b)(5).
      6. Notes:
        1. The plurality relied upon § 3(8), but the dissent relied upon § 6(b)(5). Section 3(8) is a definitional section and more generic, while § 6(b)(5) is a substantive standard related to the particular type of substance at issue.
        2. [Rehnquist] concurred, holding the statute unconstitutional on non-delegation grounds. If this position were taken seriously, most environmental regulation would be unconstitutional.
        3. No precedential effect since plurality, not majority. However, cases after Benzene were able to survive Court scrutiny because OSHA quantified the risk with a dose-response evaluation. For example, a similar regulation was upheld in Pub. Citizen Health Research Group v. Tyson, 796 F.2d 1479 (D.C. Cir. 1986), p78, because OSHA used a dose-response curve to make its decision.
        4. Although the Court required a seemingly-scientific rationale for Agency action, as discussed above, many policy decisions affect the risk assessment.
      7. Risk Management' (p91, Lester Lave)'
        • Market regulation (e.g., setting minimum safety standards)
          1. When “market inefficiencies” or imperfections (e.g. incomplete information, transaction costs, externalities) exist, the government regulates in order to address the inefficiencies -- e.g. by taxing a relatively risky product so that the price accounts for the risks
          2. Market regulation will only work with voluntary transactions.
          3. Example: If the gov’t sets minimum safety standards, the cost of all automobiles increases and may increase beyond the cost that a poor person would negotiate in the absence of such standards. Ignoring externalities, such as danger to others or medical costs borne by others, the gov’t has less reason to regulate b/c individuals presumably internalize their personal risks.
  1. No-Risk Framework
    1. Example: the Delaney Clause of the Food, Drug, and Cosmetic Act requires exposure to no additional or unnecessary risk.
    2. Not actually implemented. No-risk frameworks have evolved into “negligible risk” frameworks (1/1,000,000).
  • Technology-based standards
    1. Typically “best available technology” (BAT) standard, with costs taken into account so as not to bankrupt an industry.
    2. Benefit: does not require a formal cost-benefit calculus.
    3. Actually involves many implicit policy decisions due to uncertainty of economic effects and other factors.
    4. In practice, standards are not frequently updated due to bureaucratic inertia.
    5. Popular view – technology-based standards will be more stringent than standards set by cost-benefit analysis. Revesz disagrees b/c cba could lead to the conclusion that an industry is not worth its risks. Furthermore, an implicit cba occurs under the tbs framework – regulators rarely impose standards that will come close to endangering the existence an industry b/c of the uncertainty as to what cost level would be so endangering.
  • Risk-risk
    1. Direct: Compare (i) risks from not using a tool vs. (ii) risks that will result from using the tool – e.g. a food additive that prevents botulism but is carcinogenic to consumers.
      1. Countervailing benefits (e.g. reduction of heart disease risk by taking aspirin for headaches) are not accounted for, even though countervailing risks are.
    2. Indirect: Consider also the attenuated risks associated with using an agent – e.g. workplace hazards caused by production of a food additive (reduction of risk to consumers leads to an increased risk to workers).
      1. Countervailing risks may have been internalized – e.g. by negotiation of hazard pay (a risk premium) – which will affect how a regulatory scheme should address the risks.
    3. Drawback: Not truly a risk management framework b/c does not provide guidance for how to decide whether or not to regulate (only how to choose between different regulatory proposals).
  • Risk-benefit
    1. Allows for consideration of nonhealth effects.
    2. Not clear how to determine value for lives or other risks; Even though factors aren’t valued explicitly, decision makers will determine value implicitly (and less transparently).
  • Cost-effectiveness
    1. “Getting the most bang out of every buck.”
    2. Differs from cost-benefit b/c cost-effectiveness just looks for least costly method.
    3. Not truly a risk-management framework b/c it does not answer the question of what is the optimal level of regulation.
  1. Regulatory budget
    1. Sets a maximum economic effect for the economy as a whole with the goal of saving as many lives as possible, given the limit.
    2. Not truly a risk-management framework b/c it may not allow an agency to determine the optimal level of regulation.
  • Cost-Benefit
    1. Attempts to quantify all costs and benefits such that they can be compared and net benefits can be maximized.
    2. 'Valuing lives:'
      1. VSL (value of statistical life) really means the “value of reducing one’s chance of death.”
        • (i) Problem: Does not account for future lives. See p122.
      2. Methods of valuation
        • (i) Stated preference – contingent valuation methodology (CVM) – “How much would you pay to reduce your risk of death?”
        • (ii) Revealed preference – e.g. wage differentials in risky workplaces – What is the risk premium paid to endure certain risks, attempting to control for other relevant factors?
          • (1) (*not generally studies of carcinogenic risk due to difficulty of obtaining data – b/c people move around; however, environmental regulation focuses upon carcinogenic risk)
          • (2) When workers have no real choice, they are less likely to demand a wage premium.
          • (3) People are less risk-averse than the general population, so companies are less likely to need as much of a wage premium.
  1. Valuing non-life factors:
    • (i) Use values
    • (ii) Option values (potential for future use)
    • (iii) Existence values (valuing a resource’s existence even if no plan to use) – methodology needs to present consequences in order to be accurate.
  2. Application of CBA:
    1. EPA’s temporary suspension of the use of certain pesticides under FIFRA relied upon a reasonable cost-benefit analysis, as required by the statute. EDF v. EPA''' (CBA under FIFRA)(p108).
    2. Conversely, EPA’s prohibition of asbestos under TSCA was not based upon a reasonable cost-benefit approach, as required by statute. Corrosion Proof Fittings v. EPA (asbestos CBA)(p113).
  3. Distribution of Environmental Risks (p126)
    • Distributional Issues
      1. Many individual regulations limit lifetime individual exposure, but nothing limits the aggregate exposure (combined effects of different agents).
      2. Communities of color and poor are affected disproportionately by environmental risks (more susceptible to health effects and more likely to be exposed to harms); (see p147 re: Superfund).
        1. But communities may “choose” certain risks (see discussion p156 re: Shintech)
          • (i) Counter: What about misperceptions of risk, disproportionate political power, or paternalistic concerns (e.g., not allowing poor communities to accept “inadequate” or even any compensation in exchange for health risks)?
        2. Even a fair initial distribution could become unfair over time in the absence of a comprehensive wealth redistribution plan.
          1. Market Dynamics Theory (Vicki Been, p134): Siting a LULU (locally-undesirable land use) can influence wealthier individuals to move out of a community and poorer individuals to move in to a community (due to reduced value and cheaper options).
          2. Studies have not addressed which came first –the siting of the LULU or the people of color and the poor.
            • (i) If the sitingà remedies would have to address housing discrimination, poverty, and free market dynamics.
          3. How to determine what is a fair distribution? (p127 and p141)
            1. Pattern of distribution? (e.g., proportional burdens, progressive burdens, proportional probability of burdens, or proportional compensation)
              • (i) Perhaps concentration of LULUs is more optimal than distribution, though, and a proportional distribution would decrease overall welfare.
            2. Cost-internalization? (Generators should pay full social cost.)
              • (i) How could this be implemented?
            3. Process? (e.g., if procedure agreed upon ex ante, then fair regardless of outcome).
              • (i) But how to ensure process is fair and not influenced by disparate wealth?
            4. Misperception that cba does not take into account distributional concerns, although doing so is an operational difficulty.
            5. Addressing distributional disparities through wealth distribution
              1. Many arguments re: environmental distribution ultimately involve disproportionate wealth distribution.
              2. Solution: progressive taxation.
              3. Critique (p133): Wealth may not be the only disparity. Just b/c wealth/life expectancy correlated—not necessarily a causal link.
  • Legal Theories for Addressing Disproportionate Risks
    1. Equal Protection Clause of the U.S. Constitution (p149)
      1. Elements:
        • (i) persons similarly situated are treated differently and
        • (ii) the defendants intended to discriminate
          • (1) Arlington Heights5-factor test, p151
  1. Problem: Discriminatory intent is nearly impossible to prove, as demonstrated in E. Bibb Twiggs Neighborhood Ass’n v. Macon-Bibb County Planning & Zoning Comm’n, M.D. Ga. (1989), p150.
  2. Civil Rights Act, Title VI (p154)
    1. Prohibits discrimination on the grounds of race, color, and national origin by “any program or activity receiving Federal financial assistance.”
    2. Only enforceable by the EPA b/c Supreme Court found no private right of action. Alexander v. Sandoval (Civil Rights Act – no private right)(p155).
    3. Takeaway: The EPA is the important actor, not Congress or the courts.
  3. Executive Order 12,898 (p155)
    1. Not really had much effect, especially compared to the cost-benefit EO.
    2. Possible explanation: The cost-benefit EO established an enforcement mechanism through OMB and OIRA. EO 12,898 applies to every agency and does not have a mechanism for enforcement.
  4. State enforcement agencies and legislation

Regulatory Tools[edit | edit source]

  1. Command and control regulation' (p161)'
    • Typically takes form of BAT (best-available technology) standards.
    • Set stricter limits on more prosperous industries, those with less competition or fewer substitutes.
    • “Best” available technology = best you can get w/o destroying the industry; may destroy individual firms.
    • BAT standards, generally
      1. Equipment/design standards': Require a specific technology.'
      2. Performance standards: Determine BAT and level of emissions if BAT were used; but a company can use whatever method it wants to achieve the emissions level. *the usual statutory standard, except where emissions cannot practically be measured.
      3. Firms prefer:
        1. Performance standards (2) b/c
          • (i) They provide more options. A company can use BAT or can use something else, including changing its production process.
          • (ii) *Cost-benefit analysis may over-estimate cost of compliance due to this phenomenon of production-process change.
          • (iii) Companies don’t have to wait for new technologies to be approved to use them.
        2. Regulators prefer:
          1. Design standards (1) b/c
            • (i) easier to set;
            • (ii) easier to enforce.
              • (1) BAT standards (2) are generally set w/reference to input/output, so you actually have to be measuring a proportion.
              • (2) Regulator does not measure every day.
  2. Or performance standards (2) b/c
    • (i) design standards may require testing of emissions anyway to ensure proper implementation and use, and
    • (ii) performance standards encourage companies to be creative.
  • Comparing BAT regulation to Market Permit and Tax Scheme (See Ackerman & Stewart (p161))
    1. No incentive to reduce pollution once the BAT standard is met. Not only no incentive for firm to use it, no incentive to do the R&D to use more cost-effective measures.
      1. Market permits solve this problem by encouraging companies to reduce pollution as long as doing so is cheaper than buying a permit.
      2. Taxes solve this problem by encouraging companies to reduce pollution as long as doing so is cheaper than paying the tax.
    2. Misallocation of pollution control burden, ignoring variations among plants and industries, thus wasting money (it does not minimize aggregate costs by equalizing marginal costs – the cost to reduce by one more unit of pollution).
      1. Market permits solve this problem by allowing companies to pay the same price for the last unit of pollution (the permit amount).
      2. Taxes solve this problem by allowing companies to pay the same price for the last unit of pollution (the tax amount).
    3. Penalize successful and new products and processes. New sources are not at risk of shutdown, because they’re economically healthier, must run a lengthy gauntlet of approval processes which discourage investment, place stricter reqs on successful businesses because they can bear them.
    4. Provide a lot of points of legal vulnerability, making it often more cost-effective to litigate than to comply.
    5. Doesn’t allow an agency to set intelligent regulatory priorities.
  1. Marketable permit schemes (p159, 168-)
    • How they work:
      1. The regulator’s role:
        1. Defines the markets, sets the aggregate pollution level.
        2. Allocates initial permits.
        3. Registers transactions.
      2. Allocation
        1. Initial auction. What’s good about this?
          • (i) Doesn’t favor existing over future firms.
          • (ii) Most economists favor this b/c it will probably have lower transaction costs.
          • (iii) Provides revenue
            • (1) Problem: As with tax scheme, what to do with the money?
  2. Grandfather existing polluters but allow for trading
    • (i) Hurts later entrants. Barrier to entry.
    • (ii) More politically palatable (industry supports regulation b/c it obtains a tradable property right).
    • (iii) Most common method b/c of (ii).
    • (iv) Major problem: Creates incentive to pollute prior to allocation, although this could be addressed by using historical levels from, e.g., 2 years ago.
    • (v) Coasian analysis: Doesn’t matter where the entitlement is set, if transaction costs are low.
  3. The first unit sold: what would the firm think about when deciding to sell a permit? Cost of reducing pollution one unit.
  4. Buyer? Cost of reducing pollution versus cost of one-unit permit.
  5. When will the trades end? When the cost of reducing one unit more exceeds the cost you could get for the permit. You’re down to the more expensive ways of doing it. Decreasing marginal utility w/regard to pollution reduction.
  6. Once you’re at equilibrium amongst existing players, what happens to change the market?
    1. New producers enter.
    2. Someone wants to increase output.
    3. Deteriorating equipment
  • Benefits of marketable permit schemes
    1. Level of pollution is set at a particular level (the optimal level).
    2. Incentive to innovate:
      1. So long as the cost of developing new technology is less than the price of a permit, it’s worth doing. It’s almost like you get into the pollution-reduction industry.
    3. Eliminate the information-processing tasks agencies face
      1. Wouldn’t have to determine what the BAT was, the business would do that.
    4. Free up financial incentives and would raise money when permits are sold at auction.
    5. Do away with failure to enforce.
      1. Companies’ expectation of enforcement would influence the price that they pay for permits; failure to enforce would leave to lower auction revenue, therefore it would become a priority for the gov’t to address.
  • Disadvantages of marketable permits
    1. What happens if non-industry players buy permits to retire them?
      1. Input monopolies! FTC gets involved. Antitrust fiesta! Not plausible that it would reach a level where the impact on the market would be significant.
    2. Hot spots:
      1. Example: nine sources in a jurisdiction, ambient standards, each polluting ten units. 90 total units. Then you go to marketable permits and they all get 10 permits. Trading occurs, one firm has 81 permits, the others each have one.
      2. Are we better off or worse off? Costs have been minimized. But it depends on the curve of the pollutant. What if it’s no-threshold? If it’s convex, we’re worse off because every unit added at the top dramatically increases harm. If it’s concave, then the rise in harm comes early.
      3. In a BAT w/out ambient standards, there could be hotspots, but the claim is that marketable permit schemes exacerbate the problem.
      4. So weakness in Ackerman/Stewart argument is that it doesn’t account for the disparate environmental impacts of this (they want to do away with ambient effects as well).
      5. Remedies for the hot spot problem
        • (i) Ambient standard overlay. So trading area would have to be larger than attainment area. uses mathematical modeling to figure out impacts everywhere else of pollution at one point. You could do this through a website!
        • (ii) Size of markets.
        • (iii) Market in units of environmental degradation (p169n7). Problem is that you have an impact at an infinite number of points. Plus, you have an impact in multiple markets.
      6. This only matters for local pollutants—where the effect is concentrated in the area where the pollutant is emitted. Regional/global pollutants different.
  • The moral dynamic
    1. W/marketable permit schemes, you can pollute a lot more—you don’t have to reduce at all—so long as you can buy permits. Are marketable permit schemes, thus, a morally undesirable “license to pollute”?
    2. [this is only an issue if you treat all pollution as a moral ill. If the only “bad” pollution is overpollution, then you don’t have this problem]
  • Should environmental groups be able to buy and “retire” marketable permits? (p170n8)
    1. Depends upon how the level of pollution allowed was set.
      1. If agency tries to figure out the optimal level of pollution, then allowing people to take them out of circulation damages the social calculus.
    2. If someone tried to buy up all of them, the danger is that the government would just issue more
    3. Risk to national productivity.
  1. Effluent fees (pollution taxes) (p160, 172)
    • Tax assessed per unit of pollution.
      1. Has to be set at optimum level in order to achieve desired result (optimal level of pollution).
        1. Continuing inflation will erode costs and prices will have to be reset, which is difficult.
  • Markets v. taxes (see Baumol & Oates, p172)
    1. If you want to hit a specific performance target, then marketable permits can get you there easily. But while the government may make guesses about the relationship between taxes and level of pollution, you can’t be sure with effluent fees.
    2. Social welfare might not be maximized if the proper number of permits are not allocated (polluters will pay too much b/c the cost of compliance was underestimated) or if the tax is not set at the right amount (polluters will pay the tax and pollute more than they should if the tax is too low). [Taxes control price but not quantity; permits control quantity but not price.]
  • When a tax might be preferable
    1. When there are high transaction costs in trading permits. In tax system you have to measure emissions, create collection system, etc. Remember, though, that you’ll have to measure emissions no matter what. Experience w/running tax systems hasn’t been that great.
    2. When you want to generate money for the government. Do we want this to happen? Presumably, taxes would be lowered elsewhere. Do you prefer the gov’t to tax to internalize externalities, as opposed to some other way?
      1. You could do this by initial auction as well.
    3. When the industry is cost-sensitive (b/c taxes will impose a known cost rather than unpredictable costs of a permit scheme).
  • When a market permit scheme might be preferable to a tax scheme
    1. When regulating a threshold pollutant.
    2. If an industry is struggling.
    3. In order to get support from existing firms, you’re going to have to do this in a way that doesn’t impose initial costs.
    4. When a tax scheme might generate large amounts of revenue, which would create political difficulties in how to use the $$.
    5. When geographic flexibility is desired.
  1. Deposit-refund schemes (p169, 189)
    • Combines tax at purchase and subsidy at return/proper disposal.
    • Good in cases when it’s very hard to enforce a disposal requirement (like for batteries)
  2. Why not just have a tax?
    1. In the case of batteries, a deposit-refund scheme can ensure that you’re paying for proper disposal, rather than amelioration. You don’t want to create a disincentive for using the battery, but rather for disposing of it incorrectly. Even if use is reduced, disposal may still be inappropriate.
    2. So why don’t we have one for batteries? Revesz believes the political will does not exist.
  • How do you set the optimal level of a deposit-refund scheme?
    1. Issue of what upfront investment a purchaser can afford.
      1. Poor and large families may be hardest hit.
      2. Also, concern for theft and counterfeiting exist.
    2. How easy is it to return the item? Think about an elderly woman in a 4th floor walkup.
  1. Liability rules (ex post) (p160, 195)
    • Incentive to reduce risk based on likely liability. (A negligence standard only transmits proper incentives if it’s set at the right level.)
    • Problem: not all harm will produce a lawsuit.
      1. Company may be out of business or insolvent. If companies know they may become insolvent, they will not be incentivized to internalize the externalities, but companies don’t usually plan to be insolvent. On the other hand, companies may declare dividends and put money in the hands of the shareholders in order to reduce money available for payment. Law cannot get at this, though it can get through to parent companies by “piercing the veil.” Law in the US does not reach shareholders.
      2. Harm doesn’t show up for a long time, no signature disease, collective action issues, causation problem, etc.
        1. Causation difficulties
          • (i) Probabilistic liability/proportional causation above a certain threshold
          • (ii) Market-share liability
        2. Timing: statutes of limitations and how they are construed. Should statute begin to run at the time of the discovery of the harm (more modern rule) or at the time of exposure (traditional rule)?
        3. Poor incentives for litigation: free-rider-type problem - high costs, low individual benefits.
          1. Addressed by class action suits.
        4. Choosing between regulatory vs. liability schemes (p160):
          1. Information: Private actors have better information than regulatory authorities about an activity’s riskà liability preferable.
          2. Judgment-proof problem: Actors causing risk are likely to have insufficient assetsà regulatory.
          3. Effectiveness of remedy: Low probability of lawsuits despite harmà regulatory
          4. Consider administrative costs—likely to favor liability (p199).
  • Addressing solvency issues
    1. You may be able to pierce their corporate veil if they’re organized this way, but they usually aren’t.
    2. However, not all companies are going to be willing to operate under the structural constraints necessary to protect themselves from liability in this way. Signaling long-term presence—not everyone’s willing to go bankrupt.
    3. Insurance. You can mandate insurance, or you can just depend on the market, although the market may not exist..
    4. Bond. This is requiring in some circumstances for hazardous waste companies.
  1. Informational approaches
    • What is the best context for using informational approaches as opposed to something else?
      1. When the employee/user/consumer would be able to avoid the harm with proper warning.
      2. When the consumer would be able to rationally choose among different levels of risk because the increased risk would be worthwhile to them in certain circumstances. Honor people’s risk preferences within a certain range.
    • Problem is that people ignore warnings. Construction of appropriate warnings.


Political Context for Environmental Regulation[edit | edit source]

  1. Constitutional Limitations on Federal and State Power
    • Limitations on Federal Power
      1. The commerce clause—U.S. Const. art. I § 8 (power “[t]o regulate Commerce . . . among the several States . . .”) is very broad but has outer limits. United States v. Lopez, U.S., p206.
      2. The Tenth Amendment reserves nonenumerated powers for the states (“The powers not delegated to the United States by the Constitution, nor prohibited by it to the States, are reserved to the States respectively, or to the people.”).
        1. As a result, Congress can encourage states to regulate a particular field or a particular way but cannot require the states to do so. New York v. United States (federal objective/state means), U.S. (1992), p207:
          • (i) Facts: Congress sought to encourage states to take care of their low-level radioactive waste via (i) economic incentives, (ii) access incentives, and (iii) a “take title” provision. The State of N.Y. sought declaratory judgment that the statute violated the 10th Am. The suit was dismissed (in favor of the federal gov’t). Affirmed on appeal, but partially reversed by the Supreme Court.
          • (ii) Holding: The take title provision violates the 10th Amendment b/c it requires states to regulate in one of two particular ways which individually would be improper.
          • (iii) Reasoning:
            • (1) Because space in radioactive waste disposal facilities is frequently sold by residents of one state to residents of another, the Commerce Clause allows regulation of the interstate market in waste disposal.
          • (iv) Notes:
            • (1) The federal gov’t has figured out how to avoid the problem of NY v. USby allowing states a choice of regulating or not. Even though the federal gov’t will step in (in some vague, unspecified way) if states do not regulate, this is still a “choice” that states can make.
  2. The Eleventh Amendment grants sovereign immunity to the states (“The Judicial power of the United States shall not be construed to extend to any suit in law or equity, commenced or prosecuted against one of the United States by Citizens of another State, or by Citizens or Subjects of any Foreign State.”).
    1. The Supreme Court has interpreted this to prevent federal suits by citizens of other states as well as citizens of the state itself.
    2. The federal gov’t can waive this immunity but must do so pursuant to the Fourteenth Amendment (due process).
    3. Thus, Congress is limited in its power to allow private suits against the states.
  3. According to Revesz, while there are restraints on federal power, the restraints are relatively limited. For this reason, most of the discussion about federal vs. state regulation comes down to policy questions.
  • Limitations on State Power
    1. The Supremacy Clause
      1. A federal statute (e.g., the Clean Air Act) can explicitly state that it preempts state law.
      2. Field occupation – implicit preemption.
      3. Conflict preemption – A state regulation cannot conflict with federal law.
    2. The Dormant Commerce Clause
      1. States cannot burden out-of-state competition.
      2. Thus, states cannot reserve hazardous waste sites for only in-state waste.
    3. Policy Arguments re: Federal Regulation vs. State Regulation''''
      • In favor of Federal Regulation
        1. Race to the bottom (p248)
          1. If states could collude, they would implement relatively stringent standards from industry to achieve maximum social welfare re: environment and health. Our legal system does not allow such collusion b/c the benefits presumably are outweighed by the social costs to consumers. (Generally, it’s not a good idea to change an entire market for a localized inefficiency.)
          2. As a result, a prisoner’s dilemma situation occurs, and states implement suboptimally lax environmental standards in order to compete for industry, dropping standards until an equilibrium is reached where sufficient benefits are no longer achieved to justify the cost.
          3. Revesz argues that, despite theoretical models that attempt to demonstrate otherwise, none support the claim that, without federal intervention, states will not race to the bottom over environmental standards.
          4. Furthermore, even if the federal government mandates environmental standards, states could compete in other areas that decrease social welfare. Thus, the only solution would be to federalize all standards (labor, tax, etc.), which is not a palatable solution.
          5. A lot of arguments about the race to the bottom center on other issues (interjurisdictional competition, for example) in the end, and the race to the bottom issue doesn’t really add anything.
          6. In the end, however, competition will not necessarily lead to suboptimally lax regulation. It depends on where the bargaining power is. States can and do regulate too strictly. NIMBYism.
        2. Interjurisdictional externalities (spillover) (262)
          1. Jurisidictions don’t care about the harms imposed upon other jurisdictions. They gain economic benefits from having the particular industry in their jurisdictions, and they only see some of the costs.
          2. How strong is this argument? Jurisdictions have been found to play around with pollution in order to try to send the impact out of their borders—e.g., increased stack height.
          3. Example of clean air regulation: we have two types of standards, ambient and emissions.
            • (i) If use ambient standards, you’re probably going to need federal standards because it’s hard to prevent other jurisdictions from violating yours.
              • (1) You need some sort of allocation mechanism for figuring out who can meet them more cheaply if you want to use caps.
            • (ii) Emissions standards: because these don’t limit aggregate pollution, you may still have suboptimal transboundary effects.
          4. Interjurisdictional spillover is frequently given as a rationale for federal regulation, but the existing federal mechanisms don’t actually do a lot to take care of this. Unless it’s actually effective, can it really be a good reason for federal regulation?
        3. Economies of scale
          1. Uniformity of standards
            • (i) National standards lead to economies of scale in production processes because companies do not have to tailor their processes to meet many different state standards.
            • (ii) But it’s not clear that for items such as power plants, which are not mass-produced, that this argument applies.
          2. Economize resources
            • (i) Scientific research—e.g., is a substance a carcinogen?—can be done on the national level and not be duplicated in several jurisdictions.
            • (ii) Notice-and-comment rulemaking can occur in a centralized manner rather than in several different jurisidictions.
          3. Counter:
            • (i) Competition and smaller operations have benefits as well. High levels of centralization are not always more economical.
            • (ii) Information could still be centralized, while other functions are given to states.
          4. Quasi-constitutional argument
            1. As citizens of the United States, you have a right to a certain level of environmental quality that your state can’t take away from you. Analogous to your right to vote.
          5. Public choice concerns
            1. Environmental groups are more easily represented at the national level, due to the high costs of information and of passing state-by-state laws. Industry groups also operate at the national level.
            2. If you see environmental regulation as a product of a clash between environmental groups and industrial groups, why do we actually get environmental regulation at all?
              • (i) What about free-rider problems? You have heterogenous interests in the environmental community, you have a big aggregation, etc.
              • (ii) However, environmental groups may have great economies of scale in organization. It would be hard to participate in legislative process in the fifty states, but they might be able to get the resources together in one situation.
                • (1) What about money? Maybe you need a certain threshold of money to be effective, and once you’re past that you’re not that much more effective.
                  1. In some cases, this holds true, and in others it doesn’t. The more you can contribute to campaigns, the better you likely are.
  1. Arguments against federal regulation (and in favor of state regulation)
    1. Different regional preferences, self-determination
      1. (Per Revesz’ research, regional differences re: environmental issues outweigh party differences)
      2. But federal regulation could account for this.
        • (i) Counter: access to information is more difficult for the federal government, and political pressure is high not to favor certain areas.
      3. Differences in cost/benefits
        1. Based upon regional geography, climate, demographics, etc., it may be very expensive for a particular state to reduce pollution to the level mandated by a national standard.
        2. However, federal regulation could account for this.
          • (i) Counter: (See above).

The Clean Air Act[edit | edit source]

  1. Criteria and hazardous pollutants
    1. Criteria pollutants regulated under § 108-109 (p. V-24).
      1. Pollutants that reasonably may be anticipated to endanger public health or welfare
        1. Are present in the air due to numerous or diverse mobile or stationary sources.
        2. For which air quality criteria had not been issued before enactment of the CAA, but for which the Administrator plans to issue criteria.
        3. Example: B/c EPA made findings of adverse health effects of leadà must list lead as criteria pollutant under § 108. NRDC v. Train (lead case), (2d.), p317.
      2. Once listed, Administrator has 12 months to issue criteria. Criteria are “technical documents that identify the public health and welfare effects of the pollutant” (p. II-1).
      3. When CAA passed, criteria had been promulgated for sulfur oxides, PM, CO, hydrocarbons, and ozone. Nitrogen oxides added in 1971; Pb after citizen suit in 1976. No additional criteria pollutants have been added.
  • Hazardous pollutants under § 112.
    1. 112(b): Initial list of hazardous pollutants
    2. Can be revised, but you can’t add criteria pollutants to this list.
  1. NAAQS (§ 109), p315
    • Administrator to issue primary and secondary NAAQS simultaneously w/issuance of air quality criteria.
    • Administrator to propose NAAQS for pollutants for which air quality criteria were issued before CAA.
  2. Primary NAAQS: set at a level which, in the judgment of the Administrator and based on the air quality criteria, is requisite to protect the public health and allows an adequate margin of safety. § 109(b)(1)
    1. Cost of compliance is not to be taken into account b/c the statute does not indicate as such. Whitman v. Am. Trucking Ass’ns, U.S., p326 (see Cases). [D.C. Circuit had already held that the EPA does not have to consider feasibility. Lead Induss. Ass’n (see Cases).]
      1. Breyer concurrence (explained on p335): The statute does not require elimination of all risk. Thus, cost is implicitly allowed to be a consideration in setting the NAAQS.
      2. Notes:
        • (i) Breyer’s health-health tradeoff argument: by making this more expensive, you’re decreasing collective health.
          • (1) But health-wealth correlation does not prove causation.
          • (2) And how is the cost distributed? If most of the cost is borne by very wealthy people, it’s not going to impair their health. How regressive is your tax system? That’s the question.
        • (ii) Because CB can’t legally be considered, there is a lack of transparency in decision-making. Notice and comment can’t really work if you can’t see the agency’s full basis for its policy. Better to have explicit CBA.
  • Secondary NAAQS: set at a level which, in the judgment of the Administrator and based on the air quality criteria is requisite to protect the public welfare from any known or anticipated adverse effects associated with the presence of such air pollutant in the ambient air. § 109(b)(2)
    1. Welfare: effects on welfare are “effects on soils, water, crops, vegetation, . . . , weather, visibility, climate, property, hazards to transportation, effects on economic values and personal comfort and well-being.” § 302(h) (general provisions) (pV-279)
    2. Key distinction is between public health and public welfare. Welfare covers a wider range of things, including economic stuff—basically, everything except human health, which falls into “public health.”
    3. Primary stds are not necessarily more stringent than secondary. See chart on p316, which shows for SO a more stringent secondary std.
  • Basically, notice-and-comment rulemaking
  1. SIPs—State Implementation Plans § 110
    • Generally: “an experiment in cooperative federalism: the federal government provided a goal for air quality and the state retained discretion to achieve that goal in a manner consistent with state and local priorities” (p340).
    • Required elements of a SIP: § 110 (a)(2)
      1. Coherent plan for meeting the NAAQS and PSD or nonattainment standards through enforceable emissions limitations.
  2. Approval of a SIP (§ 110(k)): Administrator must approve the plan if it meets the requirements; if not, a disapproval of a part counts as a disapproval of the entire plan.
    1. Measures exist for conditional approval, which converts into disapproval if the conditions aren’t met.
    2. If the Administrator finds that a SIP is inadequate to attain or maintain the NAAQS, then she can issue a SIP call, which apprises the state of the deficiency and requires a SIP revision. State has 18 months. § 110(k).
  • EPA cannot reject a SIP due to technological or economic infeasibility. Union Electric Co. v. EPA, U.S. (1976), p341:
    1. P’s statutory argument: “as expeditiously as practicable” means you have to consider feasibility
    2. Court says that the phrase just means that you achieve the goal faster than three years if doing so is possible
  • States are allowed to establish state regulations that are more stringent than federal NAAQS, per § 116, but the SIPs must comply with the NAAQS. Id.
    1. Court: Look at § 116—states are allowed to set stricter standards if they want to! Why not let them do it through the SIP?
      1. Why is § 116 in here? Federalism. If states want to have stricter standards, they ought to be allowed to—it doesn’t harm other states in the way that having a weaker standard does (i.e., race-to-the-bottom/interstate externalities).
        • (i) Exemption for vehicle standards, except in cars—there is preemption of auto regulations, except in CA.
      2. Pros and cons of having the state to set the standards, rather than the federal government:
        • (i) States can more cheaply determine the best ways to meet the NAAQS, given their regional particularities.
        • (ii) But state politicians can use the system to punish opponents and reward supporters.
        • (iii) On the other hand, states can better determine which workers are less reemployable in the area and make decisions that affect them less.
      3. Why might it be easier for a state to promulgate stricter standards under a SIP than under state law?
        • (i) They’re a bit insulated from local political processes—they’ve already got to make the plan, so it’s easier to move from that to a stricter standard. (“The Feds made us do this.”)
        • (ii) Makes enforcement by private plaintiffs easier—because they’re violating federal law and can thus be sued in federal citizen-suit actions.
        • (iii) What about enforcement by the government? Fed gov and state can both enforce the SIP.
      4. Union Electric’s best argument was probably that states have avenues other than SIPs for implementing more stringent standards. However, if Union Electric had prevailed in this case, the EPA would have a much more complicated task of evaluating SIPs to ensure they meet a floor and a ceiling. This would have opened up the approval process to a multitude of challenges.
    2. Justice Powell (concurrence):
      1. Congress may not have understood the “Draconian possibility” that could result from the CAA.
        • (i) But couldn’t a state decide to shut down a utility regardless of the CAA?
      2. Federal Implementation Plans:
        • Administrator must promulgate a FIP within two years of the disapproval of a SIP, unless the state corrects the deficiency. §110(c)
        • SIP not implemented à federal gov’t required to implement FIP. §110(c), Coalition for Clean Air v. EPA, (9th), p348.
          1. Facts: No SIP for South Coast Air Basin in California despite years of litigation. Dirtiest air in the country. CA submitted a series of SIPs that the EPA didn’t approve b/c the SIPs couldn’t meet the NAAQS. EPA had proposed FIPs but w/drew them because of the potential “disruptive social and economic consequences” they would cause. CfCA sued, and EPA settled in 1989 with requirement to promulgate FIP. Then the CAA was amended in 1990.
          2. EPA eventually made an FIP. However, Congress wiped it out in a Defense appropriations rider.
            1. This was essentially a veto-proof bill, ensuring passage of the rider. Also, this type of rider received almost no scrutiny b/c the committee attaching it (a defense committee) was not an environmental committee. Lastly, a rider is typically attached at the last-minute, such that most Congresspersons do not know about it even though they vote on it.
            2. Then EPA rescinded the FIP, and California passed an SIP, which did not have to meet the NAAQS under the non-attainment provision. Thus, it took about 25 years for southern Cali to be covered by a SIP, even though the statute requires such in no more than 3 years.
  • Constitutional problem with FIPs:
    1. Can the fedgov require a state to pass a law? No. So EPA can’t, as part of a FIP, “commandeer” a state legislature. 10th Amendment. Generally, the SIPs aren’t considered commandeering because the feds will regulate if they don’t. Conditional funding is okay as well. An “option” for states, as opposed to a requirement.
    2. But if a FIP requires a state to legislate, that’s different. The fedgov can directly regulate individual polluters, but can’t mandate state legislation.
    3. Why would the state prefer to punt the responsibility to the federal gov’t?
      1. To pass the political price onto the federal gov’t. This was a situation where all the options were bad.
      2. But the federal gov’t passed on the responsibility as well.
        • (i) For one, federal enforcement would be difficult.
        • (ii) Secondly, if voters in CA don’t like the plan, federal politicians could suffer the fallout, possibly even leading to a change or repeal of the CAA.
      3. New Source Performance Standards § 111
        • Generally:
          1. Uniform, technology-based standards based on the best available technology: “…reflects the degree of emission limitation achievable through the application of the best system of emission reduction which (taking into account the cost of achieving such reduction and any nonair quality health and environmental impact and energy requirements) the Administrator determines has been adequately demonstrated.” § 111(a)(1).
          2. Administrator can vary the NSPS based on categories, classes, types, and sizes.
        • In setting a NSPS, EPA is not required to conduct CBA, but it is required to demonstrate a reasonable basis for its decision. Portland Cement v. Ruckelshaus, (D.C.), p356. (Cf. New York v. Reilly, (D.C.), p361 (EPA provided sufficient evidence to support its decision re: municipal incinerators).
          1. How do you take economic cost into account? Do you look only at cost to the industry itself? Social cost? What if an industry can’t survive in one region due to existing level of pollution or other things that make cost prohibitive, but is flourishing elsewhere? Can you regulate them into the ground since they’re alive elsewhere?
            1. What if there are close substitutes? Even moderate costs will cause the industry to collapse, because the substitute will suddenly become cheaper.
            2. What if there are close substitutes in one part of the country but not in another? Are these permissible costs under the statute?
              • (i) Yes, probably. Otherwise it’s got no teeth. And the EPA would have to demonstrate how it would work and effect competition everywhere, instead of more generally. Very difficult to do this.
            3. How did EPA “demonstrate” the technology here?
              1. EPA doesn’t have to go set every single existing plant in order to set a standard for new sources. But they needed to look at more than one. It’s really hard to argue that you can generalize properly from one—what if the one has some quirk or particularity?
                • (i) EPA could have defined the best way of controlling pollution as a dry-process plant w/this technology. They didn’t have an appropriately demonstrated technology for the wet-process plant.
              2. What if no existing plant could have met the standard? Could EPA have said, No existing plant could meet this but if you build a new plant around this standard, then it’ll meet it? And get away with no demonstration? Does it make sense to try to set a standard for new plants by examining old plants? Why does it matter if existing plants can meet the standard or not?
                • (i) Court says that EPA could have extrapolated from existing data and expert opinions to project what would be the best technology for new plants. But since they decided to base it on testing, then they have to do the testing properly!'
              3. Prevention of Significant Deterioration (PSD) (p368):'
                • Prohibits states from allowing air quality to degrade in relatively pristine regions. Initially created by regulation in response to Sierra Club v. Ruckelshaus (below). Then codified in 1977 at § 160 et seq. (p. V-118)'
                • Ambient component: a federally set baseline (169(4)) and an increment, which is set through a combination of federal and state programs. Program allows some degradation, but does constrain it. '
                • Emission component: '
                  1. New/modified major emitting facilities (MEFs): BACT. Standard set via combination of federal and state decisions under. '
                  2. Existing sources and non-MEFs: SIPs have to show how the state will take steps not to violate the PSD. § 110(a)(2).'
                • PSD permitting for new/modified MEFs: '
                  1. Categorize area as Class I, Class II, or Class III'
                    1. § 162: Entire country initially Class II, where deterioration consistent with moderate, well-controlled economic growth was acceptable.'
                    2. States could re-classify areas into Class I (typically national parks) or Class III via:'
                      • (i) effects analysis'
                      • (ii) public hearing'
                      • (iii) notice to affected land managers'
                      • (iv) Governor, legislature, and representative “general purpose unit” of local government have to approve a redesignation. Keeps a state from imposing this on a county, e.g. 164(a)'
                    3. Fed gov must approve, unless the state didn’t follow procedure or Class I reqs are violated.'
                    4. No states have sought redesignation from Class II to Class I, but some Native American tribes have. Redesignation to a more-protective standard is easier than redesignation to a less-protective standard.'
                    5. No state has ever redesignated an area to Class III. Why? Probably local governments would be very unhappy. Maybe the areas where industry wants to locate are already at a point where the NAAQS are going to constrain you. Maybe areas are clean for one pollutant but not for others—so say you can’t degrade further for SOx, because you’re in nonattainment for PM. Maybe they already get a lot of what they want from Class II designation.'
                  2. Ambient Requirement '
                    1. Baseline: § 169(4)—Baseline is set when someone applies for a permit for a major emitting facility, per § 165, in an area subject to that part. Defines current level of air pollution and provides basis for increments.'
                    2. + Increment: amount of deterioration toward the NAAQS authorized for a particular attainment area.'
                      • (i) PM and SO2 increments set by Congress'
                      • (ii) EPA has also set increments for NO2, but not for CO or hydrocarbons, which are primarily regulated via mobile source regulations elsewhere in the statute. '
                      • (iii) No increment for ozone, which is controlled via control of its precursors'
                      • (iv) EDF v. EPA (re: PSD increments), p375: EDF challenged the increment for NO2 on the grounds that EPA had merely set the same percentage increment as for SO2 and PM. Court said that this was a “contingent safe harbor,” but that an analysis under 166(c) was also required. In the end, EPA promulgated the same rule and allowed states to use alternative means of control, such as cap and grade under the CAIR. '
                    3. What happens if nobody asks for a permit for a while? Before the baseline is set, there’s not limit on degradation. '
                      • (i) Lots of sub-MEF plants built'
                      • (ii) Pollution from other areas. Until baseline is set, they can’t do anything about this! '
                    4. Why might it take a long time for an MEF to apply for a permit? '
                      • (i) No economic growth in the area. No local need for what they might produce, or no workforce/infrastructure available. '
                    5. Relationship of increment to NAAQS: '
                      • (i) Class I areas, roughly 2% of the NAAQS. '
                      • (ii) Class II areas, roughly 25% of the NAAQS. '
                      • (iii) Class III areas, roughly 50% of the NAAQS. '
                      • (iv) So the designation makes a big difference. '
                      • (v) Example: SOx: NAAQS are 80. Increment in Class II is 20. Will constrain only where the baseline is below 60. If baseline is higher than 60, NAAQS prevent you from going over 80 in any case! '
                      • (vi) So lots of areas w/air quality better than the NAAQS won’t experience additional constraint due to the PSD program.'
                    6. Emissions limits'
                      1. BACT (for new/modified MEFs). § 169(3):'
                        • (i) based on maximum emission reduction of each pollutant which '
                        • (ii) takes into account energy, environmental, and economic impacts and other costs, and'
                        • (iii) is achievable for such facility through application of production processes and available methods, systems, and techniques.'
                        • (iv) Cannot be less stringent than NSPS.'
                      2. Determined by state on case-by-case basis, but if not based on a reasonable analysis à EPA may take steps to ensure compliance w/ CAA. Alaska Dept. of Environmental Conservation v. EPA (below) (Alaska’s approval based upon a least-cost reduction standard rather than EPA’s top-down standard was criticized by EPA).'
                        • (i) “Top-down BACT”: EPA establishes a ranked list of control technologies and sources must take the top one unless they can demonstrate that on grounds of technical, energy, environmental, or economic impacts, this standard is not achievable. Then can use the next-best alternative.'
                      3. BACT contains all the elements of NSPS, although the language is different. '
                        • (i) BACT is determined on a case-by-case basis, unlike categorical NSPS. Thus, the facility at issue does not have to be representative of the industry and, if the source can meet a stricter standard, it doesn’t matter that others may not be able to.'
                        • (ii) BACT set by state-level authority while NSPS set by Administrator '
                        • (iii) If the MEF does not fall into a category that has an NSPS? A new NSPS is set. So, even though BACT cannot be less stringent than NSPS (§ 169(3)), it could be a lax standard if no NSPS is relevant that would be “technology-forcing.”'
                      4. Major emitting facility (MEF) definition:'
                        1. (see also section (ii) under New Source Review)'
                        2. '§ 169(1)—Facilities that require a permit from the state if they wish to locate in a PSD area. (See also '
                        3. “New” definition comes from NSPS: § 111(a)(2): any stationary source, the construction or modification of which is commenced after the publication of regulations.'
                        4. “Modification” (§ 111(a)(4)): physical chance in, or change in the method of operation of, a stationary source which increases the amount of any air pollutant emitted by such source or which results in the emission of any air pollutant not previously emitted.'
  • Sierra Club v. Ruckelshaus, U.S., aff’d by equally divided Court (1973), p369:'
    1. Led to statutory establishment of the PSD program. '
    2. Plain language argument: the Act authorizes EPA to “protect and enhance the quality of the Nation’s air resources.” '
      1. Basic purposes of the act under § 101(b). “On its face, this language would appear to declare Congress’ intent to improve the quality of the nation’s air and to prevent deterioration of that air quality, no matter how presently pure that quality in some sections of the country happens to be…”'
      2. Why is this a weak statutory argument on its face?'
        • (i) For one, you could actually read this to require improvement from everyone: “protect and enhance.”'
        • (ii) For some statutes, this might work better, but the CAA is a very complex, technical statute—relying on a general preamble doesn’t seem very sensible. This is a detailed regulatory regime. '
        • (iii) A statutory argument on the other side that Court didn’t look at: where’s the obvious place in the statute? 110: the SIPs. Wouldn’t you expect the section to set standards for SIPs that required prevention of further degradation? This is the section that tells you what states have to do to come into compliance with federal ambient standards. Why find this by implication when there’s a logical place for it to be? '
      3. Legislative history: '
        1. Administration officials testified before Congress that deterioration would be barred'
        2. Senate Report accompanying the bill that was eventually adopted stated that they understood it to bar SD. House Report less clear but doesn’t appear contradictory.'
          • (i) But how authoritative is this? It reads as a complete restraint on any degradation (“continued maintenance”). If this were in the statute, it would be pretty dispositive! But conference report is always more authoritative. Even though the Senate’s bill became the Act, all this really means is that the final bill had the Senate number, not the House number. The final bill rejected the Senate-version’s provision. Thus, the leg’ve hxy actually works against the Court’s opinion.'
        3. Administrator: has remarked that he believes the Act allows SD. '
          1. Rules are contradictory'
          2. Remember that this is pre-Chevron. In a Chevron world, this would be a strong argument for not deferring to the agency at all! '
        4. The part of the Rules allowing significant degradation is “contrary to the legislative policy of the Act.”'
  • Alaska Dept. of Environmental Conservation v. EPA, U.S. (2004), p379:
    1. Facts: Cominco owned a zinc mine in NW AK. Powered by diesel generators. Subject to a PSD permit for NOX. Wanted to increase production on a standby diesel generator. ADEC applied top-down BACT and proposed one method as BACT; company then proposed a non-top method but wanted to apply it to existing generators as well. ADEC approved this over EPA objections and EPA issued a stop order under § 113(a)(5).
    2. Holding: EPA’s oversight role extended to determining the reasonableness of a state permitting authority’s BACT determination in light of statutory guidelines.
    3. Reasoning: § 113(a)(5) and § 167 give EPA “encompassing” supervisory responsibility over construction and modification of facilities in the PSD area. Court finds that Congress entrusted state authorities w/initial responsibility. However, EPA has the authority to protect against unreasonable BACT designations when the state’s decision does not reflect reasoned analysis.'
  1. Non-Attainment
    • Ambient component – Reasonable Further Progress (RFP, §171(1)) – annual increment reductions
      1. Problem: only requires “the most stringent emission limitation which is contained in [a SIP] of any State for such class or category of sources.” (§ 171(A)). SIPs mainly regulate existing sources b/c new sources are regulated by NSPS. Alternatively, § 171 (B) allows “the most stringent emission limitation which is achieved in practice by such class or category of source, whichever is more stringent.” If an NSPS has not been determined, the emission limitation may not be all that stringent.
    • Emissions components –
      1. reasonably available control technology (RACT, §172(c)(1)) applicable to existing sources
      2. lowest achievable emissions rate (LAER, §§ 172(a)(2), 171(3)) for new sources
    • Note: In PSD areas and other areas, states are allowed to make decisions about whether or not they want to control existing sources, but in nonattainment areas, it is required.
    • Permits for new sources:
      1. § 172(c)(5), permits are required for “the construction and operation of new or modified major stationary sources anywhere in the nonattainment area” as explained in 173 (below).
      2. Must show the facility will comply with the LAER, provide for offsets, and that the owner is on schedule for compliance for other sources.
      3. Must show that “sufficient offsetting emissions reductions have been obtained, such that total allowable emissions from existing sources in the region, from new or modified source which are not MEFS, and from the proposed source will be sufficiently less than total emissions…so as to represent…reasonable further progress.” § 173(a)(1)(A),'
      4. Has to comply with lowest achievable emission rate: “…that rate of emissions which reflects the most stringent emission limitation” in any state implementation plan or is achieved by any source in this category. § 171(3).
      5. Revesz: distinction bw LAER and BACT isn’t clear. The offset requirement is a brand-new shiny thing, but this distinction isn’t.
    • Specific provisions
      1. Statute contains specific rules for certain pollutants. These were added by the 1990 Amendments. If your pollutant isn’t covered, you fall back on the general provisions.
      2. Example: § 181: Attainment framework for Ozone.
        1. Classifies attainment areas by degree of nonattainment, sets deadlines for attainment. Marginal to extreme. Only LA is extreme.
        2. The worse you get, the more onerous the requirements. Offset ratios are stricter (so normally, you just need to reduce more than you’ll add)—specify a ratio of reduction for new sources, frex.
        3. So you benefit from having more time, but the conditions are much tougher.
  • Case study: Citizens Against the Refinery’s Effects, Inc. [CARE] v. EPA, (offset case)(4th Cir, 1981), p390:
    1. Facts: Factory secured an offset for its emissions b/c the VA highway department switched to a different type of asphalt in some of its highway districts.
    2. Notes:
      1. But the state had already switched to this type of asphalt in most of its districts. Is this real “reasonable further progress”?
      2. The court reasoned that the program was voluntary previously, and now it’ll be compulsory. Thus, it is now enforceable.
      3. Presumably it’s less expensive to switch to this than to install control technology.
  • Example: a factory on the verge of bankruptcy can now “sell” the fact that it’s going to close.
    1. You’ve actually gained pollution.
    2. How do you calculate the amount of the offset? However much you were emitting at the time of the sale? That creates an incentive to produce more right before you shut down. It looks like that actually happened in CARE v. EPA: the year they picked was the year that they used the most asphalt.
    3. How can you fix this program? Set a universal baseline, for one thing. Have to manage the incentives created.
  • EPA allows industrial plants to be “bubbled” so improvements would not require a permit as a new source if total emissions of the plant will not increase. Chevron v. Nat’l Resources Defense Council, Inc. [NRDC] (bubble case), U.S. (1984), p392:
    1. What is the definition of a stationary source?
      1. Reagan administration proposed a “plantwide” definition. NRDC wanted you to look at a particular unit.
    2. § 302(j), Major stationary source= “any stationary facility or source of air pollutants…”
    3. But the definitions section for nonattainment, specifically, does not define a major stationary source.
      1. MEF definition in PSD: different tonnage requirements.
      2. Why is the court even looking at § 111 definitions? Because the definition of modification in § 171 is the same as in § 111(a)(4). Tenuous link.
    4. Does the court hold that the plantwide definition is required by the statute? No. Ambiguous statutory guidance. This is where Chevron deference gets developed.
    5. EPA’s interpretation of the term “source” hadn’t been consistent. Different interpretation for PSD vs. nonattainment.
      1. In nonattainment areas, EPA had taken the more restrictive interpretation: non bubble concept.
      2. In PSD, it had allowed a bubble-type interpretation.
      3. Can the same term be interpreted differently by an agency in different statutory provisions?
      4. Under Chevron’s admin rule, it seems like you could interpret it in light of the specific programs created by the statute.
    6. Why did the Reagan administration want this change?
      1. Create an incentive for big units to be built w/o having to comply w/NSPS. Plus, then you will potentially get ramp-up of emissions from existing facilities prior.
      2. But isn’t this inconsistent with the purpose of the nonattainment provisions? It may not clash with the statutory language, but it’s in collision with the purpose.
      3. What’s the most charitable construction? The statute’s about a tradeoff bw economic growth and environmental improvement.
        • (i) Reagan administration might claim that these regulations would be so burdensome that you’d never get a new source.
      4. Structure of the environmental laws tend to “enormously” privilege existing sources. Big barrier to entry: new plants would have to get a permit! Restricts economic growth, to a certain degree.
  • Deadlines: EPA cannot extend an attainment deadline, other than by reclassifying an area. Sierra Club v. EPA (nonattainment deadline)''''''''''(D.C.), p397.
  1. New Source Review' (p403)'''''
    • Generally'''''
      1. Proposal to construct a new source of air pollution may be subject to new source performance standards (§ 111) or the permit requirements of the PSD (§ 165) and nonattainment (§ 173) programs.
      2. New Source Review encompasses the PSD and NonA permitting standards but also applies to “modifications” at existing sources.
      3. New Source Review was supposed to ensure that modifications triggered more stringent standards
        1. But the EPA issued regulations in the 80s that limited modifications to “major modifications,” which did not include routine maintenance. (approved in Wis. Electric Power Co. [WEPCO] v. Reilly, 7th Cir., section (iii) below, p407—held that the EPA could have reasonably determined that the unprecedented “maintenance” at issue in the case was not “routine” and, thus, not exempted from NSR.)
      4. New sources are controlled more strictly than existing sources. Regulation of existing sources mostly left to states under SIP program
        1. Why have stricter standards for new sources?
          • (i) Differential compliance costs for new vs. existing sources
          • (ii) But what about incentive for anticipating change?
          • (iii) And what about equity?
        2. What if they can’t do anything to the plant?
          • (i) Extends the period during which it’s profitable to operate the plant, basically.
        3. So what happens when you have a plant that comes into compliance, and then EPA makes the regs more stringent?
          • (i) Desire to protect pro-environmental investments. We see this under CWA. If a company comes into compliance, and then the standard is amended, the company gets 10 years of protection of the investment before it has to comply.
        4. Note that we’re very concerned about settled expectations in CAA, but not in Superfund. Retroactive liability.
  • New or modified source definition:''''
    1. PSD provisions cross-reference NSPS definition of “modification” (§ 171(4), referring to § 111).
      1. Regulations re: NSPS (further defining modification as an hourly rate of emissions) do not necessarily apply to PSD (which EPA further defined more restrictively in reference to total emissions). EPA’s interpretation that “modification” is defined differently is reasonable. Environmental Defense v. Duke Energy Corp., U.S., p417. (Duke challenged the more restrictive interpretation).
    2. “construction” in PSD defined to include modification as in § 111 (§ 169(2)(c))
    3. The modification must lead to a significant increase in air pollution (§ 111) in order to trigger NSR.
      1. 2002: new rule for establishing baseline and future emissions in order to project the “significance” of an increase.
        • (i) EPA created a 10-year lookback period to give sources flexibility to account for slow periods in the business cycle (prev: 24 mos. prior). Challenged in NY v. EPA and upheld.
        • (ii) Plantwide Applicability Limitation: allowed sources to voluntarily adopt plantwide ceilings or camps on emissions and avoid NSR
        • (iii) Use of emissions limitations rather than actual emissions to measure emissions increases in the Clean Unit program
        • (iv) Exemption of certain control projects that result in “net overall environmental benefits” from NSR.
        • (v) Exemption for sources from recordkeeping requirements if they determine that there is “no reasonable possibility” that their projects constitute modification for NSR purposes. Challenged in New York' EPA and remanded for a “reasoned justification” of this decision.
      2. Reasons for including modification:
        1. Would encourage the installation of modern pollution controls on existing sources. Both commentators and courts have adopted this theory, based on legislative hxy.
        2. However, what’s actually happened is that the cost-benefit calculus has been tipped towards maintaining older facilities and patching them up at just below the level that will trigger NSR. “Old plant effect”. Depends on what the cost of environmental regulation is.
        3. How could you restructure this to avoid an old plant effect?
          • (i) Cap and trade system – existing sources get them, but they face an incentive to reduce their pollution.
          • (ii) Time limit on grandfathering. Think of this as similar to nonconforming uses in land use. Prior estimated natural life of the plant. That’s all they would have expected ordinarily.
        4. Defining modification and measuring increase has been, historically, a significant challenge.
          1. EPA had defined it differently w/in the NSPS context (see Duke Energy, above, and even in the same rule (see New York v. EPA). In 1980, EPA issued a standard rule for the two, which defined modification as “any physical change in or change in the method of operation of a major stationary source that would result in a significant net emissions increase of any pollutant subject to regulation under the Act.”
          2. Physical change exempted a number of things (see list p. 91)
          3. New emissions increase was “any increase in actual emissions from a particular physical change or change in the method of operation at a stationary source”—had to account for contemporaneous increases or decreases in emissions of the pollutant. Contemporaneous offsets had to occur in the five years prior. “Significant” was a set tons/year amount, specified for each pollutant.
  • Wisconsin Electric Power Co. v. Reilly, (7th Cir. 1990), p407:
    1. Question: was WEPCO’s program a “major modification” triggering NSR under the statute?
    2. Physical change: Replacement of various drums and air heaters.
      1. Plant argued that modification should be read to require a “fundamental change”, not replacement in kind. Court found this not to be supported by the statute or precedent.
      2. This would open “vistas of indefinite immunity”: like-like replacement would mean that the application of NSPS and PSD “might be postponed into the indefinite future”; the legislative history of the act suggests that this was not the intended result. Congress wanted to speed up improvement in air quality.
      3. The statute itself strikes an environmental-economic balance; not sticking strictly to the statute could upset that balance.
    3. Is this “routine maintenance”? EPA makes a case-by-case determination whether or not something is routine; in this case they found that it wasn’t.
    4. Omitted section: EPA improperly relied on potential to emit, rather than projected actual emissions to determine if there was a significant net increase.
    5. Post-WEPCO:
      1. Administration comes out w/ the 20% rule and the 10-Year Lookback Rule (see above under “significant increase). This was challenged in New York v. EPA, 2005, and the rule was upheld.
        • (i) Lookback Rule means they’ll choose the highest 24-month period in the preceding ten years as their baseline. What’s the rationale for doing this? Need to look back over a full business cycle. Surrendering capacity issue.
      2. What could you do instead? Some kind of averaging thing? Why didn’t the court require the agency to define more specifically what the length of the business cycle was? EPA says that this may vary from industry to industry and that picking 10 years facilitates certainty and clarity. Note that they erred on the side of permissiveness.
  • New York v. EPA, (D.C. 2005), p410: EPA’s definition of “modification” upheld under Chevron 2 deference.
  1. Interstate provisions''''
    • Problem: Upwind Pollution -----> Downwind States (p422)
      1. Upwind states create pollution that is externalized to downwind states.
      2. Coasian bargaining could limit the level of pollution to that which each state was willing to pay for or accept payment to receive.
        1. But Coasian bargaining is not likely to occur because
          • (i) High transaction costs relative to benefits
          • (ii) Baseline entitlements are not well defined by law.
            • (1) In Kentucky v. EPA, the transaction costs were presumably not that high. But until the case was settled, the entitlement was not established. (During the Carter administration, the downwind state had the entitlement; but the EPA resolved the issue under the Reagan administration).
          • (iii) Issues of causation are very complicated.
        2. The original CAA “was a poor tool to address such externalities” b/c emissions standards do not control (i) the geographic placement of sources, (ii) the size of each source, or (iii) the number of sources located in a state. In fact, the CAA incentived states to increase their interstate pollution.
  • Solutions
    1. Section 123 (1977) (tall stacks provision)
      1. Attempted to limit the use of “tall stacks” by limiting the amount of credit a state can receive for emissions limitations to the level consistent with a stack height built according to “good engineering practice.”
      2. This strategy is only effective to the extent that the emissions limitation affects the regulation of tall stacks. After the market permit regulation of SO2, this provision became less relevant.
    2. SIP Challenge via § 110(a)(2)(D) (1977)
      1. SIPs must meet NAAQS and must prohibit in-state sources from “contribut[ing] significantly” to nonattainment or interfering with maintenance or from interfering with PSD measures (violating increments) of another state. [If no baseline under PSD, no challenge.]
        • (i) If SIP does not address à can be challenged under judicial review, per § 307(b).
      2. State must bring a challenge w/in 60 days of approval of a SIP or SIP revision.
        • (i) Thus, a citizen cannot bring a suit under § 126 but could seek judicial review of the Administrator’s decision to approve the SIP.
      3. Agency review of SIP revisions should only focus upon emissions from the single pollution source at issue, not the emissions caused by aggregate sources. New York v. EPA (SIP Challenge), (7th Cir. 1983), p424. Challenges regarding aggregate emissions should be brought under § 126.
      4. EPA is allowed to consider costs in the determination of “significant contribution” under Section D. Michigan v. EPA, (D.C. Cir. 2000), cert. denied, p431:
        • (i) Facts: EPA issued SIP calls for 22 states and D.C. to mitigate the interstate transport of ozone. EPA then issued a rule that required the areas to reduce NOxs by “highly cost-effective controls” (defined as controls that EPA found to be capable of removing NOx for $2000/ton or less). As a result, “the ultimate line of ‘significance’ . . . would vary from state to state depending on variations in cutback costs.” Various parties challenged the EPA’s decision.
        • (ii) Holding: The EPA reasonably acted in accordance with its authority under § 110 when it (1) considered costs in defining “significant contribution” under Section D, (2) required uniform controls that have non-uniform effects, and (3) established NOx budgets for each state.
        • (iii) Reasoning:
          • (1) Cost consideration:
            1. In the Benzene case, the plurality opinion suggests that a “significant” risk is defined, at least in part, by reference to the cost of eliminating it.
            2. Ps have not proposed any other method by which EPA could determine “significance.”
            3. Rule in D.C. Cir: Clear congressional intent to preclude consideration of costs is required to reach a finding which bars agencies from doing so. Ps have not demonstrated such intent in this case.
          • (2) Non-uniformity: Although the EPA’s decision will lead to a non-uniform financial impact, the EPA considered this issue and found that a more individualized regional approach would not provide a significant improvement in air quality or a substantial reduction in cost.
          • (3) NOx budgets: States have a “real choice” in how they want to achieve the “NOx budget” allocated to them. Thus, these budgets “do not fall within the realm of impressible SIP call regulation as defined in Virginia [ EPA (1997) (see Note (2) below)] and Train [v. NRDC].”
        • (iv) Dissent: The statute does not grant the EPA authority to consider cost of alleviation when assessing SIPs. The text clearly grants the EPA to require states to act in regard to “amounts which will . . . contribute significantly to nonattainment.”
        • (v) Notes:
          • (1) Why was cost-consideration upheld here re: §110 but not upheld later in Trucking re: § 109, where the Supreme Court declined to “find elephants in mouseholes”?
            1. § 109 called for NAAQS “requisite to protect the public health.” But § 110 does not explicitly allow for cost consideration either, and under the same analysis, “adequate margin of safety” in § 109 could allow for cost consideration.
          • (2) Virginia v. EPA, (D.C. Cir. 1997), p436n3, held that the CAA did not authorize EPA to condition SIP approval on the adoption of particular measures. Michigan limited this holding b/c it allowed the EPA to essentially require certain measures if some “choice” was still available to states.
          • (3) What was EPA trying to accomplish via its ruling? Cost-minimization: Meet the NAAQS at the cheapest cost to the nation as a whole.
          • (4) What would have happened if this cost-minimization standard had been applied in APCD Jefferson County v. EPA, below under §126? It would likely have resulted in requiring Gallagher to take control measures—since it had not done much, it would presumably have cost relatively less to do something. The difference in this case is that the EPA issued the rule.
          • (5) Why are these two EPA decisions inconsistent? Different administrations (’84 decision vs. ’01 decision).
          • (6) PSD Application: This is approach for nonattainment areas. What about PSD areas? Under § 110(a)(2)(D)(i)(II), what is the standard for “interfer[ing] with measures required to be included in the applicable implementation plan...”? If a baseline has not been set (by construction of a new source), there are no PSD requirements.
  1. Section 126 petition (§ 7426-“Interstate pollution abatement”) (1977)
    1. State can petition the EPA at any time for a finding that a major source or group of sources “emits or would emit any air pollutant in violation of the prohibition of [§ 110(a)(2)(D)(i)]”—i.e., the sources would contribute significantly to the challenging state’s nonattainment or maintenance or would interfere with its PSD measures.
      • (i) Administrator must make a finding w/in 60 days after petition and after a public hearing.
      • (ii) If EPA finds a violation, it must set a schedule to bring the sources into compliance “as expeditiously as practicable” but no later than 3 years after the finding.
    2. Prior to construction or modification, a new or modified source that is subject to PSD requirements or might contribute significantly to an interstate nonattainment must provide written notice to the potentially-affected states.
    3. Issues:
      • (i) Only states and political subdivisions can bring claims under this, not individuals. The EPA Administrator is the decision-maker (§126(b)).
      • (ii) How do you determine if pollution is “significant”?
        • (1) violation of § 110(a)(2)(D)(i)—so there’s not a substantive standard.
  2. Whereas § 110 does not allow the EPA to mandate particular control measures directly, § 126 does allow such federal mandate. Appalachian Power Co. v. EPA, (D.C. Cir. 2001), p437:
    • (i) Facts: Eight states initiated a § 126 challenge to the SIP call at issue in Michigan v. EPA. The EPA promulgated an “automatic trigger mechanism” which would result in § 126 findings if an upwind state’s response to the SIP call was not approved.
  3. A § 126 petition will only be granted if (i) emissions in the challenged state substantially contribute to (ii) a violation of NAAQS in the challenging state. Air Pollution Control District of Jefferson County [Kentucky] v. EPA, (6th Cir. 1984), p425:
    • (i) Facts:
      • (1) 1972—EPA approved a Kentucky SIP that set an emissions limitation at 1.2 lbs SO2/MMBtu.
      • (2) 1973—EPA approved the same limitation for Indiana’s SIP.
      • (3) 1974—Indiana exempted the Gallagher Power Station, which is just across the border from KY, from the limitation.
      • (4) 1978—Jefferson County designated as nonattainment area.
      • (5) 1979—Indiana submitted a SIP revision setting the limitation at 6 lbs, which was the current emissions rate of Gallagher. The Administrator approved Indiana’s revision w/ respect to the primary NAAQS but not the secondary ones.
      • (6) KY filed a § 126 petition. EPA denied the petition, finding that
        1. the Gallagher plant does not cause or significantly contribute to Jefferson County’s nonattainment b/c only about 3% of the SO2 concentration at any location in the county where the NAAQS are violated was attributable to Gallagher (although 34-47% SO2 concentration in other parts of the County were attributable to G) and
        2. the interstate pollution abatement provisions only protect against interference with national standards, not more stringent, local standards—although G limits the “margin for growth” contemplated by the KY SIP, the PSD regulations are inapplicable since Jefferson is a nonattainment area, not a PSD area.
      • (ii) Holding: The EPA’s determination is not unreasonable and, thus, should be upheld.
      • (iii) Reasoning:
        • (1) Although the statute does not explicitly refer to a significant contribution test,[1] reference in § 126(a) implies that Congress intended it.
        • (2) Prohibiting de minimis contributions by one state to another state’s NAAQS violation would “in effect ‘hold one state hostage to another’s failure to enact the pollution control strategies necessary to conform to the requirements of the Clean Air Act.’”
      • (iv) Notes:
        • (1) As a result of this ruling, although states are free to set their own, more stringent standards under § 116, interference with such standards cannot be enforced against upwind states under § 126. Thus, if interstate pollution is a problem, a state would probably not want to enact more stringent standards that would only affect in-state sources.
        • (2) The court provides no analysis for what counts as a “significant” contribution.
      • (v) Class Notes:
        • (1) The entire county was classified as nonattainment, even though the court parsed out areas of the county that were not in nonattainment.
        • (2) Allowing downwind states to set more stringent standards that constrain upwind states would allow upwind states to be hostage to downwind states (a first-in-time rule that is generally disfavored b/c it encourages wasteful use of resources). Conversely, the ruling of this case allows upwind states to hold downwind states hostage.
        • (3) Ironically, the CAA was enacted as a federal law in part to address interstate externalities, but the very provision that would seem to address that does not allow states to do so!
  4. SIP Call via § 110(k)(5) (1990)
    1. Administrator must issue a SIP call to “require the State to revise the plan as necessary to correct such inadequacies” when he finds a SIP is substantially inadequate to
      • (i) attain or maintain the relevant NAAQS,[2]
      • (ii) adequately mitigate interstate pollutant transport per §§ 176A & 184, or
      • (iii) “otherwise comply with any requirement of this chapter.”
    2. Challenges to SIPs and interstate provisions:
      • (i) Petition the EPA to issue a SIP Call. If that petition is denied, then the denial can be appealed in court.
      • (ii) Although there are three potential avenues for challenging SIPs (SIP Challenge, SIP Call, and a State Petition under § 126), they all have the same substantive standard b/c they all point to § 110(a)(2)(D).
      • (iii) Per § 307(b), which allows for judicial review: Challenges to the NAAQS have to be filed in the D.C. circuit, but challenges to SIPs must be filed in the relevant circuit, unless the administrator’s decision has national scope or effect.
        • (1) Most likely the circuit of the state whose SIP is being challenged.
        • (2) There’s no time limit in § 126, but there’s a 60-day time limit in § 110 (SIPs). The EPA Administrator is the decision-maker in §126, but in § 110, it’s the relevant federal judge.
      • (iv) SIP amendments
        • (1) Courts have held that the only thing that can be challenged after an amendment is the effects of the amendment.
        • (2) Problem with this is that it enables you to change a SIP bit by bit to sneak in pollution that would otherwise be impermissible.
  5. Alternative solutions for interstate air pollution
    1. Internalize the externality by making any downwind pollution created by a state count towards their compliance with NAAQS. Problem is trying to measure this.
    2. Ban downwind pollution. Pollution isn’t random—prevailing winds dictate pretty well where pollution will go. But if you just banned downwind pollution entirely, you might just stop a state from having any industry.
    3. What could a judge do? Sometimes they sort of try to impose taxes, but that’s frowned-upon. It’s hard to fix this w/o a statutory scheme.
    4. Conceptualize the air in a downwind state as a resource to be shared between the upwind and the downwind state. Set possibility of more stringent state ambient standards aside. Assume it’s met in the downwind state. But what’s a plausible regulatory tool/risk management strategy for minimizing the problem? Want to minimize costs. Figure out where you can regulate most cheaply, and do that. So you’re sort of “averaging” air quality between the two.
      • (i) But what if the downwind state wants to do better?
      • (ii) What about in a PSD situation? Assume no baseline but the state really wants cleaner air and doesn’t want to get dragged up to the NAAQS. Should an upwind state be allowed to impose whatever pollution it wants on the downwind state until the baseline gets set?
  • Acid Deposition Marketable Permit Scheme (p442)
    1. History of attempts to control SO2
      1. 1970 CAA Amendment—authorized EPA to set new standards for SO2. EPA allowed:
        • (i) low-sulfur coal or
        • (ii) scrubbing
      2. 1977 Amendment—required new electric plants to use scrubbers (essentially eliminating the low-sulfur option).
    2. 1990 Amendments- Market-based tradable permit scheme
      1. “Big dirties”—EPA distributed allowances to certain electric-generating units (EGUs) at emission rate of 2.5 lbs SO2/MMBtu (heat input), multiplied by the unit’s baseline MMBtu.
      2. Others—EPA distributed allowances to other EGUs and emissions units at 1.2 lbs SO2/MMBtu times unit’s baseline.
      3. Cap (maximum) at 8.95MM lbs.
        • (i) Policy: Incentivizes utilities to reduce their SO2 emissions so that they can profit off the sale of their allowances/permits.
        • (ii) Effect: Low-sulfur (western) coal again became a viable option.
      4. Widely considered a success, per p452n5.
    3. Some states attempted to protect business (dirty coal) interests, which were affected by the MPS. However, the Dormant Commerce Clause does not allow the type of discrimination against out-of-state coal producers that the Illinois Coal Act established, even when economic costs to a region result from unrestricted interstate commerce. Alliance for Clean Coal v. Miller, (7th Cir. 1995), p443:
      1. Constitutional Background (p451n3): The Constitution, through the Commerce Clause (art. I, § 8, cl. 3) grants Congress plenary authority to regulate interstate commerce. Conversely, the “Dormant Commerce Clause” jurisprudence has established that this authority limits states’ ability to interfere with interstate commerce by discriminating against out-of-state parties or by otherwise burdening the interstate flow of articles of commerce.
        • (i) Policy: Establish a unitary national economy.
      2. Facts: In response to the 1990 CAA Amendment, which could potentially cause lost jobs in high-sulfur coal-mining regions such as Illinois, the Illinois Legislature passed the Coal Act. A Virginia trade association (including railroads, who transport coal) sued the Illinois Commerce Commission (ICC), who was charged with implementing the Coal Act.
      3. Holding: The Coal Act and the compliance plans approved in accordance with it are invalid b/c they impermissibly restrict interstate commerce.
      4. Reasoning:
        • (i) The Act discriminates against out-of-state coal producers by requiring:
          • (1) that the ICC to take account of costs to the local economy when considering whether to approve compliance plans,
          • (2) that certain plants to install scrubbers “to continue to burn Illinois coal,”
          • (3) that the cost of using high-sulfur coal be passed on to the consumers, and
          • (4) that the ICC consider the impact on local employment before approving a utility’s plan to decrease its use of Illinois coal by 10% or more.
        • (ii) Although Ds argued that the Act only “encouraged” the use of Illinois coal, the Court found that the Act’s mandates amounted to “ingenious discrimination.”
        • (iii) “[T]he fact that the rate-payers [(i.e., taxpayers)] are footing the bill does not cure the discriminatory impact on western coal producers.”
        • (iv) The “market participant” exception does not apply.
        • (v) Protection against local economic harm is not a justification for such discrimination.
      5. Concurrence [Cudahy]:
        • (i) “[T]he Commerce Clause effectively precludes consideration of local economic damage as a legitimate reason to handicap interstate commerce.”
        • (ii) Federal preemption, through the operation of the Supremacy Clause, might also cause the Coal Act to be invalid, if the court were to determine that Congress intended to occupy the field with a market-based approach via the CAA Amendment.
      6. Nor can downwind states provide for greater protection of environmental interests. The Supremacy Clause prohibits state laws from impeding the execution of the full purposes and objectives of federal laws by interfering with the methods by which the federal laws were designed to achieve those goals. Clean Air Markets Group v. Pataki, (2d Cir. 2003), p447:
        1. Facts: Because Title IV of the CAA Amendments of 1990 created a nationwide marketable-permit scheme for SO2, emitting units in states downwind from NY could just buy more permits rather than reducing their SO2 emissions. To address this problem, NY passed the Air Pollution Mitigation Law in 2000 (“Section 66-k”).
        2. Statute: Section 66-k imposed a “mitigation offset” (fine), assessed at the purchase/trading price of the allowance, to NY utilities who sold or traded their allowances to facilities in certain upwind states. The law also imposed the offset for allowances not sold to those states but subsequently transferred to one of the states. Accordingly, to avoid such a fee, the utilities must attach restrictive covenants to allowances (not sold to one of the identified states) that prohibit their subsequent transfer to one of the identified states.
        3. Holding: Section 66-k is invalid under the Supremacy Clause of the Constitution.
        4. Reasoning:
          • (i) Although NY argued that its law and the federal law have the same “ultimate goal” (environmental protection), the NY law impermissibly interferes with the methods by which the federal law was designed to achieve this goal.
            • (1) The text of the federal law clearly indicates Congress’ intent to effectuate a nationwide tradable-permit scheme.
            • (2) Legislative history: Congress considered and rejected a plan that would place regional restrictions upon trades.
            • (3) EPA regulations: mandate that state programs will not interfere with trading.
          • (ii) Like the Illinois Coal Act, the NY law may not explicitly restrict the actions of facilities. However, “it clearly interferes with their ability to effectuate [allowance] transfers.”
            • (1) The NY law requires utilities to forfeit 100% of any proceeds from trades to identified upwind states.
            • (2) The law decreases the value of allowances traded to non-identified states by essentially requiring restrictive covenants on such trades.
  1. NAAQS case study: Lead (p317)''''
    1. Background
      1. Administrator forced to add lead because he’d made findings about the dangers of lead when promulgating the unleaded gasoline rule under § 211 of the CAA.
      2. W/in Administrator’s discretion to make a finding of adverse effect, but once he’d done it, had to list lead as a criteria pollutant.
        1. However, in a subsequent challenge to the sulfur oxides NAAQS, court held that the content of a revision (as opposed to a new standard) was w/in the Administrator’s discretion.
  • Setting the lead standard
    1. Determine the criteria: what the effects are and who’s at risk.
      1. What does the statute say about the stringency of the standard?
        • (i) Primary: requisite to protect the public health, allowing an adequate margin of safety.
        • (ii) Secondary: protect the public welfare from known or anticipated adverse effects.
      2. Determine the critically sensitive population
        1. Young children age 1-5 b/c still developing, so greater possibility of developmental effects (MR); also, greater likelihood that little kids will ingest something (eat paint chips).
          • (i) Critical population is the one that sees an adverse effect at the lowest concentration. Could be pregnant women, asthmatics, old ppl, etc.
        2. Why look at the most-sensitive population? If they’re protected, everybody else is as well. Meets no-risk goal.
        3. (Young children aren’t always the most critically-sensitive group.)
      3. Determine the pivotal adverse health effect (lowest level at which adverse effect is observed)
        1. The health effect that occurs at the lowest concentration is chosen as the “signal” via which to set standards. This is sometimes called a critical population/critical effects test.
          • (i) Point is to stay under the threshold by determining where it exists.
        2. The EPA determined the air lead standard to prevent the occurrence of EP elevation in children, rather than ALAD inhibition. Why?
          • (i) ALAD shows an effect earlier, but the effect doesn’t lead to impairment—not a harmful physiological change—as with EP elevation.
            • (1) But you can imagine that the Administrator might have chosen this level instead—a detectable physiological change at a lower level. Precautionary approach due to lack of studies.
          • (ii) Also, CDC had used it.
        3. Determine the mean population blood level that would protect the CSP.
          1. Although CDC established an individual threshold for risk at 30 µg Pb/dl (30 micrograms of lead per deciliters of blood), the EPA is looking at mean population exposure, meaning some individual children will be below and some above the mean. So, EPA set the mean at 15µg Pb/dl.
          2. Based upon a normal statistical distribution, about 20,000 kids (0.5%) will still be above the CDC individual level of 30µg Pb/dl. Did the EPA not care about them? On what basis did the Administrator make this judgment?
        4. Determine the relationship b/w air lead exposure and blood lead level
          1. Ratio of exposure to blood lead increase set at 1:2 (for every 1µg of lead in the air, blood lead increases by 2µg).
          2. This is conservative estimate, based on studies showing a range of responses, but it’s not entirely clear why EPA chose this level.
        5. Determine the allowable blood lead increment from air
          1. Problem: most of blood lead comes from non-air sources.
            • (i) Basically, kids in public housing were getting lead from eating paint chips off the wall or paint turning into dust in the air. EPA didn’t have statutory authority to regulate this, though.
          2. Since non-air sources contribute most, NAAQS must be set low to keep average exposure under 15.
            • (i) Estimates of non-air contributions ranged from 10.2 to 14.4; EPA chose 12.
              • (1) Not much explanation of why 12.
              • (2) Probably due to imbalanced regional impact. No NAAQ will be protective in all locations, but EPA doesn’t think that § 109 allows them to set location-specific standards. “[I]f EPA were to use a larger estimate of non-air contribution to blood lead, the result would be an exceptionally stringent standard, which would not address the principle source of lead exposure. Conversely, EPA believes that it should not adopt an estimate of non-air contributions below the level shown in available studies to be the lower mean blood leave level documented in the criteria document.”
            • (ii) Target is 15, minus 12 from non-air sources = 3µg/dl.
            • (iii) This must be divided by the ratio determined in step 5, making the appropriate standard 1.5µg m3.
          3. Margin of safety, b/c based on lowest threshold and attuned to most sensitive population.
            1. EPA has selected a ratio at the “protective end of the range”—notice, however, that they could have picked a much more stringent standard and didn’t!
            2. What happens if this gets challenged in court?
            3. How does an Administrator set a standard?
              • (i) Maybe it wouldn’t actually be that hard to reduce emissions to 0.6! Maybe that would be super-cheap. Theoretically, you’re not allowed to look at this. Eads says that you do anyway.
            4. Cases: Cost-benefit under CAA''''
              • Lead Induss. Ass’n v. EPA, p333 (1980) (II-9)'
                1. Court previously reviewed and approved the unleaded gas regulations. In this case, asked to review the NAAQS.
                2. Question: must the Administrator consider economic and technological feasibility in setting the NAAQS? No.
                  1. Petitioner bases this on the statutory provision allowing an adequate margin of safety—believes this requires the Administrator to weigh these factors in deciding what that margin is.
                  2. Court doesn’t buy it. Nothing in the Act or the legislative history back this up.
                    • (i) Congress told Administrator to consider this when they wanted him to: 111 requires it to be considered in NSR provisions, frex.
                    • (ii) Expressio unius: explicit language lists lots of factors that are supposed to be considered.
                    • (iii) So industry isn’t arguing that you can consider costs in the scientific analysis, but is arguing that in the decision on an “adequate margin of safety” you can do so.
                    • (iv) Court says that “adequate margin of safety”, on its face, doesn’t have anything to do with costs, and that if they’d wanted costs to be considered there, they would have put it there.
                      • (1) Congress knows how to do this if they want to.
                      • (2) But why assume against it? Scalia’s argument goes to this—it’s too important and too likelihood to strongly influence the application of the statute.
  1. Question: is the Administrator allowed to set the standards based on likely harm, rather than proven harm?
    1. Yes: Congress wanted them to err of the side of caution.
      • (i) Economic and technological feasibility to be subordinated to goal of protecting public health.
      • (ii) CSP to be focus of regulation.
      • (iii) Absence of adverse effect on those individuals the goal.
  • Whitman v. Am. Trucking Ass’ns, U.S. (2001), p326:'
    1. Challenge to the PM and ozone NAAQS'
    2. Does “the public health” encompass an economic-impact factor?
      1. No. Congress was aware of the potential economic impact of air quality regulations. § 110 permits the Administrator to waive the compliance deadline if there are potential adverse impacts on the public health or welfare; NSR provisions; auto provisions—basically, Congress knows how to give this authority if it wants to!
      2. The CAA does not “hide elephants in mouseholes.” Also, the statute is explicit in other sections, indicating that Congress knows how to make it clear when they want to.
      3. “adequate margin” and “requisite” do not constitute authority to consider economic impact. These “modest words” do not have “the power to determine whether implementation costs should moderate national air quality standards.”
    3. Does the “Administrator’s judgment” allow him to consider economic impacts?
      1. It may allow him to include certain unenumerated impacts. However, not cost of implementation—that’s too direct and too important and too likely to severely impact what’s done—that would sure have been mentioned if Congress had wanted it considered.
      2. Conversely, the statute could have clearly articulated that costs should not be taken into account.
    4. Minor questions:
      1. Do the provisions that allow the states to receive information on the economic impacts of abatement technologies constitute permission to weigh these factors? No.
      2. Did § 109(b) as interpreted by the Administrator lack an “intelligible principle” to guide the EPA’s authority? Lower court found that EPA’s interpretation was a nondelegation violation.
        • (i) An agency’s interpretation doesn’t create a nondelegation problem. That’s in the statute.
        • (ii) This statute is well within the outer limits of nondelegation doctrine.
      3. Breyer’s concurrence'
        1. See above in NAAQS section.'
      4. Mobile Sources (p452)'
        • Generate more than half of U.S. air pollution
        • Regulated by Title II
          1. § 202 – vehicle emissions
            1. “which reflect the greatest degree of emission reduction achievable through the application of technology which the Administrator determines will be available for the model year to which such standards apply, giving appropriate consideration to cost, energy, and safety factors...”
          2. § 211 – fuel content
          3. Technology standards do not limit aggregate pollution
        • Preempt state standards.
          1. Set a floor and ceiling. § 209(a).
            1. Ceiling on any pre-sale regulation of motor vehicles: California’s South Coast Air Quality Management District enacted “Fleet Rules” with stringent requirements for the purchase of fleet vehicles. The lower courts found that purchase requirements did not qualify as “standards” within the meaning of § 209(a) and were not preempted, but the Supreme Court disagreed. Engine Mfgs Ass’n v. S. Coast Air Quality Mgt. Dist., U.S., p463.
          2. Exception, § 209(b):
            1. EPA can waive certain state standards, unless (A) state’s determination was arbitrary and capricious, (B) state did not have a necessary and compelling need, or (iii) state’s standard is inconsistent with § 202(a) technological feasibility.
            2. Calif. and identical programs (11 other states) have received such waivers.
              • (i) California’s waiver for in-use maintenance restrictions was upheld in Motor & Equip. Mfg. Ass’n v. EPA, (D.C.), p459.
            3. Market participant exception – discussed in Engine Mfgs, &p469.
  • EPA’s grant of a deadline waiver for The Auto Industry was reversed by the court b/c the risks to the auto industry were outweighed by the potential environmental risks of delay. Int’l Harvester Co. v. Ruckelshaus, (D.C. Cir. 1973), p454
  • Greenhouse Gas Regulation for Mobile Source CO2 Emissions
    1. CAA § 202 requires EPA, in response to a petition for rulemaking, to make a judgment about whether CO2 is a pollutant that endangers public welfare. Massachusetts v. EPA, U.S., p470.
      1. Dissent (Scalia): Not a nondiscretionary duty—The statute says nothing about the reasons for which the Administrator may defer making a judgment.
    2. Hazardous air pollutants' (p486)'
      • Regulated under § 112: “shall establish any such standard at the level which in his judgment provides an ample margin of safety to protect the public health from such air pollutant.”
  1. Pre-1990 § 112:
    1. Don’t have to establish ambient standards or review SIPs. Thus EPA can directly control them via a 2-step rulemaking process:
      1. Listing as hazardous
      2. Uniform national emissions standards
    2. Results disappointing!
      1. Only seven pollutants listed in the 20 years after the passage of the Act
      2. Regulations that were promulgated were often ineffectual: “quintessential example of ineffective halth-based standard-setting”
    3. NRDC v. EPA: court rejected a standard explicitly based on technical and economic feasibility. Court allowed EPA to consider feasibility only w/i the framework of an “ample margin of safety.”
      1. Health standard set first, then margin of safety established.
    4. NRDC v. Thomas: DC Circuit upheld trail court’s dismissal of a case that sought to compel the Administrator to list eight pollutants for which he had made a finding of carcinogenicity. Court found that the notices were not the “functional equivalent” of a finding that these were HAPs.
    5. DC Circuit, in a 1990 decision, interpreted this to allow consideration of costs.
  2. 1990 reforms:
    1. Technology-based approach to regulation, w/ health as a backstop.
    2. Initial list of HAPs set by Congress.
    3. Requirement to establish source categories
    4. For each category, an emission standard to be established following a schedule.
    5. Standard to be set according to the Maximum Available Control Technology (for major sources): “the average emission limitation achieved by the best performing 12 % of the existing sources or, for categories containing fewer than 30 sources, the emission limitation achieved by the best performing five sources.”
    6. 8 years after development of the initial standards, a health-based standard could be established to eliminate residual risks
      1. Still with the foot-dragging
    7. 112(f)(2): Allows costs to be considered in setting the second set of standards – but can costs be considered when setting standards to protect the public health (first part of the sentence) or only when “environmental effects”? Note that the first part of the sentence is basically identical to the pre-1990 language, which was the subject of the court’s holding.
      1. “the Administrator shall, within 8 years after promulgation of standards for each category or subcategory of sources pursuant to subsection (d) of this section, promulgate standards for such category or subcategory if promulgation of such standards is required in order to provide an ample margin of safety to protect public health in accordance with this section (as in effect before November 15, 1990) or to prevent, taking into consideration costs, energy, safety, and other relevant factors, an adverse environmental effect.”
  • National Mining Association v. EPA
    1. EPA’s final rule on the implementation of emissions standards for HAPs.
    2. Definition of major sources: “any stationary source or group of stationary sources…” – rule language nearly identical to statutory language. Emissions from all sources of HAPs w/in a plant site to be aggregated, so long as geographically adjacent and under common control. If total emissions exceed standard, then MACT applies to all sources.
      1. Petitioners argued that a minor facility that happens to be located at a larger industrial site will be impermissibly regulated. Argue that these have to be w/I the same source category.
    3. Held: Statutory language accords w/EPA’s interpretation. “Compelled” by the statute. Statutory language:
    4. Held: EPA is required to consider state and local controls, even if not included in the SIP, when calculating the amount of hazardous air pollutants a statonary source can emit. EPA is not limited to considering “federally enforceable” controls when calculating the maximum capacity of a plant (not that operational controls are treated as reducing the plants emitting capacity).
  • Cement Kiln Recycling Coalition v. EPA (2001)
    1. Congress imposed minimum stringency requirements that apply w/o regard to costs or to ther factors. For new sources, must achieve the same level as the best controlled similar sources. For existing sources, MACT. EPA implements these via two-step process.
    2. EPA issued standards limiting emissions from three types of sources that produce 80% of HAP emissiosn in the US. W/i each category, EPA set standards for a series of HAPs.
      1. Then EPA set the MACT floors for each one: id’d the best-performing 12% of sources (MACT pool), id’d emission control technology used by most of the sources. Two techniques: end of stack control technology, feedrate. EPA then expanded the pool to include all sources using the MACT control and set the floor at the worst emission level achieved by any source in the expanded pool.
      2. Then EPA set five beyond-the-floor standards as req’d by 112(d)(2).
    3. Sierra Club argues that 112(d)(3) requires floors to reflect emissions actually achieved by the best-performing sources and doesn’t allow EPA to set a lower floor based on what it believes to be achievable by all sources.
      1. EPA argues that 112(d)(3)’s floor provision incorporates 112(d)(2)’s ahcievability standard. Court finds that this conflicts w/the plain language of the statute.
  • Sierra Club v. EPA (2004) (154)
    1. Sierra Club challenges the promulgation of regulations governing HAP emissions from primary copper smelters. EPA found that copper smelters used PM controls to control HAP emissions, and thus set standards in terms of PM, not HAP.
    2. Sierra Club argued that EPA didn’t consider non-air quality health and environmental impacts as required by 112(d)(2) (“taking into consideration the cost of achieving such emission reduction, and any non-air quality health and environmental impacts and energy requirements…”).
      1. Non-air quality impacts are results of emissions that don’t manifest themselves as air pollution, like acid deposition, according to SC.
      2. According to EPA, however, they’re health and environmental imapcts that may result from measures to achieve the emission reductions.
    3. Congress didn’t define the term, and EPA’s interpretation passes the Chevron test. Non-air quality impacts are groups with consideration of the costs of emission reductions and energy requirements, which strongly suggests that this is meant to allow the administrator to consider the costs and benefits of the control measures.
    4. Furthermore: this construction would suggest that EPA was supposed to consider these less-direct impacts now but more direct impacts later (when the the risk-based standards are promulgated).

The Clean Water Act[edit | edit source]

  1. Generally
    1. § 101(a)(1) “…it is the national goal that the discharge of pollutants into the navigable waters be eliminated by 1985.”
      1. Ambient standards CAA version of water quality standards
      2. Emissions rules CAA version of effluent standards.
      3. All new sources required to meet strict performance standards. All of the work of the CWA for the first twenty years was done via effluent limitations. Water quality standards took about 20 years to be established. So the “work” that the ambient standards due in the CAA is different than the “work” that water quality standards do in the CWA. CWA driven by ELs.
  • Effluent limitations
    1. § 301(b)(1)(A): Effluent limitations have to be established.
    2. Must require the best practicable control technology currently availably as defined by the administrator pursuant to…
      1. Note that this does NOT say “for categories/classes”. However, the DuPont court said “yes” to categories.
    3. § 304(b): Must identify the characteristics of the pollution. May consider total cost of technology in relation to effluent benefits, age of equipment, engineering aspects, non-water quality environmental impacts for both BPT and BACT.
    4. § 301(b)(2)(A): For specific toxic pollutants, must specify the use of the “best available technology economically achievable for such category or class, which will result in reasonable further progress toward the national goal…”
    5. § 306(a), (d): National standards of performance.
      1. “standard of performance” means a standard that reflects the “greatest degree of effluent reduction which the Administrator determines to be achievable through application of the best available demonstrated control technology”
      2. “new sources” defined
      3. “source”= “any building, structure, facility, or installation”
      4. “construction”=any placement, assembly, or installation of facilities or equipment at the premises.
      5. (d): preemption for 10 years.
    6. Best management practices allowed as a type of effluent limitation under CWA
  • Major method for carrying this out: National Permit Discharge Elimination System.
  • Phased approach to setting standards for existing sources.
    1. First round: “best practicable control technology currently available” (BPT) § 301(b)(1)(A)
      1. The 1977 standards were to be based on the “best practicable control technology currenty available”. Had to be set by class and category of structure. Practicable was intended to limit the standard to the use of existing technology only in cases where additional technology would yield a reduction wholly out of proportion to cost. So EPA must conduct a CB analysis on switching to new technology. Congress was concerned about overburdening industry.
    2. Second round: “best available technology economically achievable for such category or class” (BAT) Probably not a lot of difference bw CWA BAT and CAA BACT.
      1. Explicitly set by category/class. Cost was a less important factor: no CB requirement.
      2. How does EPA determine what technology is “available”?
        • (i) Courts have allowed EPA to set standards based on data from one facility.
        • (ii) Technology that had not yet been applied in practice was also allowed so long as it was reasonable to assume that it would be applied.
      3. § 304(b)(1)(B) (Effluent Limitation Guidelines): BPT. Comparison of total cost of application of the technology to the effluent reduction benefts—the comparison factors, v. the consideration factors listed later.
      4. § 304(b)(2)(B): BAT. All are consideration factors. No comparison factors.
      5. Why the distinction?
        1. Why might Congress have wanted to set a BPT standard that requires benefits to exceed costs, even if industry could have met a higher standard?
          • (i) Strong technology forcing component to the statute. Congress wants to force change regardless of the costs, but only eventually. Unless there was really a strong environmental benefit, they weren’t going to force it early.
  • Controls on point as well as nonpoint sources.
  1. Effluent limits and state designation of use of waters. Goal for all waters to eventually be “fishable/swimmable”, but many still are not.
    1. Once the use has been designated, the state determines the water quality criteria necessary to support the designated use.
  • More on new sources:
    1. BPT and BAT apply to existing sources, while the standards apply to new sources. Set for categories and must be based on technology demonstrated to be available presently (though can be one plant only).
  1. The CWA v. the CAA''''
    • Less emphasis on water quality standards (the equivalent of ambient air quality standards) and more on the effluent limitations (the equivalent of emissions standards)
    • Extensive federal role in setting effluent limitations vs. state role of setting emissions standards; conversely, states set the water quality standards vs. federal setting of NAAQS.
    • Existing sources
      1. By 1977, existing sources had to meet federal BPT standard and by 1983 (though later extended by statute), BAT standard. No grandfathering.
    • Non-point sources
      1. A lot of concern about pollution that doesn’t come out of a single, identifiable point. Agricultural runoff, frex.
      2. Point sources used to be a big chunk of the water quality problem. Now, however, we’ve been able to make a big reduction in effluent from point sources. Non-point sources are much harder to regulate, and so the shift over time means that NPS are now a bigger chunk of the water quality problem.
        1. Sort of like fugitive emissions in CAA. But that’s less of a problem—appears to account for a small chunk.
      3. E.I. Du Pont de Nemours v. Train, U.S. (1977), p509:
        • Question presented: Does EPA have the authority under § 301 to issue industry-wide regulations limiting discharges by existing plants? Must EPA allow variances for individual plants to the new source standards issued under § 306?
          1. DuPont argues that the § 402 permit is the only statutory authority for the issuance of enforceable limitations on the discharge of pollutants by existing sources.
          2. § 301 does not state who shall set the effluent limitations. EPA argues that it has the authority, and that the 402 permits simply incorporate the across-the-board limitations that it is allowed to make.
        • It looks, from the statute, sort of like the EPA administrator was supposed to set guidelines pursuant to 304 before setting [?] limitations under 301 and national standards for new sources under 306. This didn’t happen.
          1. EPA looked at plants with really good pollution control and used that information as well as technical data and economic studies to determine what degree of pollution reduction could be achieved by various levels of technology in the statute.
          2. § 509(b)(1) provides for judicial review of EPA’s actions in setting effluent limitations. However, no judicial review of § 304 guidelines. If the regs are “effluent limitations” under 301, then they’re reviewable in a court of appeals, but if DuPont is right and they’re 304 guidelines, then they can only be brought in the district court.
            1. This section also suggests by its language (“review of the Administrator’s action in setting an effluent limitation”) that this is the Administrator’s job. Problem is that this could refer to things other than § 301(b) standards.
  • Court: statutory language makes it clear that the 1983 effluent limitations are to be set by regulations. These effluent limitations are to have coercive power over and entire category or class of source, not in the course of issuing a permit to one member.
    1. Held: 301 unambiguously provides for the use of regulations to exstablish the 1983 effluent limitations.
  • Court: The language is different in the 1977 section—“point sources” rather than “categories and classes of point sources.” However, nothing else suggests a “radically different” mechanism. Thus is appears that this is merely poor drafting.
    1. Held: 301 also provides for setting the 1977 limitations by statute.
    2. Class/category issue: statute mandates the creation of class- and category-wide standards. This doesn’t make sense if you’re letting standards be set by each state, because you won’t have uniform standards. It’ll also dampen the technology-forcing effect, bc if these are state-by-state, you’ll only get a push to the best-performing source in the state. However, that’s not in the section that covers the 1977 BPT standards.
    3. Based on an argument, basically, that it would be really hard for EPA to set these standards by permit; we’ve already found that EPA is to set the standards (not state), and thus it can’t work this way.
    4. Rests this on the variance clause: “We conclude that the statue authorizes the 1977 limitations as well as the 1983 limitations to be set by regulation, so long as some allowance is made for variations in individual plants, as EPA has done by including a variance clasue in its 1977 limitations.”
  • Court: The statute isn’t the clearest thing in the world, but other sections indicate that these regulations shall be set by the Administrator.
    1. Held: the § 301 limitations are to be adopted by the Administrator, based primarily on classes/categories of plants, and will take the form of regulation.
    2. Held: the § 304 guidelines are basically preparatory work for the act.
    3. Held: § 306 contains no variance provision and none is required.
  1. Weyerhaeuser v. Costle''''''''''
    • Pulp and paper industry challenge to the 1977-1983 effluent limitations for their category. Argued that EPA should have done a better cost benefit analysis. EPA should also have balanced benefits against non-water quality environmental impacts to arrive at a net environmental benefit conclusion. This is based on some LH and on the requirement to take cost “into account”.
    • § 304(b)(1)(B) governs the process of setting the effluent limitations. Groups the factors that must be considered into two sets: cost v. benefit (“comparison factors”) and other factors (“consideration factors”).
      1. Congress “mandated a particular structure and weight” for the comparison factors (a “limited” balancing test): “By singling out two factors…for separate treatement…Congress elevated them to a level of greater attention and rigor.”
      2. Congress didn’t create a balancing structure for the consideration factors. EPA had discretion to decide how to weigh those.
        1. Limited because you’re not fully accounting for costs and benefits.
      3. How could you argue for CBA of all of the factors? It’s irrational not to do it, since you already have to go down that road AND you also have to consider these factors. Rationality requirements of the APA.
    • Did EPA weigh the comparison factors via a limited balancing test as required by Congress?
      1. Petitioners wanted an examination of the marginal costs and benefits. Try to distinguish the case from American Paper by pointing out that they supplied an accounting of the marginal costs and benefits here. Also look to some LH.
        1. EPA did a total approach, not a marginal analysis. Cf. Corrosion-Proof Fittings: that req’d an incremental approach. That’s a different statute, so it doesn’t control here.
      2. Court: EPA has some discretion with regard to how it decides to perform CB analysis. The 1977 standard did not place “primary importance” on the cost of compliance.
        1. But: when an incremental analysis has been submitted, the Agency should not ignore it. In this case, however, ignoring it didn’t matter because the incrmental anlysis wouldn’t have changed EPA’s decision.
        2. Marginal cost wasn’t high—stayed fairly constant.
      3. Variances''''
        • Types of CWA variances:
          1. § 301(c): Timetable for compliance can be modified by the administrator upon a showing by the operator that the modified requirement represents the best technology w/in the economic means of the point source AND represents reasonable further progress. Applies only to 301(b)(2)(A). Individual-firm bankruptcy variance—BAT standard over all is can go up to the point of wiping out the industry, but this variances allows individual firms to escape.
          2. § 301(g): Timetable can be modified for certain “nonconventional pollutants.” Applies to 301(b)(2)(A) only. Water quality variance
            1. Requires state approval
            2. At a minimum, compliance w b(1)(A) or (C) will occur (1977 limitations). So you have to meet the BPT standard.
            3. No additional requirements will be imposed on any other source as a result.
            4. Will not interfere with attainment of water quality necessary to protect public drinking water, wildlife, and recreational activities, won’t be an unacceptable risk to human health or the environment.
          3. “Fundamentally different factor” variance.
          4. NO VARIANCES EXIST FOR NEW SOURCES
        • EPA v. National Crushed Stone Association
          1. Challenge to pollution discharge limitations in the coal mining and crushed-stone industries.
          2. To get a variance from the 1977 BPT requirements, must show that factors of the production process are “fundamentally different.”
            1. A greater than normal cost of implementation will be considered in a variance request, but economic inability to meet costs will not.
            2. 301(c) variance does allow consideration of costs, but only in the BAT context. However, the DuPont decision found variance to be a fundamental part of the regulatory context.
          3. Must variance in the BPT context take the economic capability of the discharger to meet requirements into account?
            1. On its face, 301(C) is limited to variances from the BAT requirements.
            2. Language of 301(c) cannot easily be “exported” to the BPT requirements, because the underlying considerations are different.
              • (i) The 301(c) variance rules are designed to create a sort of mini-BAT for that industry: the best available technology economically available for that source (as opposed to category) that rep’s future reasonable progress.
              • (ii) However, the “reasonable further progress” requirement, for example, doesn’t make sense w/BPT—no baseline!
              • (iii) BAT variance can require you to regulate to the maximum economically feasible level, but BPT standards don’t require that of non-variance cases.
            3. Finally, BPT understood that a substantial number of point sources might close. A variance based on economic infeasibility doesn’t make sense. Congress understood the potential economic hardships.
          4. Chemical Manufacturers Association v. NRDC''''''''''
            • EPA developed FDF variance for cases that didn’t fit the baseline conditions it had used to create its models. Post-1977 amendments, EPA still used FDF variations now and again, and created some by regulation.
            • § 301(l) states that EPA may not modify any of the toxics requirements in § 301. EPA argues that this applies only to modifications explictly allowed by other provisions of § 301—just holds out toxics.
              1. An amendment to the standard is a permissible modification. What wouldn’t be permissible? Removing something from the toxics list by reg?
            • Are the FDF variances a “modification”?
              1. NRDC says yes.
              2. Court says that other provisions permit EPA to revise its own standards from time to time. “Modify” thus has no plain meaning and must be construed by EPA and the courts.
                1. EPA argues that the FDF variance is more like a revision permitted by 307 than it is like the other modifications in the statute. Court agrees: “It is, essentially, not an exception to the standard-setting process, but rather a more fine-tuned application of it.”
                  • (i) Does this make sense? Is this really a “modification”?
                  • (ii) What if EPA was just rewriting the categories? They could create a new subcategory if they wanted to. That’s how the court is treating this—as a redefinition of a subcategory.
                2. Court does not think that FDF variances threaten the structure of the act—“spelled out both by the Court and by the Agency itself.”
              3. Congress has not spoken plainly, open to construction, EPA’s constsruction is reasonable.
            • Problems with the logic of this decision:
              1. Categories created via notice and comment rulemaking.
              2. In this case, similar sources would, however, still have to get variances themselves.
              3. Okay, say there are 10 plants that are different and so the Admin decides to create a subcategory. He’ll force everyone to the level of the best source in those 10. If he used a variance, then none of them would be forced to shift. You lose the technology-forcing characteristic of the statute.
            • Is this a well-founded decision under Chevron?
              1. In terms of intent of Congress, does it make a lot of sense to allow 301(l) to kill the statutory variances and not the judicially-created non-statutory one?
            • Dissent in this case made the following argument:
              1. This is not a better standard for a new subcategory. It’s an impermissible move that defeats the technology-forcing elements of the statute.
              2. Agency’s interpretation is inconsistent w/the clear intent of Congress. Congress had expressed an intent to prevent modification of the toxics standards.
            • Post this decision, Congress added § 301(n) and created the FDF variance by statute. This was seen as a serious weakening of the toxics provision. So they added provision for the FDF variance in a limited number of cases to keep it from popping up everywhere. 301(n) is a much more stringent standard. Does not allow you to consider cost, frex.
          5. Non-point' Source' Pollution
            • We’ve been very successful at controlling point-source pollution. Yay, big success story! However, non-point sources haven’t been controlled effectively. Biggest problem: agricultural runoff.
            • How do you control nonpoint sources?
              1. BMP: how much pesticide you use, when you feed your animals, how you manage water, etc.
            • Statutory regime:
              1. § 319: State identifies the waters at risk; identifies the categories of sources that cause pollution; designs the process; submits a plan to the administrator for control of NPS; admin must approve plan.
              2. Similar to SIPS
            • Why not marketable permits? How would this work?
              1. Has the same economic incentives: point sources will have an incentive to buy permits bc it’s very expensive for them to reduce pollution due to stringent regulations. The NPS, having done nothing, will be able to reduce fairly cheaply bc they haven’t done anything.
                1. But the PS still have to meet the BAT standards so couldn’t pollute more. The permits may allow them not to have to further reduce their pollution or allow them to expand.
                2. Problem: What unit will the tradable permits be in? What if the NPS is not measuring a particular pollutant that it does not produce? Based upon expectation. What determines what it is expected to pollute? Develop a table that correlates Lbs of input (e.g., pesticide) to amount of output.
              2. Initial allocation: some kind of baseline measurement. Grandfather the NPS, who’d then make money by selling the permits. However, you have to establish the baseline level of pollution.
                1. This is where the “unit of environmental degradation” idea comes in.
              3. How do you define the permissible trading area? It’s going to have to be within the same watershed. Difficulty in defining this. Some watersheds huge!
              4. This is the argument of the unit of degradation measure—but raises problems, because you’ll have to buy units everywhere.
            • NPS has grown as a percentage of the problem because PS have been forced to reduce their effluent levels. However, WQS are still being violated, because of NPS.
              1. Two possible groups that can face more stringent regulations: PS and NPS.
              2. Trading is a way of aligning the incentives bw the two groups. 2003 Water Quality Trading Policy hasn’t really been implemented.
                1. This will be more complicated, because you aren’t dealing with single units ****. (W/ ghg units, conversion is easier b/c agricultural runoff involves so many different chemicals.)
                2. We probably won’t actually be able to measure that outputs anyway! How can you work a trading scheme under those circumstances? You’ll need proxies for the amount of effluent.
                3. Maybe you can use the difference in effluent bw current management practices and better management practices. So, permits for reducing pesticide, frex. But you still need some way of measuring and certifying.
  • Urban runoff presents different problems:
    1. It would be hard to create a marketable permit scheme for each individual auto driver.
  1. Water Quality Standards''''
    • Statute initially paid little attention to them.
    • Designated uses: fishable/swimmable the goal. If you want to designate for something else, you have to justify why you can’t reach f/s. § 131.10(j), (k)
      1. Minimum uses
        1. What use will actually be required?
      2. Anti-degradation
        1. How to require high quality over that required for a particular use.
        2. Statutory scheme: § 131.12
          • (i) Antidegradation program must maintain level of water quality necessary to protect existing uses.
          • (ii) Support wildlife and recreation, unless allowing lower water quality is “necessary to accommodate important economic or social development.”
          • (iii) Still have to protect existing uses and comply w/other PS requirements.
        3. Compared to PSD
          • (i) You can degrade down to existing use, even if you’re way above it.
          • (ii) But unlike in PSD, even if you’re not at top-level quality, you can’t degrade the hell out of something.
        4. Set by the states. Must submit plan to EPA for approval. WQS Regs § 131.5.
          1. EPA reviews for protection of PH and welfare, purposes of CWA, procedural requirements.
        5. UAA: can only downgrade due to natural factors, irremediable human-created conditions (dams that can’t be removed), severe economic dislocation leading to inability to meet 301(b) or 306. § 131.10(g). Economic provision allows you to prevent the closure of important plants.
        6. Why didn’t Congress force everyone to f/s standard? States can have bodies of water that are never f/s.
          1. CAA makes all air breathable. But we can’t escape the air. You can “escape” the water. Less of an impact on interstate commerce.
          2. Huge costs on existing sources by not grandfathering them. But not willing to impose the cost of bringing all waters to f/s standard.
          3. Note that this isn’t about drinking water—that’s the SDWA.
  • Criteria:
    1. Translation of the use into the pollution restrictions that support that use.
    2. You don’t get a restriction on what any particular pollutant can do, just on concentration. Allocating the pollution control burden is the second stage.
  1. Total Maximum Daily Loads (TMDLs)' (p569)'
    • Are the required effluent standards “not stringent enough to implement any water quality standard applicable to such waters” (§ 303(d))? à State must '
      1. list noncompliant bodies of H2O,'
      2. set the TMDL for each identified pollutant “at a level necessary to implement the applicable water quality standards with seasonal variations and a margin of safety which takes into account any lack of knowledge concerning the relationship between effluent limitations and water quality,”'
      3. and allocate TMDLs among polluters. (NPDES permits must be consistent with this.)'
    • Relationship bw this and 301(b)(1)(C)?
      1. Concern was that otherwise the Administrator would have to do this on a one-off basis every time a permit was issued. This is about aggregate reductions.
      2. Basically, just another oversight tool
    • No action on this for decades, but series of litigation put pressure upon states and EPA.
    • TMDLs are required by § 303(d) even for waters only polluted by non-point sources. Pronsolino v. Nastri, (9th Cir. 2002), p570:'
      1. Facts: EPA required California to list the Garcia River as a water body with insufficient pollution controls, even though the river was only polluted by nonpoint sources and, thus, effluent limitations did not apply. When Calif did not implement its draft TMDLs, the EPA implemented TMDLs. The Pronsolinos received a timber harvesting permit on the condition that they mitigate 90% of the controllable road-related sediment run-off and limit harvesting to specified seasons. '
      2. P’s argument: '
        1. Effluent standards “not stringent enough” implies that effluent standards apply to the body of water; however, if a water body is polluted only by nonpoint sources, es do not apply; thus, § 303(d) does not apply to such water bodies.'
        2. This interpretation is reasonable b/c the statute treats point and nonpoint sources differently in many instances.'
      3. Holding: The language of § 303(d) indicates that it applies to water bodies polluted only by nonpoint sources, and to the extent the language is ambiguous, the EPA’s interpretation is reasonable enough to be upheld.'
      4. Reasoning:'
        1. The CWA does not treat point and nonpoint sources differently under all circumstances.'
        2. If the statute were to require differentiating between
          (1) waters with one insignificant point source and substantial nonpoint source pollution and (2) waters with only nonpoint source pollution, this distinction would, “for no apparent reason, require the states or the EPA to monitor waters to determine whether a point source had been added or removed, and to adjust the § 303(d)(1) list and establish TMDLs accordingly.”'
        3. Federalism issues: The EPA’s actions did not upset the traditional federal/state balance b/c a state can choose both if and how it will implement the TMDL, although it will lose federal grant money if it does not.'
      5. Notes:'
        1. Lack of federal enforcement since states are not required to implement TMDLs.'
        2. State enforcement of a TMDL for nonpoint sources will not occur via effluent limitations but via nonpoint regulation methods.'
        3. § 301(b)(1)(C) allows the federal gov’t to set more stringent standards, but it would be very cumbersome to do so on a case-by-case basis.'
  • Estimation of margin of safety: NRDC v. Muszynski, (2d. Cir. 2001), p577:'
    1. Facts: NYC drinking water sources upstate have suffered increasing phosphorous pollution due to sewage and nonpoint runoff. NY placed nineteen reservoirs on a § 303(d) list and submitted TMDLs. EPA approved some but rejected others b/c not all the reservoirs had pollution that exceeded the level required for TMDLs. NRDC sued, alleging that the margin of safety required by § 303(d)(1)(C) and included with the approved TMDLs was insufficient.'
    2. (NRDC’s) P’s argument: The margin of safety (MOS) in the TMDLs does not address the “lack of knowledge concerning the relationship between effluent limitations and water quality,” as required by the CWA.'
    3. Holding: EPA’s reliance on the Reckshow Model for its calculation of MOS and its approval of the TMDLs was reasonable b/c based upon substantial evidence.'
    4. Reasoning:'
      1. P’s view would nullify Agency judgment: Requiring EPA to use a “rigorous scientific methodology” would prevent the Agency from acting to address clear public health or environmental dangers when the magnitude of such dangers could not be effectively quantified.'
      2. EPA’s approval contemplated revision of the MOS as more information becomes available.'
  • Leeway allowed before federal plan required—While EPA must establish a TMDL if a state’s TMDL is disapproved, EPA is not required to do so when a state either fails to submit or submits an inadequate TMDL (constructive failure to submit). San Francisco BayKeeper v. Whitman, (9th Cir. 2002), p583:'
    1. Facts: California did not submit TMDLs until 15 years after the deadline. Since then, California has submitted several and established a schedule for completing them.'
    2. P’s argument: California did not technically fail to submit TMDLs, but in essence it constructively failed to do so.'
    3. Holding: Although California had failed in the past to establish TMDLs, once it began to remedy the situation, the EPA was no longer obligated to establish water quality standards for the state.'
  1. Interstate Water Pollution''''
    • (1) EPA can require an upstream polluter to comply with a downstream state’s water quality standards. (2) The CWA does not require the EPA to ban discharges into an interstate waterway when the waters already violate water quality standards. Arkansas v. Oklahoma, U.S. (1992), p586:'
      1. Interstate water pollution plant. In granted a NPDES permit for an Arkansas plant, does EPA have to weigh the Oklahoma’s water quality standards?
      2. EPA’s CJO decided that § 301(b)(1)(C) required them to avoid a “detectable” impairment of water quality in Oklahoma.
        1. EPA’s permit incorporated OK water quality standards.
      3. How does this compare to the CAA? (See p593n5)
        1. CAA provisions allow the state to consume the entire increment.
        2. In this case, there are no federal standards. What’s the problem with that, though?
          • (i) They give the entitlement to the downstream state. Basically an option to veto industrial activity in the upstream state, regardless of the potential benefits to both states. Only required “detectable,” not even “significant” degradation.
          • (ii) No balancing of the interests.
          • (iii) Problems with giving entitlement to more protective state:
            • (1) The state may be doing so for “bad reasons” (to hurt out-of-state producers while it does not limit its in-state producers, who pollute other water bodies, in the same way).
            • (2) It will probably not bear the costs (externalizes them).
  2. Ricky things that in the CAA case, the balance is too far towards the upwind source of the pollution, and that in the CWA case, the balance is too far towards the downstream state.
  3. Maybe Coasian bargaining is more possible in this context?
  4. Inquiry made more complex by the absence of federal standards. So the downstream state’s stricter standard gets to control.
  5. Court is essentially saying that EPA interpretation is permissible, not that it’s compelled. Read narrowly.

HAZARDOUS SUBSTANCES – RCRA and CERCLA[edit | edit source]

  1. RCRA (Resource Conservation and Recovery Act) – ex ante regulatory scheme.
    • Provides ex ante regulation of generators, transporters, and operators of disposal sites. Much of the regulation based on tracking the waste. Chain of manifests that theoretically allows you to track the waste.
  2. TSD (treatment, storage & disposal) facility.
  • Extensive permitting and licensing reqs, esp. for TSD operator.
  • Only applies prospectively.
  • Even with this system, leaks can occur, however. And then CERCLA kicks in. So even if you followed all the rules of RCRA, the possibility for CERCLA liability still exists.
  • CERCLA isn’t just about past actions. At the time of passage, it was, but under statute it applies to both the new and the old.
  1. CERCLA (Comprehensive Environmental Response, Compensation and Liability ACT) aka Superfund – ex post regulatory scheme.
  2. Why have both regulatory schemes?''''
    • Same regulation v. liability arguments as we’ve seen before. A mixed regime can address the pros and cons of the two separate types.
    • Liability puts the decision in the hand of the risk-causing entity. This removes the burden from govt. to regulate every single thing. Regulating the TSD is easier for the government, but regulating the amount of waste you can put out is harder.
    • If it’s just ex ante regulation, TSD has no incentive to step in and do the cleanup. Under an ex post system you do, because you might be able to do it more cheaply. You also have an incentive to find the pollution earlier, because it’ll cost you less.
    • HOWEVER, these incentives only work if there are enough parties with solvency out there who will worry about putting that solvency at risk.
  3. PRPs—Potentially Responsible parties § 107(a)(1)-(4)
    • Definitions in § 101.
    • Current owner or operator of a vessel or facility at which the release occurs. Always liable, regardless of whether disposal occurs during tenure. New York v. Shore Realty Corp., (2d Cir. 1985), p642:
      1. Facts: Shore wanted an interpretation of § 107(a) that would exempt a current owner from liability. Argued that b/c (a)(2) limited liability to owners at the time of disposal, (a)(1) should as well.
      2. This is a pretty strained reading. It would make (a)(1) irrelevant. There is no way to make both provisions relevant and to accept Shore’s reading.
      3. Current-owner liability means that before accepting gifts of or purchasing property, institutions and companies should conduct environmental assessments.
    • Prior owners and operators
      1. only if disposal of hazardous substances occurred during tenure
      2. or if sold/transferred the property w/o disclosing knowledge of release or threatened release
    • Generators who arrange for disposal, treatment, or transportation
    • Transporters, if they are involved in choosing the site: “selected by such person.”
      1. Some legal questions about what involvement you have to have in the selection.
      2. If the generator specifies the facility and you just take it there, you’re probably not liable.
      3. If you drive in a leaky truck, are you liable? Argue this as an implicit site selection: you chose for things to leak all over the road.
  4. Release § 101(22)
    • Once a hazardous waste barrel is placed on the site is it a release? Probably not. Is it a threatened release? Maybe if it’s starting to degrade.
    • Proper disposal in an improper site probably counts as a threatened release as well.
    • “Threatened” is not defined. Ricky thinks you tie this to some increased probability of release—if it’s a lot higher than the probability if the stuff was being treated properly.
    • “Response costs”: “response” includes remove, removal, remedy, and remedial action and enforcement activities related to these things.
      1. Difference bw removal and remedial action is important: govt can only spend on remedial action if the site is on the NPL.
    • So the EPA inspector can’t run in and sue the responsible party if they just see the possibility of release—some cost has to be incurred first. However, if you do an assessment and incur costs, you can ask for an injunction requiring the PRP to pay the prospective response costs.
  5. Hazardous substance § 101(14)
    • Any substance designated under the CWA, any substance designated under CERCLA, any substances with the “characteristics identified” by the SDWA—even if it’s not actually listed, it’s a hazardous substance for these purposes if it has the characteristics, toxic pollutants, hazardous air pollutants, but NOT hazardous wastes.
  6. Extent of liability § 107(a)''''
    • § 107(a)(4)(A): removal or remedial action costs incurred by a public actor “not inconsistent with the National Contingency Plan”
      1. National Contingency Plan: § 105, for some reason, delegates power to the President (he’s delegated it back to the Administrator). Sets forth procedures for the conduct of cleanups. If the govt wants to recover under § 107(a)(4)(A), has to follow those procedures.
    • § 107(a)(4)(B): removal or remedial action costs incurred by private parties, if “consistent with” the NCP. So govt gets more leeway, can do additional things if NCP doesn’t forbid them, but private parties can’t.
    • § 107(a)(4)(C): natural resource damages. Difference bw this and removal damages? If remediation consists of removing all the barrels, carting off soil, etc. NRD would include the persistent damage: so now you have a big hole in the ground. Losses that attach after the removal/remedial action has taken place. The hazardous substance may be gone, but the quality of the resource may remain permanently degraded.
    • § 107(a)(4)(D): health assessment costs. Cost of figuring out if there have been health impacts as a result.
      1. What’s missing? Damages for the health effects. This isn’t a federal toxic tort statute. An earlier draft would have created a federal cause of action, but it didn’t pass. They can sue under a state law. No preemption of those claims.
    • § 107(f): NRD only apply to public lands. Private landholders don’t get compensation for NRD. What it means for land to be controlled by the gov isn’t entirely clear—what about an easement, frex?
    • Strict liability statute. Challenges to the retroactivity, but Congress clearly intended this.
  7. Joint and Several Liability
    • Not explicitly provided for in the statute.
      1. But § 113, added with 1986 SARA amendments, contains a provision for contribution, which would only make sense if liability were joint and several.
    • Rule: If harm divisible or reasonable basis for apportionment à not J&S liability.
    • Rule: If harm indivisible à J&S liability.
      1. You can claim against one party for the full harm. This allows you to recover full costs even when some PRPs are insolvent.
      2. If there’s not J&S liability, it is the plaintiff’s responsibility to find all the parties. With J&S liability, the plaintiff can choose whom to recover the judgment from.
    • So one PRP may bear the burden of litigation, of finding the other PRPs, and the upfront cost of the judgment.
    • This assumes there is J&S liability w/contribution. Absent contribution, that PRP would have not relief.
      1. Common-law rule of J&S liability didn’t contemplate a right of contribution.
      2. § 113(f) of CERCLA has an express contribution provision.
    • What about indemnification? § 107(e)
      1. Seller/buyer, frex.
      2. Transfers entire liability, as opposed to a share.
      3. 107(e)(1): liability “under this section” can’t be transferred.
        1. If the govt goes after the buyer, the buyer can’t use the indemnification as a defense. But it would be a separate action.
        2. If the seller’s insolvent, the buyer has to pay the full action. They can sue upon it, but the suit is worth nothing.
        3. The beneficiary of the indemnification agreement is taking the risk w/r/t the indemnifying party’s solvency.
      4. Divisibility of harms and contribution § 113(f)(1)
        • Divisibility of harms
          1. Where does the distinction come from? Common law
          2. But if the court finds the harm to be indivisible, the PRP will subsequently pursue the other party in contribution. Then the court will have to divide it. Does this make sense?
            1. At the contribution phase, the court uses equitable factors to assess the party’s responsibility. The “Gore Factors” (p. 43).
              • (i) Most important one: volume of waste.
              • (ii) Second one: hazardousness of particular type of waste.
            2. So really? This doesn’t make any sense. Sometimes Gore factors are used to determine that something’s indivisible. Sometimes they’re used at contribution phase. See Chem-Dyne.
          3. Courts are always willing to apportion in a contribution action. So when is it apportionable in a way that defeats J&S liability?
            1. Obviously govt will favor broad application of J&S liability bc if you apportion at the contribution phase, then the parties bear the cost of insolvent parties.
            2. So maybe the Gore factors are most appropriate at the contribution phase (aside from the first one, which focuses on the distinguishability of the shares).
              • (i) The others aren’t causation-based.
            3. Revesz thinks there must be a conceptual difference bw these stages. First one must be causation-based. Note that these came out of the R2T, but recent Supreme Court case (Burlington Northern & Sante Fe Ry.) confuses them and allows apportionment at the liability stage. Unclear how this will affect future cases.
  • PRPs that do not satisfy the civil action requirement of § 113(f) cannot bring a contribution action under § 113(f). Cooper Induss., US, p729.
    1. PRP had undertaken a voluntary cleanup and then sued other PRPs for contribution under § 113. The court said that if a cleanup is undertaken voluntarily, there is no right to contribution.
    2. Language of § 113(f)(1): “may seek contribution during or following any civil action under section § 9606.” So since nobody sued them, they can’t recover.
    3. Can a party engaging in voluntary cleanup ever recover under § 113? No.
      1. But maybe the implied right to contribution in § 107(a)(4)(B): “any other person” can recover costs. Court remands for consideration of potential § 107 claim. Key-Tronic found an implied cause of action for PRPs under § 107. Sort of dicta in a case over attorneys’ fees.
    4. So why didn’t the court just let them recover under § 107? Maybe they didn’t want to decide if this is cost recovery or contribution.
      1. Because if they have to class this as cost recovery, then they have to decide about what happens with all of the shares of insolvent parties.
      2. If it’s characterized as cost recovery, then you have the chance to get there first and impose J&S liability on the other PRPs. But then they might file contribution actions against the other responsible parties and try to impose some of the orphan shares back onto you.
    5. Policy issues
      1. Want to incentivize private party cleanup by favoring them in order to expedite cleanup actions.
      2. Cleanup party (e.g., current owner) does not get an unfair advantage, necessarily, because the court can apportion “orphan shares” of liability equitably.
    6. Defenses: § 107(b)''''
      • Act of God: “something no reasonable god would do.” § 107(b)(1)
      • Act of War § 107(b)(2)
      • Act or omission of a third party, other than an employee or agent of the defendant or whose action occurs in conjunction with a contractual relationship btwn the two. § 107(b)(3)
        1. Must be the sole cause of the release
        2. Responsible party must have exercised due care w/regard to hazardous substnace AND taken precautions against foreseeable acts or omissions of third parties.
        3. If a third party caused only part of the harm, you wouldn’t have a defense. “damages resulting therefrom were caused solely by…”
      • So if you own some land, and one day you turn up and somebody randomly put hazardous waste on your property, are you responsible when the fedgov sends you an enormous bill for the release?
        1. The PRP carries the burden of the defense, so you’d have to know who had done it.
        2. Once you figure out that it’s Random Dude X, what do you have to do? You must have exercised due care when you discovered the substance. Maybe this means consulting the government, etc.
        3. What about precautions against foreseeable acts? Like if you’d found one barrel, thus it would be foreseeable that more dumping would occur. You’d be expected to take all reasonable precautions to avoid further dumping. The first one wasn’t foreseeable, but subsequent ones probably were.
        4. Final hurdle: Random Dude X actually has to be able to pay. Because if you can’t establish the defense, you can bring a contribution action against the third party so long as he is solvent. Even if you’re liable, you may be able to get it all in contribution.
        5. If you remove the stuff yourself, consistent with the NCP, then you can try to recover under § 107(a)(4)(B).
      • How do we know that the Δ carries the burden of asserting the defense?
        1. “caused solely by” in the defenses. Burden is on the PRP to show that someone else caused the problem.
          1. If the burden was on the P, then the Δ would escape liability absent a showing of causation.
          2. But here the P didn’t have to show that the Δ caused the problem. PRP shows causation to escape liability.
  • Innocent Landowner Defense, § 101(35)
    1. Technically a subset of the Third-Party Defense
    2. Clarified that the third party defense can be used by current owners
  1. The Superfund' § 111'
    • Fund that the government has built up through various taxation schemes: corporate taxes, general revenue, sector-specific taxes, etc.
    • Who should pay?
      1. General revenue—by everybody, basically
      2. The industry where the problem is caused, regardless of the responsibility of individual firms
      3. People who actually have a connection to the problem. Pay via a liability regime.
    • Basically, the question is who you prefer. Someone with some connection to the problem, or just general revenue. Your connection is likely to be attenuated no matter what.
    • Superfund is very unpopular, for one b/c companies can be liable even if they complied with all applicable law at the time.
      1. In the end, there are a lot of contaminated sites and the alternative is to raise more money through taxes. But taxes have to be voted on, whereas the liability regime is already in place. Funds raised in liability actions, untraceable to a member’s vote.
    • Currently, the tax has expired and has not been renewed.
  2. Lender liability § 101(20)(A) (p658)
    • Pre-1996, banks were liable post-foreclosure and at any point pre-foreclosure when they become heavily involved in operation of the mortgaged asset.
    • 1996 Amendments (added § 101(20)(E)). Owner means:
      1. Person who owns a facility or vessel
      2. Person who previously owned, operated, or controlled a facility if the facility was conveyed to state or local gov’t via bankruptcy, foreclosure, tax delinquency, abandonment, or similar means. (§ 101(20)(A)(iii)).
      3. Not a unit of government who ends up with title to a thing via foreclosure.
      4. Not a person who, without participating in the management or ownership, holds indicia of ownership as a security.
        1. Problem: how do you interpret “without participating in the management of the investment or facility”? Is monitoring participation?
      5. § 101(20)(E)(ii): you’re not an owner or operator if you didn’t participate in management prior to foreclosure and you seek sale “at the earliest practicable, commercially reasonable time.”
        1. But nobody’s going to want to buy this, so what should you do? Advertise it every six months or something? But you can’t misrepresent the nature of the property either.
      6. § 101(20)(F): defines participation in management. Must be actual participation, not an unexercised right to do so. Providing financial counseling in an attempt to avoid default or diminution in value of the facility is exempted.
    • Why give banks such a good deal?
      1. Don’t want to reduce the availability of credit too far.
      2. But then why not exempt banks all the way? Don’t want to let them go ahead operating a polluting site.
        1. Want to create some incentive for the bank to inspect land before they extend credit. Bank is a repeat player, and thus can probably do this more cheaply and effectively than a small company that may rarely do this sort of thing.
        2. Their security interest will be worth nothing if the property becomes a Superfund site. Though if it’s then cleaned up by the gov’t, they get a windfall.
      3. Revesz thinks liability at foreclosure was actually a good idea. Prior to foreclosure, some liability (because of potential scams). Banks could always choose not to foreclose and lose their investment.
    • US v. Aceto Chemical Corp
      1. US attempting to recover from agricultural companies that send chemicals to Aidex to be made into pesticides. Aidex a formulator. Argument is that though Aidex formulated the pesticides, Aidex and the other companies owned the chemicals and the finished pesticides throughout. Argued that because waste creation was a necessary incidental effect of pesticide formulation, that Aceto and other companies had therefore “arranged for” disposal.
      2. How does it serve the purposes of the statute to go after Aceto etc.? You’re going to get Aidex anyway.
        1. Maybe if Aceto is exposed to liability, this will create an incentive for the company to choose the formulator that is most likely to be careful with regard to its disposal practices.
        2. Is the best thing to do to internalize the costs all along the supply chain? Do you want to create incentives for everyone to deal with the most responsible parties?
        3. Lets you bring more solvency into the liability regime.
      3. Why not do this?
        1. The holding of this case depends on formal ownership. If that’s the full basis, then companies will transfer ownership prior to and after formulation. If it was the sale of a hazardous waste (which has no value), then the court’s probably going to see that as a sham. But if there’s actually a useful product, then the distinction does seem to turn on who has title to the substance.
      4. What about individuals? What about when you take your suits to the try cleaner?
        1. Most individuals throw out hazardous substances in their household garbage.
        2. PRPs: willing to press suits against small towns and municipalities because they know that municipalities going out of business will provoke a backlash in Congress. So they oppose reform that will protect these entities.
          • (i) tries not to sue “small attractive parties” but PRPs will bring them in via contribution actions.
        3. Land-buyer liability: the “innocent landowner” defense
          • Is a contract for the sale of land a contract for superfund purposes?
            1. It matters because of § 107(b)(3)—the third-party defense. If you want to say that the problem was caused solely by the seller of the land, it matters if a contract for the sale of land creates a contractual relationship.
            2. The defense: § 101(35): Contractual relationship includes “land contracts, deeds, easements, leases…unless the real property on which the facility is concerned is located was acquired by the defendant after the disposal or placement of the hazardous substance…and one or more of the circumstances below is also established.”
              • (i) defendant didn’t know and had no reason to know
              • (ii) defendant is a government and got the land involuntarily
              • (iii) defendant acquired the land by inheritance or bequest.
            3. So if disposal continued after the contract, it’s a K for Superfund purposes.
  3. So to establish the defense, what do you need?
    1. The above, and then you have to go back to § 107(b)(3) and show that it was solely the responsibility of that party whose in a purchase relationship with.
      • (i) That you took action against foreseeable acts or omissions of the party (not really relevant here, bc you weren’t present when the acts or omissions took place)
      • (ii) That you took due care (§ 107(b)(3)(b))
    2. Courts call this the “innocent landowner” defense, but Revesz argues it doesn’t really create a new defense—just redefines “contract” under the third party defense.
    3. Reason to know
      1. § 101(35)(B): all appropriate inquiries; took reasonable steps to stop any release or prevent a threatened release or exposure.
        • (i) (ii) Administrator to promulgate regs including criteria in (iii).
        • (ii) (iii) Environmental assessment, interviews w/past owners, operators, etc.; reviews of historical sources/title docs, record searches, visual inspections, relationship of purchase price to uncontaminated value (did you get a suspiciously good deal?), commonly known or ascertainable information, degree of obviousness of the contamination. Specialized knowledge of the defendant matters.
      2. Statute used to equate all appropriate inquiry with “customary standards”. So you’d check the chain of title, do some visual inspection, etc.
      3. Pacific Hide & Fur a good case on this
      4. Westwood a good case on sellers asserting this defense
    4. The Clean-Up Process (p733)
      1. Site identification
        1. No federal effort to go find the superfund sites. They do have a database (CERCLIS), and things get onto that list via citizen reporting, state and local governments, etc. About 36,000 sites.
  • Steps in the process
    1. Preliminary Assessment/Site Investigation
    2. HRS—if it’s high enough, goes on the NPL.
    3. Once it’s on the NPL, RI/FS
    4. Then EPA creates the Record of Decision, which says what the remedial action will be. Have to examine the alternatives, do public consultation, etc.
    5. RD/RA: remedial design and remedial action.
    6. Takes a really long time (12-16 years)
  • NPL v. non-NPL sites
    1. NPL sites get the structured assessment, and superfund money can be used for removal and remedial actions. For non-NPL sites, Superfund money can only be used for removal actions.
    2. Superfund money can be used for cleanups when there are not solvent parties.
      1. And to fund cleanups prior to cost recovery. Use the Superfund money to get the case started, and then reimburse the fund post.
    3. There’s not a lot of money in the Superfund. Government hasn’t appropriated much money. Has mostly stopped putting things on NPL, because not much is coming off. Very few cleanups have actually been completed. This is embarrassing for EPA.
  • Clean-up standards § 121
    1. First, must be in accordance with the NCP
    2. Permanent and significant reduction in the volume, toxicity, or mobility of the pollutant is preferred. § 121(b)(1)
    3. Relevant standards, § 121(d)(2): cleanup must meet the “standards, requirements, criteria, or limitations” of other federal laws, ARARs.
      1. SDWA: § 121(d)(2)(A)(ii) requires you to meet MCLGs under the SDWA. However, Ohio v. EPA, p739, held that EPA was reasonable in enforcing only the MCL requirement for Superfund sites when MCLG is set at zero. Under SDWA, the MCL is the enforceable standard, not the MCLG. But here it’s “converted” into an enforceable standard. EPA argued that MCLGs are not “relevant and appropriate under the circumstances,” as required by the statute, when the MCLG is not measurable.
      2. Why would Congress have wanted Superfund sites to clean water to a greater extent than required for most sources under the CWA? Perhaps the only explanation is that we can’t control the future use of the groundwater and want to preserve the potential use of it for drinking water, per Viscosi article (p754).
    4. ARAR = state environmental requirement or standard that is (1) timely promulgated, (2) more stringent than federal standards, (3) legally applicable or relevant and appropriate, and (4) timely identified. See § 121(d)(2)(A)(ii)
      1. Relevant and appropriate = standard under federal or state law that “address[es] problems or situations sufficiently similar to those encountered at a site that their use is well situated to that site.” U.S. v. Akzo Coatings of Am., Inc., p748, 752n38.
    5. Cost-benefit and clean-up standards: “The President shall select appropriate remedial actions…which provide for cost-effective response. In evaluating the cost-effectiveness of proposed alternative remedial actions, the President shall take into account the total short- and long-term costs of such actions, including the costs of operation and maintenance for the entire period during which such activities will be required.” § 121(a).
      1. However, cost-benefit doesn’t really come in where it would help the most: you’d assume that we’d want to impose stricter standards in areas of greater population density, but ARARS are most likely to be waived in urban areas instead.
      2. Note that this is cost-effectiveness, but not really CBA. No structural analysis requirements like the comparison factors in CWA.
    6. Hamilton & Viscusi article, Land-Use Solutions (p754)
      • The principle risk being controlled by Superfund cleanups is the risk to future residential users. Typically not even a risk to workers at a site if the site became a residential facility. So why spend all of this money when we could control this via the land-use regime? Deed restrictions, e.g.
        1. How much can you count on land-use restrictions? Love Canal example—ended up building a school on top of it.
        2. Informational issue—ensuring that future potential users will know that this was a Superfund site. NCP requires a public hearing and the affected community. But what happens 30 years later?
      • Think about the brownfields issue. Industrial facilities have to locate somewhere. Should we create incentives for industrial facilities to locate in greenfields or to go to sites that are already somewhat polluted, do some necessary cleanup, and preserve the greenfields for clean activities?
      • The Bona Fide Prospective Purchaser protection, § 101(40)
        1. In the early days of Superfund, it was difficult to get industrial facilities to move into brownfields due to the threat of liability.
        2. EPA didn’t want to have an across-the-board policy, but they’d issue “comfort letters.” However, comfort letters weren’t binding, and didn’t keep other people from suing them (other PRPs).
          1. For a long time, only the Chicago region did this.
        3. Bona fide prospective purchaser protection, § 101(40)
          1. Purchasers who know about toxic materials present on the site, and who takes takes appropriate steps to control it is protected.
          2. So why doesn’t everyone do this? Because you do still incur response responsibilities (§ 101(40)(D). More likely to be industrial operators who take this on.

The National Environmental Policy Act (NEPA)[edit | edit source]

  1. First of the “big” environmental statutes. At the time, it was thought to be very important, but it hasn’t turned out to be such a big deal in retrospect.
  2. Policy in favor of Informational Regulation (p793):
    • Enhances market efficiency
      1. Managers can’t correct problems they don’t know about
    • Create incentives for self-policing
      1. Businesses may be concerned about consumer choices or legislative/regulatory backlash
    • Improves the effectiveness of command-and-control regulation
    • Promote democratic values by making information available to the public to make individual choices.
  3. § 2 - General aspirational goals
  4. § 101(b) – sets out more specific goals
    • Why do we need any environmental laws with these provisions? Why couldn’t the EPA promulgate regulations that would be reviewed under the APA?
      1. Congress may not have contemplated the CAA and CWA, etc. Other countries have laws like NEPA as their main environmental statute.
      2. Given the difficulty proponents of environmental standards have experienced under CWA and CAA, it’s not clear that a broader statute would have been very effective.
  5. § 102(2)(C) - Environmental Impact Statements (EISs)
    • Is the project a “major federal action” or “proposal for legislation” that “significantly affects the human environment”? à EIS required. § 102(2)(C).
      1. “Significant” effect—normally determined in reference to context and intensity. 40 CFR § 1508.27 (CEQ regulation).
      2. To make a threshold determination of “significance,” an agency must provide the public with notice and opportunity to comment. Hanly v. Kleindienst, (2d Cir. 1973), p828
      3. non-federal action but “federal nexus”? à considered federal action.
      4. Uncertainty re: local impact = controversial = significant à EIS required. Anderson v. Evans (Alaskan whale hunting case), (9th), p835.
      5. Non-discretionary action of agencyànot subject to NEPA’s EIS requirement b/c the action is not the legal cause of any environmental impact (Agency can’t not do it). DOT v. Public Citizen, U.S. (2004), p870.
    • Regulations to determine if an EIS is required:
      1. Circumstances under which an EIS is normally required? à EIS, unless gov’t demonstrates the existence of no significant impact
      2. If EIS not normally required à no EIS (categorical exclusion), unless substantial evidence is presented that it will have a significant impact.
      3. If unclear à EA to determine if EIS is warranted.
  6. § 202 - Created the CEQ (Council on Environmental Quality) as an advisory board to the President and to promulgate regulations under NEPA.
  7. NEPA was ultimately interpreted by the Supreme Court to have no substantive requirements.
    • Federal agencies must give sufficient weight to environmental factors. Calvert Cliffs’ Coord. Com., Inc. v. U.S. Atomic Energy Comm’n, (D.C.), p798.
    • Court review of agency action under NEPA: procedural, not substantive review. Strycker’s Bay Neighborhood Council, Inc. v. Karlen, U.S. (1980), p808 (citing Vermont Yankee).
    • What is substantive or procedural? Played out in Robertson v. Methow Valley Citizens Council, U.S., p810
  8. Other rules
    • An EIS is triggered by a proposal for federal action, not mere contemplation of such action. Kleppe v. Sierra Club, p819.
    • Based upon CEQ regs, for multiple actions: if “connected” or “cumulative” à cumulative EIS is required. Thomas v. Peterson, (9th Cir. 1985), p822.
      1. “connected actions” = necessary, interdependent actions
      2. “cumulative actions” = have cumulatively significant impacts.
    • Positive environmental impacts à no EIS required. Friends of Fiery Gizzard v. Farmers Home Administration, (6th), p845n5.
    • When an Agency has decided not to consider certain alternatives, doing so is only required if the alternatives are raised in a clear manner. Vermont Yankee Nuclear Power Corp. v. NRDC, Inc., U.S., p846:
    • Uncertainty: When an Agency, after preparing an EIS, has weighed uncertainty re: environmental effects and decided the effects are outweighed by the benefits of proceeding à defer to Agency. Alaska v. Andrus, (D.C.), p851.
    • New information:
      1. Although an Agency normally has discretion to decide whether or not a supplemental EIS is warranted by new info, the ACE (Army Corps of Engineers) did not demonstrate in this case a reasonable consideration of info known at the time. (Injunction upheld). Sierra Club v. U.S. Army Corps of Engineers, (2d.), p856.
      2. After the completion of an EIS, if the major federal action is still pending and new information indicates the remaining action could significantly affect the environment à same standard as EIS (hard look + reasonable consideration of info). Marsh v. Oregon Nat. Resources Council, U.S., p861.
  9. No private right of action.
  10. Judicial review only per APA.
    • Proposal of legislation—No challenge under NEPA b/c no private right of action. APA challenge possible? Is the alleged proposal for legislation agency action?
      1. If No (e.g., it was Presidential action)àno review under APA.
      2. If Yesà was it a final action (does it have the capacity to affect interested parties)?
        1. If Noà no review under APA.
        2. If Yesà judicial review under APA. Public Citizen v. U.S. Trade Representative [OTR], (D.C.), p865.
  • Judicial review of agency decision to conduct an EIS or not:
    1. (1) Did agency take a “hard look” at possible environmental effects?
    2. (2) Was the agency’s decision arbitrary or capricious?
    3. Mitigation Example: Forest Service justified FONSI based upon mitigation measures (making road appear that it wasn’t there so that ATV users wouldn’t cause an adverse impact) but discussed no evidence that the mitigation measures would be effective or even implemented. Nt’l Audubon Society v. Hoffman, (2d), p841.
  1. The legacy of NEPA? (p814n7)
    • Requirement to do EISs has precipitated potential changes to internal agency culture by the need to hire people w/ diverse expertise and viewpoints
    • Environmental groups with limited resources benefit b/c
      1. other parties pay for information gathering
      2. and are provided opportunities to participate in decision making process and put media or political pressure upon agencies
        1. But pro-development agencies conduct EISs
  • If the EIS is inadequate, then an agency can get sued on it. The agency takes a risk, because the project will be enjoined if the EIS is procedurally inadequate. Lots of delay involved.
  • Also, many state statutes have substantive components.

'

The Endangered Species Act[edit | edit source]

  1. Generally
    • Listing: § 4—Has Secretary (of Interior or Commerce) listed the species as endangered or threatened?
      1. Either in response to a petition (w/in 12 months) or of own initiative.
      2. Based “solely on the basis of the best scientific and commercial data available” (w/o consideration of cost of protection). § 4(b)(1)(A).
      3. Definitions:
        1. Species, § 3 = species, subspecies, or any distinct population segment (DPS)
          • (i) “distinct” (per NFMS regs, Nt’l Ass’n of Home Builders v. Norton (9th Cir. 2003), p899) =
            • (1) discrete = delimited by boundaries (incl. int’l boundaries)
            • (2) significant to the taxon
          • (ii) Very different conservation measures – discrete but not necessarily significant
        2. “endangered” = in danger of extinction through all or a significant portion of its range
        3. “threatened” = likely to become endangered in the foreseeable future
  • Critical Habitat Designation
    1. If species listed à CH designation, per § 4(a)(3)(A)(i).
      1. May take economic impact or other relevant impact into consideration.
        • (i) But excluding an area from CHD cannot result in extinction. 4(b)(2).
      2. Features to consider (primary constituent elements –PCE) and to be listed along with c.h.d., per Fish & Wildlife Service (FWS) and Nt’l Marine Fisheries Service (NMFS) regulation, p929:
        • (i) space for individual and population growth, and for normal behavior;
        • (ii) food, water, air, light, minerals, or other nutritional or physiological requirements;
        • (iii) cover or shelter;
        • (iv) sites for breeding, reproduction, rearing of offspring, germination, or seed dispersal; and generally
        • (v) habitats that are protected from disturbance or are representative of the historic geographical and ecological distribution of a species.
      3. All economic impacts must be considered; thus, the baseline method (excluding impacts of listing) is invalid, per N.M. Growers Ass’n v. U.S. FWS, p929.
      4. Critical habitat must be designated “to the maximum extent prudent and determinable.” § 4(a)(3).
        • (i) Per FWS/NMFS regs, p940
          • (1) Not prudent =
            1. Species threatened by taking or other human activity, and identification of c.h. can be expected to increase such threat, and/or
            2. Designation would not be beneficial to the species.
          • (2) Not determinable =
            1. Info sufficient to perform required analysis of impacts of designation is lacking, and/or
            2. Biological needs of species are not sufficiently well known.
          • Federal jeopardizing action: § 7(a)(2) –Is action authorized, funded or carried out by a federal agency
            1. likely to jeopardize the continued existence of any endangered species or threatened species
              1. Exception—nondiscretionary action of a federal agency (e.g., delegating enforcement of CWA to a state authority) à not subject to jeopardy question. p918n3.
              2. God Squad exception – see below.
            2. or result in the destruction or adverse modification of habitat of such species which is determined to be critical (critical habitat destruction – CHD)?
              1. Prohibited even if no actual members of the species were harmed.
            3. Consultation – a mandatory procedural requirements for agencies:
              1. Agency inquire of Secretary—Whether any threatened or endangered species “may be present” in the area of the proposed action? § 7(c)(1)
              2. If yes, agency prepare a biological assessment to determine whether such species “is likely to be affected” by the action? §7(c)(1). (Can be part of EA or EIS)
                • (i) (can be challenged under § 11)
              3. If so, agency consult with Secretary resulting in a “biological opinion” (BiOp) specifying if action would jeopardize or destroy CH:
                • (i) If violate ESA, may go forward only if alternative cures violation. 7(b)(3)(A).
                • (ii) If not violate ESA, Secretary may still require measures to minimize impact.
              4. Agencies must substantially comply with ESA’s procedural requirements. Thomas v. Peterson, (9th Cir. 1985), p945 (granting an injunction against a Forest Service project b/c the agency did not consult with the FWS or complete a biological assessment)
  • “God Squad” Exemption, § 7(e)
    1. Endangered Species Committee can grant an exemption from “no jeopardy” requirement if:
      1. No reasonable or prudent alternatives
      2. Benefits clearly outweigh benefits of alternative action consistent with conserving the species or its critical habitat, and such action is in the public interest
      3. Action is of regional or national significance and
      4. No irreversible or irretrievable commitment of resources prohibited by sub§ (d) has been made by the Agency concerned or exemption applicant.
    2. Must formulate reasonable mitigation and enhancement measures
    3. Decisions appealable to circuit courts.
      1. Example: Timber group requested exemption. God Squad members had discussions with President and ruled for petitioner. Environmental groups challenged. Court ruled that the APA prohibition on ex parte communication was applicable. Remanded to determine if the timber sale exemption involved such communications. If soà exemption invalid. Portland Audubon Society v. ESC, p922.
  • Federal obligation to conserve, § 7(a)(1), p950
    1. No affirmative obligation. Agency has discretion and is not obligated to choose the least burdensome alternative to an action just b/c it may be more conservative of an endangered or threatened species. Pyramid Lakes, p950.
  • Private taking action: § 9 (a)(1)(B)—Has a person (including private and government entities, but see p967n9 for exception)
    1. “take[n]” a listed species? (i.e., harassed, harmed, pursued, hunted, shot, wounded, killed, trapped, captured, or collected, or attempted to engage in such conduct)
      1. harm = “significant habitat modification or degradation that actually kills or injures wildlife.” Babbit v. Sweet Home, p955 (upholding regulation).
      2. Per O’Connor, limited by ordinary principles of proximate cause (p964).
    2. Habitat modification that impairs the breeding of a species = harm? (p965, Palila decisions)
      1. Yes. Marbled Murrelet v. Babbitt, (9th Cir. 1996).
      2. Not to an identifiable animal, per Scalia in Sweet Home.
      3. Yes, to an identifiable animal, per O’Connor in Sweet Home.
    3. Exemption for “scientific purposes” - § 10(a)(1)(A)
    4. Exemption for “incidental taking”- § 10(a)(1)(B)
      1. Secretary can grant a permit for a taking, “if such taking is incidental to, and not the purpose of, the carrying out of an otherwise lawful activity”
      2. Condition: Permit applicant must submit a habitat conservation plan (HCP) that will “minimize and mitigate” the impact of the taking authorized by permit. § 10(a)(2)(A)(ii).
        • (i) Example of HCP—NWF v. Norton, p970.
        • (ii) Typically these involve some level of land-swap mitigation—we’ll conserve this and manage it specially. Basically a form of offset. The big issue usually ends up being equivalence. Is 1:1 replacement enough? No, qualitative issues.
        • (iii) It’s easy to put these requirements on the government, in large part because it’s such a large land manager. But the situation is tougher for private landowners, because they don’t have the resources.
        • (iv) Problems:
          • (1) it’s not clear the land would be destroyed or developed otherwise.
          • (2) Oversight—what assurances HCP will be followed?
            1. The agency could require an endowment as insurance.
          • (v) Why allow exception via § 10?
            • (1) If all value were removed from the land à regulatory taking, requiring compensation.
            • (2) But political pressure to repeal the ESA could result if the gov’t began making many regulatory takings.
            • (3) HCP allows private actors to make choices about how best to use their resources w/o imposing a command-and-control scheme.
            • (4) Cost-shifting to private parties to achieve social objectives.
          • Enforcement
            1. Civil and criminal penalties. § 11.
            2. Citizen suits/private right of action.
              1. Against violators. § 11(g)(1)(A)
              2. Against Secretary for failure to perform nodiscretionary duty. § 11(g)(1)(C).
  • Broad conceptual question: why protect animals only when they’re threatened? Why not have broader protections for animals and plants in general?
    1. Concerns about biological diversity come into play most strongly when numbers are dwindling.
    2. Maybe protection on the species level does more to preserve the system?
    3. Human-centered aspect: usefulness of species, biological diversity as protection of agricultural stock, undiscovered use for animals.
  1. Case study: TVA v. Hill, p910:
    • Tellico Dam case. Dam was authorized prior to the passage of the ESA and was almost finished when the enforcement action was brought. Court found that ESA required an injunction of the virtually completed project b/c the Secretary had determined that the project would eradicate an endangered species, despite Congressional appropriations for the project subsequent to passage of the ESA.
      1. Congress had continued to reauthorize funding for the Tellico dam despite its awareness of the ESA issues. What’s the significance of this reauthorization? Does that constitute implied repeal? The court found that it didn’t. Would it have made a difference what type of appropriation it was? Was Congress then authorizing this? Court almost seems to be saying that if they want to do this, they have to “own” it.
      2. Congress subsequently attached a rider to an appropriations bill to exempt this project from ESA protections. Also created the Endangered Species Committee (“God Squad”) and the statutory exemptions.
      3. Was stopping the dam actually a tragic waste of resources? Sunk costs were high. But what were the potential benefits? And the cost of completion? Turns out that this wasn’t justified on CB terms even if you attach no value to the snail darter.
    • How does the statute trade off costs? § 7(h)(1)(A)(ii)
      1. “the benefits of such action clearly outweigh the benefits of alternative courses of action consistent with conserving the species or its critical habitat and such action is in the public interest.” So this appears to let you weigh the benefits of going ahead with the project and the benefits of the conservation action, but not the costs of either. Benefits of building the dam v. benefits of preserving the snail darter.
      2. Do you have to assign an economic value to the snail darter? '
  2. Distinct Population Segment Policy: National Association of Home Builders v. Norton
    • Pygmy owl DPS case.
      1. Under DPS, FWS had designated the AZ population of these owls as different from the Mexico population of the owls. Mexico population was in good shape, but AZ population wasn’t (there’s also an eastern population).
    • Why were NMFS and FWS responsible for developing the DPS policy?
      1. Authority delegated to the Secretary of the Interior and the Secretary of Commerce. § 3(15). Interior gets things on land, Commerce gets things on the ocean.
    • What qualifies a sub-group under DPS?
      1. Must be discrete in relations to the remainder of the species
      2. Must be significant to the species to which it belongs
    • Court found that the population was discrete, but that it was not genetically different, and therefore not significant, to its taxon as a whole. The taxon as a whole includes the Mexican population.
      1. What about the international border issue? p. 6—differences in conservation status…that is significant in light of section 4(a)(1)(D) of the Act.” This is about discreteness, however, not about significance. You still have to be significant.
      2. Basically, if something is abundant across international borders, and therefore not significant, it doesn’t matter if it’s not well conserved there. Once it becomes endangered there, it still might not be a subspecies here, but you won’t need to designate it as a subspecies in order to protect it.
    • Note that if an entire species is endangered in the US, we protect it even if it’s abundant elsewhere.
    • Ecosystem v. species question again—if it’s about ecosystem protection, then we don’t care if they’re abundant elsewhere.
  3. Obligation to conserve, § 7(a)(1)' (p950)'
    • All agencies “shall . . . carry[] out programs for the conservation of endangered species and threatened species.” § 7(a)(1).
      1. “conservation” = “the use of all methods and procedures which are necessary to bring any endangered species or threatened species to the point at which the measures provided pursuant to this chapter are no longer necessary.” Etc. § 3(3).
    • No affirmative obligation: Agency has discretion and is not obligated to choose the least burdensome alternative just b/c it may be more conservative. Pyramid Lake Piutes Tribe v. U.S. Dept. of Navy, (9th Cir. 1990), p950.
      1. No challenge under 7(a)(2) b/c no jeopardy (the additional diversions won’t do much).
    • Notes:
      1. This hasn’t been a very important avenue of litigation. Why has § 7(a)(1) been interpreted to have no teeth?
        1. Judges don’t want to get involved in compelling an agency to take all actions that it possibly can that are consistent with its authorities. They could characterize action as “inaction,” giving agencies much discretion.
      2. Contrafactual, what if diversion will have a significant affect on the fish? Still side with D b/c of discretion (?).
  4. Private takings, § 9(a)(1)(B)
    • Unlawful for any “person” to “take” any endangered species.'
      1. “Take” is defined to mean “harass, harm, pursue, hunt, shoot, wound, kill, trap, capture, or collect, or to attempt to engage in any such conduct.”'
        1. Exception: Secretary may grant a permit for a taking “if such taking is incidental to, and not the purpose of, the carrying out of an otherwise lawful activity.” §10(a)(1)(B).'
      2. Secretary of Interior’s interpretation reasonable that “significant habitat modification or degradation where it actually kills or injures wildlife” = harm and, thus, constitutes a taking. Babbitt v. Sweet Home Chapter of Communities for a Great Oregon, U.S. (1995), p955.
        1. Reasoning [Stevens]:
          • (i) Dictionary definition of “harm” includes indirect acts. Also, interpreting “harm” not to include indirect acts renders the word surplusage—no additional value over the other words in the definition of take.
          • (ii) The broad purpose of the ESA supports agency’s interpretation allowing indirect harms.
          • (iii) Avoid absurdity of § 10 exception—no need for an “incidental” takings exception if indirect harms were not included within the definition of “taking.”
        2. Dissent [Scalia]:
          • (i) Focuses upon “harm” within the list, not as an isolated term. Other words are active words. By noscitur a sociis, “harm” is active, too.
          • (ii) 10 not necessarily absurd—could have been enacted to protect against the taking of endangered animals while taking non-endangered animals.
        3. Notes:
          • (i) How could Congress have addressed Scalia’s concern? Separate “harm” from the other terms in the list that indicate direct acts.
          • (ii) Connection bw § 10 and § 9: 10 is the incidental take provision, allows permits for takings described in § 9 if they are “incidental” to another important action. Has to be incidental to, not the purpose of, an otherwise lawful activity.
        4. Concurrence [O’Connor] (p964):
          • (i) Re: impaired breeding, feeding, and sheltering.
          • (ii) Actual, rather than just speculative harm, should be required.
          • (iii) Proximate cause principles should be used—how close is the connection between the act and the harm?
        5. Marketable Permit Scheme and Endangered Species
          • Advantage: hot spots could work well for species preservation—cf. Nature Conservatancy’s buying of large, contiguous tracts of land.
          • But how would the value of different species be measured? Not necessarily fungible.
          • Also, possibility of manipulation—traded permits might not be needed (b/c land was going to be protected regardless), so no gain but a loss potential for the permit.
  1. The statute was subsequently amended to add “significantly contributed.”
  2. Are the relevant NAAQS in-state NAAQS or does it include out-of-state NAAQS? Subsections (1) and (2) seem to limit it to in-state NAAQS, but this would probably apply only if the EPA realizes it erred in an earlier approval. Thus, subsection (3) establishes the substantive standard of “otherwise comply[ing] with any requirement of this chapter.”