Category: Environmental Law Review Syndicate

Mitigating Greenhouse Gas Emissions in the Northeast and Mid-Atlantic Transportation Sector: A Cap-and-Invest Approach

James D. Flynn*

This post is part of the Environmental Law Review Syndicate


In recent years, states in New England and the mid-Atlantic region have made significant progress in reducing climate change-inducing greenhouse gas (GHG) emissions from the electricity generation sector.[1] Several factors¾including the effects of the economic recession, shifts in energy markets from coal to natural gas and renewable energy sources, and carbon pollution mitigation and clean energy programs like renewable portfolio standards¾have been identified as principal drivers of these reductions.[2] Another is the Regional Greenhouse Gas Initiative (RGGI), a cooperative effort among nine northeastern and mid-Atlantic states to reduce carbon dioxide (CO2) emissions from the power sector.[3] RGGI employs a cap-and-invest approach in which the participating states set a regionally uniform, decreasing cap on CO2 emissions from covered power plants, periodically auction off emission allowances, and invest auction proceeds in other programs including end-use energy efficiency, renewable energy, greenhouse gas abatement, and direct customer electric bill assistance.[4] One study estimates that CO2 emissions in the RGGI region would have been approximately 24 percent higher in 2015 but for the program, which took effect in 2009.[5] At the same time, it is estimated that through 2015, RGGI generated approximately $2.9 million in net economic benefits,[6] and that the investment of RGGI allowance auction proceeds in 2015 alone will return $2.31 billion in lifetime energy bill savings for consumers.[7]

Over approximately the same period of time, however, CO2 emissions from the transportation sector in RGGI states have remained relatively level or have increased. Transportation accounts for 44 percent total CO2 emissions in the region, more than any other sector.[8] Each RGGI member state has adopted a long-term GHG reduction goal, set by statute or executive order, or in climate- or energy-related plans, “generally consistent with achieving an 80 percent reduction of GHG emissions by 2050 from 1990 levels.”[9] Most states’ goals do not include sector-specific emission targets, but because transportation is the largest source of emissions in the region, shifting to a cleaner transportation system is a “critical component of the action needed to meet economy-wide goals and to avoid further catastrophic harms of climate change.”[10] RGGI states already employ a variety of policy mechanisms aimed at decarbonizing transportation,[11] but have been considering whether to employ a cap-and-invest approach similar to RGGI or California’s multi-sector cap-and-invest program, which includes the state’s transportation sector.[12]

This paper first discusses the mechanics of RGGI and California’s cap-and-invest program generally, including how auction proceeds are invested. It then discusses the potential to use a cap-and-invest approach to mitigate GHG emissions from transportation in the Northeast and mid-Atlantic and addresses two key policy considerations: the type of fuels to be covered and the point of regulation. It concludes that, if properly designed, a cap-and-invest approach could achieve significant GHG reductions from transportation in the region and generate substantial funds for other GHG mitigation and climate change adaptation initiatives.

I. The Cap-and-Invest Model

Cap-and-trade programs generally operate as follows.[13] The government sets an overall emissions target¾the cap¾and determines which facilities will be covered. Emission allowances, each generally equal to one ton of emissions, are periodically auctioned or distributed without cost—or both—to covered facilities.[14] The total number of allowances is equivalent to the cap number, which decreases over time.[15] A market is created in which covered facilities may purchase or sell allowances from other covered facilities. Covered facilities are required to hold enough allowances to cover their emissions at the end of a compliance period, which may range from one to three years.[16] If a facility lacks sufficient allowances, it will be assessed a monetary penalty in addition to having to purchase enough allowances to cover the shortfall.

This market-based approach provides covered facilities three options: (1) they may reduce their emissions to meet the number of allowances they purchase or receive; (2) they may purchase additional allowances on the market and emit more; or (3) they may reduce their emissions below the allowances they hold and sell the remainder on the market.[17] The advantage of cap-and-trade programs is that facilities that can reduce their emissions more cost-effectively will do so, while those that face higher emissions reduction costs will purchase additional allowances at auction or on the market.[18] Accordingly, cap-and-trade schemes provide firms with flexibility to design cost-effective, tailored emissions plans, and the regulator achieves its policy objective by means of the overall emissions cap.[19] “Cap-and-invest” refers to cap-and-trade programs that invest their proceeds into other policy initiatives intended to address the pollutant or its effects.


RGGI is the first market-based regulatory program in the United States designed to reduce GHG emissions.[20] It is a cooperative effort among the states of Connecticut, Delaware, Maine, Maryland, Massachusetts, New Hampshire, New York, Rhode Island, and Vermont to cap and reduce CO2 emissions from the electricity generation sector.[21] RGGI is composed of individual CO2 budget trading programs implemented in each participating state. Through independent regulations, each state’s CO2 budget trading program limits emissions of CO2 from electric power plants with the capacity to generate 25 megawatts or more (some 164 facilities), issues CO2 allowances, and establishes participation in regional CO2 allowance auctions.[22]

RGGI began with discussions among the governors of seven New England and mid-Atlantic states, which led to a 2005 Memorandum of Understanding that outlined the program.[23] In 2008, the RGGI states issued a Model Rule that participating states could use as guidance to establish and implement their individual programs.[24] RGGI’s designers expected the initial program might be expanded in the future by covering other emission sources, sectors, GHGs, or states.[25] CO2 emissions from covered facilities in RGGI states account for approximately 20 percent of GHG emissions in the region.[26]

At the end of each three-year compliance period, covered facilities must surrender one allowance for each ton of CO2 emissions generated during the period.[27] Covered facilities are permitted to bank an unlimited number of emission allowances for future use.[28] Over 90 percent of allowances are distributed through periodic auctions, and a reserve price sets a price floor for allowances.[29] RGGI employs a “cost containment reserve” that allows for additional allowances to be auctioned if certain price thresholds are met.[30] In limited circumstances, covered facilities may also submit offsets, which are measurable reductions, avoidances, or sequestrations of emissions from non-covered sources, in lieu of emission allowances.[31] The RGGI states agreed that each would use at least 25 percent of its individual auction proceeds “for a consumer benefit or strategic energy purpose.”[32]

Member states invest the proceeds from allowance auctions in a variety of consumer benefit programs at scale.[33] In October 2017, RGGI, Inc. (the corporate entity that administers RGGI) released a report that tracks the investment of RGGI auction proceeds in 2015 and the benefits of these investments throughout the region.[34] The report estimates that “[t]he lifetime effects of these investments are projected to save 28 million MMBtu of fossil fuel energy and 9 million MWh of electricity, avoiding the release of 5.3 million short tons [4.8 million metric tons] of carbon pollution.”[35] The report also notes that “RGGI investments in 2015 are estimated to return $2.31 billion in lifetime energy bill savings to more than 161,000 households and 6,000 businesses which participated in programs funded by RGGI investments, and to 1.5 million households and over 37,000 businesses which received direct bill assistance.”[36] RGGI states have discretion as to how they invest RGGI proceeds.

The report breaks down these investments into four categories. Energy efficiency makes up 64 percent of investments. Funded programs are expected to return $1.3 billion in lifetime energy bill savings to over 141,000 participating households and 5,700 regional businesses.[37] Clean and renewable energy makes up 16 percent of investments, and investments in these technologies are expected to return $785.8 million in lifetime energy bill savings to 19,600 participating households and 122 regional businesses.[38] Greenhouse gas abatement makes up 4 percent of investments and are expected to avoid the release of 636,000 short tons of CO2.[39] Finally, direct bill assistance makes up 10 percent of investments accounting for $40.4 million in bill credits and assistance to consumers.[40] One independent report notes that while RGGI states each have their own unique auction revenue investment programs, “[o]verall, greater than 60 percent of proceeds are invested to improve end-use energy efficiency and to accelerate the deployment of renewable energy technologies,”[41] which far exceeds the 25 percent investment “for a consumer benefit or strategic energy purpose” required by the Model Rule.

Whether or not RGGI has been successful is the subject of debate. As designed, it applies only to CO2 and only to emissions from some 164 power plants with the capacity to generate twenty-five megawatts or more.[42] Since CO2 accounts for only 20 percent of total GHG emissions in the RGGI states, and electricity generation accounts a fraction of total CO2 emissions, RGGI’s potential is limited.[43] The Congressional Research Service has thus described the initiative’s contribution to global GHG reductions to be “arguably negligible.”[44] In addition, RGGI significantly overestimated emissions from member states for its first compliance period and set an initial emissions cap that was actually above realized emissions levels.[45] This limited participation in the program and allowed participating facilities to bank substantial amounts of unused allowances. After the 2012 program review, RGGI lowered the cap by 45 percent between 2014 and 2020.[46] And after the most recent review in 2016, RGGI lowered the cap by an additional 30 percent between 2020 and 2030.[47] The extent to which these adjustments will hasten emissions reductions to be seen. On the other hand, several studies have shown that the combination of the price signal created by RGGI and the investment of allowance auction proceeds in other environmental programs has been the dominant driver of the recent emissions decline in the region.[48]

b. California’s Cap-and-Invest Program

In 2006, California enacted its landmark climate change law, the Global Warming Solutions Act, also known as AB (“assembly bill”) 32.[49] The statute established an aggressive goal of reducing GHG emissions to 1990 levels by 2020, and an 80 percent reduction from 1990 levels by 2050, across multiple sectors of the state’s economy.[50] AB 32 directed the California Air Resources Board (CARB), the state’s air pollution regulator, to implement a cap-and-trade program, which went into effect in 2013.[51]

According to CARB, the program, which covers approximately 450 entities, “sets a statewide limit on sources responsible for 85 percent of California’s greenhouse gas emissions, and establishes a price signal needed to drive long-term investment in cleaner fuels and more efficient use of energy.”[52] It is “designed to provide covered entities the flexibility to seek out and implement the lowest-cost options to reduce emissions.”[53] The 2013 cap was set at about 2 percent below the emissions level forecast for 2012, declines an additional 2 percent in 2014, and declines 3 percent annually from 2015 to 2020.[54]

Unlike RGGI, California’s program distributes free allocations of emission allowances earlier in the program, but those allocations decrease over time as the program transitions to an auction process.[55] The allocation for most industrial sectors is set at approximately 90 percent of average emissions and is updated annually based on each facility’s production.[56] Electrical distribution and natural gas facilities receive free allowances on the condition that the value of allowances must be used to benefit ratepayers and achieve GHG emission reductions.[57] The allocation for electrical distribution utilities is set at about 90 percent of average emissions, and for natural gas utilities, is based on natural gas supplied in 2011 to non-covered entities.[58] The program includes cost containment measures and allows for the banking of allowances, has a three-year compliance period with an annual obligation to surrender 30 percent of their previous year’s emissions, and allows for offsets of up to 8 percent of a facility’s compliance obligation.[59] AB 32 also employs a substantial penalty mechanism for facilities that fail to meet their compliance obligations: “If the compliance deadline is missed or there is a shortfall, four allowances must be provided for every ton of emissions that was not covered in time.”[60]

California’s cap-and-trade program became linked with Québec’s cap-and-trade system on January 1, 2014 and became linked with Ontario’s cap-and-trade program on January 1, 2018.[61] All allowances issued by the California, Québec, and Ontario programs before and after the linkage can be used for compliance interchangeably across jurisdictions.[62] The three jurisdictions also hold joint allowance auctions.

On January 1, 2015, suppliers of transportation fuels, including gasoline and diesel fuel, became covered under the program.[63] A fuel supplier is defined as “a supplier of petroleum products, a supplier of biomass-derived transportation fuels, a supplier of natural gas including operators of interstate and intrastate pipelines, a supplier of liquefied natural gas, or a supplier of liquefied petroleum gas.”[64] All fuel suppliers that deliver or import 10,000 metric tons or more of annual CO2 equivalent emissions are subject to a reporting requirement, but only suppliers that reach a 25,000 metric ton threshold are covered by the cap-and-trade program.[65]

Proceeds from the allowance auctions are deposited in the state’s Greenhouse Gas Reduction Fund and are appropriated by the state legislature for “investing in projects that reduce carbon pollution in California, including investments to benefit disadvantaged communities, recycling, and sustainable transit.”[66] As of 2017, some $3.4 billion had been appropriated to state agencies implementing GHG emission reduction programs and projects, collectively referred to as the California Climate Investments.[67] Of that amount, $1.2 billion has been expended on projects “expected to reduce GHG emissions by over 15 million metric tons of carbon dioxide equivalent.”[68]

II. Applying a Cap-and-Invest Approach to Northeast and Mid-Atlantic Transportation Sector

Under business-as-usual trends, carbon emissions in RGGI states will be 23 percent below the 1990 baseline in 2030.[69] These states must achieve much deeper emissions reductions across multiple economic sectors in order to achieve their “greenhouse gas emission reduction targets for 2030 that range from 35 to 45 percent, centered around a 40 percent reduction from 1990 levels.”[70] Since transportation represents the largest share of GHG emissions in the RGGI states, that sector should be a primary focus of policymakers’ attention.

One study finds that the levels of emissions reductions necessary to meet the GHG reduction goals of the states in the region could be accomplished “through a suite of clean transportation policies” including financial incentives for the purchase of clean vehicles, such as electric and hybrid light-duty vehicles and natural gas powered heavy-duty vehicles; investments in public transit expansion including bus rapid transit, light rail, and heavy rail; promotion of compact land use; investment in bicycle infrastructure; support for travel demand management strategies; investment in system operations efficiency technologies; and investment in infrastructure to support rail and short-sea freight shipping.[71]

One potential mechanism for achieving the levels of reductions necessary for the RGGI states to meet their targets “would be to implement a transportation pricing policy, which could both achieve GHG reductions and generate proceeds that could be used to fund clean and resilient transportation solutions.”[72] For example, “carbon-content-based fees, mileage-based user fees, and motor-fuel taxes” could “generate an average of $1.5 billion to $6 billion annually in the region.”[73] A mid-range pricing policy that generated approximately $3 billion annually “would create a price signal that would promote alternatives to single-occupancy vehicle travel and result in modest additional emission reductions. It would also raise a cumulative $41 billion to $46 billion for the region during 2015-2030.”[74] Proceeds from such a pricing policy would offset projected declines from existing state and federal gasoline taxes and could be used to fund other clean transportation initiatives.[75]

A hypothetical regional cap-and-invest program for vehicle emissions might be structured as follows. Member states would establish a mandatory regional cap on GHG emissions from the combustion of fossil transportation fuels calculated using volumetric fuel data and fuel emission factors available from the Environmental Protection Agency.[76] The cap would decline over time. States would auction allowances equal to the cap and establish an entity like RGGI, Inc. to administer the program, auction platform, and allowance market.[77] Regulated entities would achieve compliance by purchasing allowances at auction or from other market participants, and possibly with offsets earned from reductions in other aspects of their operations.[78] As with RGGI, individual member states would commit to invest a percentage of their auction proceeds into other initiatives aimed at reducing GHG emissions, including from transportation, and could retain the discretion to decide individually how to allocate those funds.[79]

Because power plants are stationary and relatively few in number, their GHG emissions can be regulated directly, i.e., at the stack. Vehicles, however, are mobile and far more numerous. To regulate the emissions from every fossil fuel powered vehicle at the tailpipe would entail a substantial and possibly prohibitive administrative burden, and would likely be politically unpalatable. An alternative is to use transportation fuel as the point of regulation. Determining which types of fuels and which entities in the fuel supply chain to cover under the cap-and-invest program will be critical.

Transportation fuels that could be covered include gasoline, on-road and off-road diesels, aviation fuels, natural gas, propane/butane, and marine fuels.[80] Considering both the volume of each type of fuel consumed and the comparative emissions resulting from its consumption, the program should cover, at a minimum, gasoline and on-road diesel, which account for approximately 85 percent of carbon emissions from transportation in the region.[81] Other fuels may make up too small a portion of total emissions to justify the additional technical and regulatory burden of covering them.[82] In addition, because all states in the region currently require reporting on gasoline and on-road diesel, the most straightforward approach would be to regulate those fuels. Covering other fuels would require at least some states that do not already require reporting of these fuels to establish new reporting requirements.[83]

Another key design choice is the point of regulation: which entities within the transportation fuel supply chain should be subject to the regulatory obligation to hold sufficient allowances. Because all states in the region have existing reporting and enforcement mechanisms for gasoline and on-road diesel (and many also tax off-road diesel and aviation fuel), one option would be to regulate existing state points of taxation for these fuels.[84] However, state points of taxation are not uniform throughout the region. They can include many different types of entities in the supply chain and in some states the point of taxation is different for different fuels.[85] State regulations also differ with respect to what actions by covered entities trigger the reporting requirement.[86] Many states have points of regulation low in the supply chain, such as entities that purchase fuel from the terminal rack and distribute it to retailers.[87] Thus, while using existing state points of taxation to regulate transportation fuels would make use of existing state regulatory mechanisms, it would also require regulating over one thousand entities across the region, many of which are smaller distributors.[88]

Another possible point of regulation would be one that is as far upstream as possible, i.e., entities that refine fuel in the region for use in the region, and those that import fuel into the region for use in the region.[89] This would include refineries, and for fuels refined outside the region, the first importers into the region.[90] Eight refineries in the region and an unknown number of first importers, including foreign suppliers and suppliers from U.S. states outside the region, would be subject to regulation.[91] This option would require reporting of the destination of all fuel produced in or that enters the region to ensure that a fuel to be used outside the region is not inadvertently covered.[92] While the Energy Information Administration (EIA) and the Environmental Protection Agency generally require destination data from refiners and importers into the U.S. and from interstate suppliers, the agencies do not publicly disclose this data.[93] Thus, regulating refiners and importers would likely cover many fewer entities as compared to existing state points of taxation, most of which would be large petroleum companies.[94] However, because only three states in the region have refineries within their borders, and because importers are not systematically tracked throughout the region, accounting for fuels that are transported through states to prevent double-counting would likely require the establishment of new regional reporting requirements that would include points of origin and destination.[95]

A third possible point of regulation would be entities known as prime suppliers, defined by the EIA as “suppliers who produce, import, or transport product across state boundaries and local marketing areas and sell to local distributors, local retailers, or end-users.”[96] For the region, this includes approximately 30 refiners, other producers of finished fuel, interstate resellers and retailers, and importers.[97] EIA requires these entities to report the amount of fuel, including gasoline, diesel, and aviation fuel, sold or transferred for end use by state on a monthly basis.[98] Although EIA does not publicly provide disaggregated prime supplier data because of statutory privacy restrictions, organizations may enter into data-sharing arrangements with EIA to obtain individual prime supplier data.[99] Thus, while the prime supplier group would include a larger number of regulated entities than importers and refiners, it would provide a consistent definition of a point of regulation already understood by the regulated entities.[100] Regulating prime suppliers, most of which are higher in the supply chain than existing state points of taxation, would also relieve most smaller entities of compliance obligations.[101]


States in states in New England and the mid-Atlantic region must make much deeper emissions reductions in the transportation sector in order to meet their overall GHG emission reduction targets. Recognizing this reality, representatives from Connecticut, Delaware, Maryland, Massachusetts, New York, Rhode Island, Vermont, and Washington, D.C., at the 2017 Conference of the Parties to the United Nations Framework Convention on Climate Change, signed a joint statement affirming their commitment to reducing GHG emissions from the transportation sector. In that statement, they identified “market-based carbon mitigation strategies” as potential pathways to achieving needed emissions reductions.[102]

Despite its early struggles, the cap-and-invest approach to mitigating emissions in the northeast and mid-Atlantic electricity generation sector has achieved, at a minimum, some emissions reductions, substantial investment in other GHG mitigation efforts, and overall net benefits within the region. California has achieved substantial GHG emissions reductions across multiple sectors, including transportation, and has invested substantial sums in a suite of other green programs. These examples demonstrate the potential of using a cap-and-invest approach to accomplish environmentally and economically sound policy objectives, both within the RGGI region and in the context of transportation. If properly structured, such an approach could achieve significant emissions reductions in the region and raise substantial funds for other GHG mitigation and climate change adaptation initiatives.

How would a cap-and-invest approach to transportation emissions be structured? The fundamental aspects of RGGI and California’s cap-and-invest program are similar in most respects. California occupies a unique position in federal regulation of automobile emissions and had the benefit of constructing a program applicable only to itself, although its program is now linked with programs in other jurisdictions. RGGI already covers much of the Northeast and mid-Atlantic region, could be expanded to include other sectors of those states’ economies, including transportation, and could be linked with the California-Québec-Ontario cap-and-invest system to create a larger and more efficient allowance market.

Owing to the practical differences between directly regulating emissions from power plants and indirectly regulating transportation emissions by fuel type and supply chain point, the mechanics of using a cap-and-invest approach to mitigate transportation emissions, especially across jurisdictions, poses some potentially challenging design issues. The program should cover, at a minimum, gasoline and on-road diesel. Identifying the appropriate point of regulation will require policymakers to consider a host of technical, administrative, and policy issues. Existing state points of taxation are numerous and vary by jurisdiction and by fuel type within jurisdictions. Upstream refiners and importers are far fewer in number but regulating these entities would likely require the development of new regional reporting mechanisms that might make this option administratively undesirable. While the prime suppliers group is larger in number than refiners and importers, regulating prime suppliers would provide a consistent state-based definition of a point of regulation already understood by the regulated entities, and would not subject most smaller entities to compliance obligations.

* James Flynn is an LL.M. candidate at New York University School of Law and the graduate editor of the NYU Environmental Law Journal.

[1] See Energy Information Administration, State Carbon Dioxide Emissions Data (last visited Feb. 10, 2018),

[2] See Gabe Pacyniak, et al., Reducing Greenhouse Gas Emissions from Transportation: Opportunities in the Northeast and Mid-Atlantic, Georgetown Climate Center 8 (2015),

[3] Regional Greenhouse Gas Initiative, Welcome (last visited Feb. 10, 2018),

[4] Regional Greenhouse Gas Initiative, RGGI Benefits (last visited Feb. 10, 2018),

[5] Brian C. Murray and Peter T. Maniloff, Why have greenhouse emissions in RGGI states declined? An econometric attribution to economic, energy market, and policy factors, Energy Economics 51, 588 (2015).

[6] See Paul J. Hibbard, et al., The Economic Impacts of the Regional Greenhouse Gas Initiative on Nine Northeast and Mid-Atlantic States, Analysis Group 5 (July 14, 2015),; Ceres, The Regional Greenhouse Gas Initiative: A Fact Sheet (2015),

[7] Regional Greenhouse Gas Initiative, The Investment of RGGI Proceeds in 2015 3 (Oct. 2017),

[8] See Energy Information Administration, supra note 1; Gerald B. Silverman and Adrianne Appel, Northeast States Hit the Brakes on Carbon Emissions From Cars, BNA (Oct. 16, 2017),

[9] Pacyniak, supra note 2.

[10] Id.

[11] See Gabe Pacyniak, et al., Reducing Greenhouse Gas Emissions from Transportation: Opportunities in the Northeast and Mid-Atlantic, Appendix 3: State GHG Reduction Goals in the TCI Region, Georgetown Climate Center 4-13 (2015),

[12] See, e.g., Center for Climate and Energy Solutions, California Cap and Trade (last visited Feb. 10, 2018),

[13] See Joel B. Eisen, et al., Energy, Economics and the Environment 326 (4th ed. 2015).

[14] Id.

[15] Id.

[16] See id.

[17] Id.

[18] See id.

[19] See id.

[20] See Regional Greenhouse Gas Initiative, supra note 3.

[21] See id.

[22] Regional Greenhouse Gas Initiative, Program Design (last visited Feb. 10, 2018),

[23] Regional Greenhouse Gas Initiative, A Brief History of RGGI (last visited Feb. 10, 2018),

[24] Id.

[25] Jonathan L. Ramseur, The Regional Greenhouse Gas Initiative: Lessons Learned and Issues for Congress,

Congressional Research Service 3 (May 16, 2017),

[26] Id.

[27] Id.

[28] Id.

[29] Id.

[30] Id.

[31] Id. at 4.

[32] Id. at 3.

[33] See Brian M. Jones, Christopher Van Atten, and Kaley Bangston, A Pioneering Approach to Carbon Markets: How the Northeast States Redefined Cap and Trade for the Benefit of Consumers, M.J. Bradley & Associates 4 (Feb. 2017),

[34] Regional Greenhouse Gas Initiative, The Investment of RGGI Proceeds in 2015 (Oct. 2017),

[35] Id. at 3.

[36] Id.

[37] Id.

[38] Id.

[39] Id.

[40] Id.

[41] Jones, supra note 33.

[42] Id.

[43] See id. at 3.

[44] Id. at 17.

[45] Id. at 4.

[46] Regional Greenhouse Gas Initiative, Elements of RGGI (last visited Feb. 10, 2018),

[47] Regional Greenhouse Gas Initiative, Summary of RGGI Model Rule Updates 1 (Dec. 19, 2017),

[48] See Murray, supra note 5 at 25-26; Man-Keun Kim and Taehoo Kim, Estimating impact of regional greenhouse gas initiative on coal to gas switching using synthetic control methods, Energy Economics 59, 334 (2016).

[49] California Air Resources Board, Assembly Bill 32 Overview (last visited Feb. 10, 2018),

[50] Id.

[51] Id.

[52] Id.

[53] California Air Resources Board, Overview of ARB Emissions Trading Program 1 (last visited Feb. 10, 2018),

[54] Id.

[55] Id.

[56] Id.

[57] Id.

[58] Id.

[59] Id. at 2.

[60] Id.

[61] California Air Resources Board, Facts About The Linked Cap-and-Trade Programs 1 (updated Dec. 1, 2017),

[62] Id.

[63] California Air Resources Board, Information for Entities That Take Delivery of Fuel for Fuels Phased into the Cap- and-Trade Program Beginning on January 1, 2015 1 (last visited Feb. 10, 2018),

[64] Id. at 2.

[65] Id.

[66] California Air Resources Board, 2017 Report to the Legislature on California Climate Investments Using Cap-And-Trade Auction Proceeds i (2017),

[67] Id.

[68] Id. at v.

[69] Elizabeth A. Stanton, et al., The RGGI Opportunity, Synapse Energy Economics, Inc. 3 (revised Feb. 5, 2016), Notably, this study took into account the anticipated effect of the Clean Power Plan, which President Donald Trump and Environmental Protection Agency Administrator Scott Pruitt propose to repeal. See id. at 4.

[70] Id. at 2.

[71] Pacyniak, supra note 2 at 22. The Georgetown Climate Center serves as the facilitator for the Transportation Climate Initiative, which is “a collaboration of the agency heads of the transportation, energy, and environment agencies of 11 states and the District of Columbia, who in 2010 committed to work together to improve efficiency and reduce greenhouse gas emissions from the transportation sector throughout the northeast and mid-Atlantic region.” Id. at i.

[72] Id. at 25.

[73] Id.

[74] Id.

[75] Id. at 26-27.

[76] Drew Veysey, Gabe Pacyniak, and James Bradbury, Reducing Transportation Emissions in the Northeast and Mid-Atlantic: Fuel System Considerations, Georgetown Climate Center 7 (Nov. 13, 2017),

[77] Id.

[78] Id.

[79] See id.

[80] Id. at 9.

[81] See id. at 11-13.

[82] See id. at 33.

[83] Id. at 20.

[84] Id.

[85] Id. at 16.

[86] Id.

[87] Id. at 17.

[88] Id. at 33.

[89] Id. at 21.

[90] Id.

[91] Id.

[92] Id. at 22.

[93] Id.

[94] Id. at 33.

[95] Id.

[96] Id. at 24.

[97] Id.

[98] Id.

[99] Id. at 25.

[100] Id. at 33

[101] Id.

[102] See Transportation and Climate Initiative, Northeast and Mid-Atlantic States Seek Public Input As They Move Toward a Cleaner Transportation Future (Nov. 13, 2017),; Sierra Club, Northeast and Mid-Atlantic Governors Lauded for Announcement on Transportation and Climate, Press Release (Nov. 13, 2017),

Reinstating CERCLA as the “Polluter Pays” Statute with the Circuit Court’s Mutually Exclusive Approach

By Brianna E. Tibett[i]

This post is part of the Environmental Law Review Syndicate


The purpose of the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA) is to facilitate the “timely cleanup of hazardous waste sites and to ensure that the [cleanup costs are] borne by those responsible for the contamination.”[ii] The proper application of CERCLA’s two private causes of action is necessary to achieve these goals. When applied properly they encourage private parties to voluntarily cleanup hazardous waste sites, effectively spread the cost of cleanup to the responsible parties, and encourage settlement.

For example, when a private potentially responsible party (PRP) voluntarily cleans up a site before any action regarding the site is commenced the PRP eliminates their exposure to uncertain liability, and avails itself of the “arguably preferred recovery vehicle for a PRP,” the cost recovery action. The private cost recovery action, under § 107(a)(4)(B), allows private parties to seek to recover the costs they incurred in voluntarily cleaning up a contaminated site from PRPs (regardless of their contribution to the site’s contamination).[iii] The PRP subject to the § 107(a)(4)(B) cost recovery action, can counterclaim in or bring against multiple other PRPs a § 113(f)(1) contribution action, requiring the equitable apportionment of the response costs.[iv] The remedy and the shorter statute of limitations afforded by contribution actions incentivizes PRPs to immediately locate other PRPs and initiate lawsuits sooner.[v]

The Supreme Court’s framework for the application of these private causes of action created in Atlantic Research[vi] jeopardizes CERCLA’s mechanisms that encourage PRPs to settle with the EPA. The Court’s framework identifies the cause of action that applies exclusively in some circumstances but not all. Specifically, the framework leaves open the availability of both causes of action in situations in which costs are directly incurred as a result of forced cleanup. Uncertainty around the cause of action that a court will allow in circumstances of compelled cleanup may cause PRPs to stray away from settling with the Environmental Protection Agency (EPA), and thus make it more difficult for the EPA to negotiate cleanup and reimbursement settlements.[vii] Or it could incentivize PRPs to attempt to pass their tab on to another PRP by settling [to cleanup] and then bringing a cost recovery action to recover those cleanup costs. Which if permitted would leave the defendant unable to counter-sue for contribution, because of the plaintiff-PRP’s contribution bar, defeating CERCLA’s goal to have the responsible parties pay for cleanup.[viii]

The United States’ Courts of Appeals, have advanced a mutually exclusive framework that fully clarifies the applicability of and the interplay between the private causes of action. This Article supports the mutually exclusive approach. First, the Article provides a brief overview of the history and development of the private causes of action. Second, the Article highlights the issues regarding the applicability of the private causes of action left unresolved by the Court. Third, the Article demonstrates how the mutually exclusive framework, established by the U.S. courts of appeals, seamlessly resolves those issues and facilitates the advancement of CERCLA’s goals.

I. History and Development of CERCLA’s Private Causes of Action

A. CERCLA’s Enactment

Congress’s prime motivation for passing CERCLA was to provide the EPA with the ability to promptly respond to the country’s hazardous waste sites and to place the cost of the response on the responsible parties, the “polluters.”[i] To that end, Congress furnished the EPA with the means to undertake cleanup itself,[ii] sue PRPs for reimbursement,[iii] and the authority to compel PRPs to clean up contaminated sites.[iv] However, Congress recognized that the EPA would not be equipped on its own to address 30,000 to 50,000 improperly managed hazardous waste sites.[v] CERCLA would also have to induce private parties to perform cleanup.[vi] Accordingly, Congress included § 107(a)(4)(B), to enable private parties to recover their costs of cleanup from PRPs.[vii]

Because of CERCLA’s liability scheme, the remedies available to PRPs were in dispute.[viii] Under § 107(a)(4)(A) the courts have interpreted CERCLA’s liability to apply retroactively, strictly, jointly, and severally.[ix] Additionally, CERCLA liability extends beyond polluters to also include those who would benefit from cleaned sites, such as current owners and operators.[x] Thus, a current owner of contaminated property who did not contribute to the release of hazardous waste, or a past owner who only contributed a small part of the waste, may be a PRP.[xi] PRPs may find themselves subject to a cost recovery action, and if so, ultimately liable for the entire cost of cleanup.

To mitigate these harsh results, some courts held either that § 107(a)(4)(B) or federal common law provided litigants subject to a § 107 cost recovery claim an implied right to contribution.[xii] This allowed PRPs to either counterclaim for contribution or sue other PRPs for contribution.[xiii] A successful contribution action permits the equitable apportionment of costs among PRPs,[xiv] ameliorating the harsh effects of joint and several liability. As a result, more PRPs were required to pay their proportionate share of the cleanup instead of leaving a single PRP liable. Despite these efforts, extensive litigation continued, necessitating a CERCLA amendment.[xv]

B. SARA’s Contribution Action and Contribution Protection

Congress passed the Superfund Amendments and Reauthorization Act (SARA) in 1986 to address: (1) the EPA’s inability to timely recover response costs; (2) the threat that the courts would erode joint and several liability into a “fair share” allocation; and (3) the effectiveness of contribution actions in spreading the cost of cleanup to responsible parties.[xvi] SARA created an express cause of action for contribution and incorporated statutes of limitations.[xvii] The right to contribution, codified in § 113(f)(1),[xviii] allows a PRP, “during or following” a § 106 (compelled clean-up) or § 107 civil action, to seek contribution payments from another PRP that has not resolved its liability.[xix] The “settlement bar” created by SARA in § 113(f)(2), provides parties who have reached an “administrative or judicially approved settlement” with “contribution protection”—immunity from contribution claims that concern matters within the agreement.[xx]

The new provisions, although preserving contribution, did not fully resolve existing issues and indeed generated new ones. For example, SARA did not answer whether an implied right to contribution still remains when contribution pursuant to § 113(f) is unavailable—i.e., whether PRPs may pursue contribution only through § 113(f).[xxi] Many United States Courts of Appeals, while attempting to navigate § 107(a) and § 113(f) claims, have held that a claim for contribution under § 113(f) was the exclusive remedy for PRPs.[xxii] By preventing PRPs from pursuing an action under § 107(a), § 113(f) served as PRPs’ sole avenue to seek contribution.[xxiii] Still, some courts expanded § 113(f)’s provisions to allow recovery actions even in the absence of a suit under § 106 or § 107.[xxiv]

C. The Supreme Court’s Cooper Industries Decision

In Cooper Industries, Inc. v. Aviall Services, Inc. the Court addressed the expanded application of § 113(f)(1) and ultimately limited its availability to PRPs “during or following” a § 106 or § 107 civil action.[xxv] The Court held that current property owners who voluntarily cleaned up the contaminated site could not maintain a contribution action under § 113(f)(1) because the claim did not arise out of a § 106 or § 107 civil action.[xxvi] First, the Court held that “may” in § 113(f)(1) should not be read as permissive; it should be read to only authorize § 113(f)(1) contribution claims “during or following” § 106 or § 107 civil actions.[xxvii] The Court stated that reading “may” to allow a PRP to bring a “contribution action at any time, regardless of the existence of a . . . civil action,” would render the language “during or following” superfluous, along with § 113(f)(3)(B), which permits contribution actions after settlement.[xxviii] Second, the Court found that § 113(f)(1)’s saving clause, does not change its reading.[xxix] The Court specified that the saving clause functions only to prevent the loss of “any cause(s) of action for contribution that may exist independently of § 113(f)(1).”[xxx] Therefore, it does not expand the scope of § 113(f)(1) or create a cause of action, it only “rebuts any presumption that the express right of contribution provided by . . . [§ 113(f)(1)] is the exclusive cause of action for contribution available to a PRP.”[xxxi]

Following this application of § 113(f)(1), several Courts of Appeals reconsidered whether PRPs have any right of action under § 107(a)(4)(B).[xxxii] After revisiting this issue, some courts permitted private cost recovery actions under § 107(a)(4)(B).[xxxiii] However, the Third Circuit continued to hold § 113(f) as the exclusive cause of action available for PRPs.[xxxiv] Accordingly, the Third Circuit in E.I. DuPont De Nemours and Co. v. U.S. held that there was no cause of action for PRPs who engaged in “sua sponte voluntary cleanups,”[xxxv] effectively disincentivizing voluntary cleanup.

D. Supreme Court’s Atlantic Research Decision

The Court again revisited the scope of CERCLA’s private causes of action. In United States v. Atlantic Research Corp., the Court: (1) held that PRPs have a right to cost recovery under § 107(a)(4)(B);[xxxvi] (2) clarified that §§ 107(a) and 113(f) provide distinct remedies; and (3) provided a framework for the application of §§ 107(a)(4)(B) and 113(f)(1) actions.[xxxvii] The Court made the inference that Congress sculpted § 113(f)(1) based on the traditional sense of contribution, which is contingent “upon an inequitable distribution of common liability among liable parties.”[xxxviii] However, because the statute authorizes PRPs to seek contribution “during or following” a civil action, liability does not need to be established before bringing a contribution action under § 113(f)(1).[xxxix]

The Court held that PRPs may utilize a cost recovery action, pursuant to § 107(a)(4)(B), only to recover costs the PRP “‘incurred’ in cleaning up a site.”[xl] For instance, when a PRP reimburses another party, the PRP has not incurred its own cleanup costs and thus cannot recover them under § 107(a)(4)(B).[xli] Additionally, the Court held that § 107(a)(4)(B) is the sole cause of action to recover costs incurred during voluntary cleanup.[xlii] With these distinctions made, the Court states that the remedies available in §§ 107(a) and 113(f) “provid[e] causes of action ‘to persons in different procedural circumstances,’” and as a result they do not cause conflict, or provide an opportunity for a PRP to choose its remedy.[xliii]

To summarize, § 113(f) authorizes a right to contribution “to PRPs with common liability stemming from an action instituted under § 106 or § 107(a).”[xliv] Respectively, after a PRP pays money pursuant to a settlement agreement or a court judgment, in which they are reimbursing those parties they may, and may only, pursue § 113(f) for contribution.[xlv] On the contrary, “§ 107(a) permits cost recovery . . . by a private party that has itself incurred cleanup costs.”[xlvi] As a result, in cases of reimbursement, a PRP cannot circumvent § 113(f)(1)’s three-year statute of limitations by attempting to bring an action in cost recovery, which has a six-year limitation.[xlvii]

Lastly, the Court claims PRPs that utilize § 107(a) “will not eviscerate the settlement bar set forth in § 113(f)(2).”[xlviii] The settlement bar provision “prohibits § 113(f) contribution claims against ‘[a] person who has resolved its liability to the United States or a State in an administrative or judicially approved settlement.’”[xlix] The Court explains that although the contribution bar “does not by its terms protect against cost-recovery liability,” the defendant can trigger equitable apportionment by filing a § 113(f) counterclaim.[l] In footnote 6, the Court states that in cases of reimbursement and voluntary cleanup, §§ 107(a)(4)(B) and 113(f) have no overlap, but there may be overlap when a PRP incurs expenses pursuant a consent decree.[li] In cases of “compelled costs” a PRP does not incur costs voluntarily (which would have the effect of precluding a § 113(f)(1) contribution action) but also does not reimburse the costs of another party (which would have the effect of precluding § 107(a)(4)(B) cost recovery action).[lii] The Court did not address whether these compelled costs of response are recoverable under § 107(a) or § 113(f).

II. Issues Left Unresolved by the Supreme Court 

Although the framework provided by the Court’s Cooper Industries and Atlantic Research decisions reinstate PRP’s ability to utilize § 107(a)(4)(B) for cost recovery and § 113(f) for contribution, the decisions do not clarify the complete applicability and interplay of the private causes of action. The Court’s framework for CERCLA’s private actions is limited to the following: (1) PRPs who pay money to satisfy a settlement agreement or a court judgment—incur costs in the form of reimbursement—may only pursue § 113(f) contribution actions; and (2) PRPs who have incurred cleanup costs directly—not reimbursement costs—may only seek to recover those response costs from other PRPs pursuant to § 107(a)(4)(B). Thus, in those limited “procedural circumstances,” §§ 107(a)(4)(B) and 113(f) are mutually exclusive.[i]

This limited framework leaves unresolved the cause of action or actions available to private parties in other situations, specifically when PRPs incur costs directly.[ii] The following issues, which were unresolved by the Court’s framework, have not only spurred considerable litigation, but have also caused apprehension to settling claims:

  1. Whether settling-PRPs may sue other PRPs for cost recovery pursuant 107(a)(4)(B) to recover cleanup costs that were incurred voluntarily, i.e., costs incurred independent of the administrative or judicially approved settlement.
  2. What causes of action do settling-PRPs that incur costs directly in order to comply with settlement obligations have when such settlement does not satisfy the requirements set forth in 113(f)(3)(B).
  3. What cause of action is available to PRPs who directly incur cleanup costs under an obligation in an “administrative or judicially approved settlement?” May they bring an action in cost recovery pursuant § 107(a)(4)(B) and as a result: (1) circumvent the contribution bar which prevents them from bringing an action in contribution against the other PRPs in their settlement agreement; and (2) render both non-settling PRPs and settling-PRPs unable to counterclaim in contribution because the plaintiff-PRP can utilize the contribution bar.
  4. Whether a PRP that settled with a state entity has a cause of action under CERCLA.
  5. Whether a private entity that finances a cleanup pursuant to a private agreement has a cause of action under CERCLA to recover costs.
  6. If the statute of limitations for a PRP’s contribution claim runs out—and a PRP can no longer pursue its right to contribution—may the PRP pursue cost recovery pursuant § 107(a)(4)(B) when it had incurred cleanup costs as a result of its obligations flowing from an “administrative or judicially approved settlement.”

Strictly adhering to the Court’s framework to resolve these issues would permit either a § 107(a)(4)(B) or § 113(f) action in all the above circumstances. Allowing settling-PRP’s to choose which cause of action they can utilize could cause any PRP, regardless their responsibility of contamination, to be stuck with the entire or a significant portion of the cleanup costs while other PRPs skirt liability.[iii] For example, under the Court’s framework, settling-PRPs could pursue cost recovery actions under § 107(a)(4)(B) for costs incurred directly from cleanup required in order to satisfy the “administrative or judicially approved settlement.” As a result, defendant PRPs subject to § 107(a)(4)(B) causes of action brought by a settling-PRP, can be subject to joint and several liability without the ability to counterclaim for contribution pursuant § 113(f)(1) because of the plaintiff-PRP’s contribution bar under § 113(f)(2). This application would not advance Congress’s intent of CERCLA being a “polluter pays” statute, where the responsible parties bear the financial responsibility of the cleanup. To the contrary, under this framework CERCLA functions more like a game of Uno.[iv]

III. The Mutually Exclusive Approach Adopted by the U.S. Courts of Appeals 

Litigation over the unresolved issues has ensued in the lower federal courts since the Court’s holding in Atlantic Research.[1] The United States’ Courts of Appeals that have heard the issues, collectively hold that the causes of action available to private parties apply mutually exclusively.[2] This framework provides a seamless application of the private causes of action in all circumstances, including those that were left unresolved by the Court.

When a private party incurs costs directly, the mutually exclusive approach resolves the issue of what proper cause of action the PRP is authorized to utilize. The lower courts agree that once it is determined that either a § 113(f)(1) or § 113(f)(3)(B) contribution action is available for the costs sought, the PRP must pursue an action for contribution, and is barred from pursuing a § 107(a)(4)(B) cost recovery action.[3] If, however, contribution is not available to recover the costs sought, the private party may pursue a § 107(a)(4)(B) cost recovery action to recover its response costs.[4] The mutual exclusive approach provides a framework for determining the causes of action for each of the unresolved issues mentioned above, while simultaneously advancing CERCLA’s goals.

A. Availability of §§ 113(f)(1) and 113(f)(3)(B) Contribution Actions Under the Mutually Exclusive Approach

All contribution claims under § 113(f) are contingent upon “an inequitable distribution of common liability among” PRPs at the time the underlying claim is resolved.[5] Following the Court’s rulings in Cooper Industries and Atlantic Research, PRPs subject to a civil action under either § 106 or § 107 may only seek contribution. The unresolved issues following the Court decisions thus lie within the application of § 113(f)(3)(B).

Section 113(f)(3)(B) provides that contribution claims are available to entities who have resolved their “liability to the United States or a State for some or all of a response action or for some or all of the costs of such action in an administrative or judicially approved settlement[.]” Following the language of § 113(f)(3)(B), the agreements that trigger contribution claims must be “administrative or judicially approved settlement[s].”[6] A judicially approved settlement can take the form of a consent decree, which results from a court’s approval of a settlement that is “fair, reasonable, and consistent with CERCLA’s goals.”[7] The “defining feature of an ‘administrative settlement’ is” the resolution of a “PRP’s liability to the United States . . . for some or all of a response action or for some or all of the costs of such action.”[8]

For an administrative settlement to trigger the application of § 113(f)(3)(B), the federal government must have followed the procedures set forth in § 122(i).[9] Although § 112(i) procedural requirements apply only to the federal government, several courts have held that in light of due process concerns, CERCLA administrative settlements entered into with a state entity require hearings or public comments, as required for federal entities in § 112(i).[10] Thus, a state administrative settlement should provide non-settling parties with notice and an opportunity to be heard.[11] If procedural safeguards similar to those set forth in § 122(i) are not followed, a settlement cannot constitute an “administrative settlement” that triggers § 113(f)(3)(B).[12] Those PRPs will neither have an action in contribution, nor will they be afforded contribution protection.[13]

Congress provided contribution protection to those parties entering into settlements to further incentivize settling, as well as to support the “polluters pay” philosophy.[14] Section 113(f)(2) bars contribution claims against entities that have resolved their liability to the United States or a state in an “administrative or judicially approved settlement” if the costs arise from matters addressed in the settlement.[15] The party claiming contribution protection, whether defendant or plaintiff, must demonstrate that it is afforded such protection. Contribution protection will not be afforded to parties that cannot demonstrate the resolution of their CERCLA liability.[16] In other words, PRPs must demonstrate that they have been subject to “an administrative or judicially approved settlement.”[17]

For example, the Pennsylvania Middle District Court held that the agreement between the Pennsylvania Department of Environmental Protection and United States did not constitute an administrative settlement because it was devoid of any procedures designed to safeguard due process concerns.[18] As a result, the court permitted the plaintiff to pursue a contribution claim against the federal government because the federal government was not afforded contribution protection[19].

When adhering to the mutually exclusive approach, if the requirements to satisfy an “administrative or judicially approved settlement” are not met by the agreement that causes the PRP to incur cleanup costs directly, that party may pursue a cost recovery action to recover those costs, because it does not have an action for contribution. On the other hand, “a party who may bring a contribution action for certain expenses must use the contribution action, even if a cost recovery action would otherwise be available.”[20]

Parties cannot circumvent the mutually exclusive approach by waiting for their contribution action to run so they can employ an action for cost recovery. When a party could have brought a § 113(f) contribution claim, but failed to do so in a timely manner (three years had passed since the party had the availability of an action under § 113(f)) the party cannot evade the statute of limitations and the allocation scheme of a § 113(f) contribution claim by bringing a § 107(a) cost recovery action.[21]

Moreover, the mutually exclusive approach permits a PRP that has incurred costs as a result of both a civil action or settlement agreement and voluntary cleanup at a single site to pursue both cost recovery and contribution actions without compromising CERCLA’s liability structure. Under the mutually exclusive approach, when any of the statutory triggers for a contribution claim occurs for certain expenses the party may only bring a § 113(f) contribution action for those expenses.[22] But, the same party may also bring a § 107(a)(4)(B) action to recover expenses that fall outside of the contribution action.[23] “[A] party’s right to contribution for some of its expenses at a site does not necessarily mean that the party loses its right to bring a cost recovery action for other expenses.”[24] Thus, costs incurred from work performed outside the obligations of an “administrative or judicial settlement” are recoverable under § 107(a)(4)(B).

B. Availability of Cost Recovery Pursuant § 107(a)(4)(B) Under the Mutually Exclusive Approach

Following the Court decision in Atlantic Research, private parties may bring a cost recovery action against other PRPs to recover costs directly incurred from engaging in cleanup pursuant to § 107(a)(4)(B).[25] This distinction does not resolve the issue of what cause of action is applicable when PRPs are obligated to incur cleanup costs pursuant to a civil action, an “administrative or judicially approved agreement,” or a private agreement. In all of these circumstances, a PRP does not reimburse another entity, but rather incurs costs directly.

The Third Circuit Court in Agere Systems applied the mutually exclusive approach to determine which, if any, private cause of action is available to a private entity that is obligated under a private settlement agreement to fund a response action. The Third Circuit held that in such circumstances the private party may recover their costs with a § 107(a)(4)(B) cost recovery action.[26] In Agere, the plaintiffs that had been subject to EPA § 107(a) civil actions were required to comply with consent decrees by doing work such as cleanup at the contaminated facilities.[27] The two plaintiffs not subject to the consent decree (Agere and TI)[28] joined a private settlement agreement with the plaintiffs subject to the consent decree.[29] The private settlement agreement required Agere and TI to fund the other plaintiffs’ two consent decrees.[30] Agere and TI then brought a cost recovery action under § 107(a)(4)(B) against other PRPs.[31] The Third Circuit Court held that the Agere and TI were permitted to bring a cost recovery action pursuant § 107(a)(4)(B).[32]

The Third Circuit explains that this holding is in-line with the Court’s decision in Atlantic Research. First, Agere and TI “incurred” costs in the ordinary sense since they were paying for ongoing work.[33] Second, when the Court made the statement that payments made pursuant to a settlement agreement are not recoverable with a § 107(a)(4)(B) cost recovery claim, those parties had § 113(f) contribution claims for their settlement amounts.[34] In contrast, the two Agere plaintiffs did not have such contribution claims, and as a result they would not have an avenue to recover those amounts under CERCLA if they were not permitted to utilize § 107(a)(4)(B).[35]

The Third Circuit goes on to explain that Congress could not have intended such an outcome because CERCLA’s goal is “to encourage private parties to assume the financial responsibility of cleanup by allowing them to seek recovery from others.”[36] CERCLA should not be read to discourage private entities’ participation in cleanup in situations where they have not yet been sued, but are aware that they may bear some responsibility for cleaning up hazardous waste.[37] The Third Circuit correctly explained that private entities would be less likely to settle if it is uncertain whether they can seek to recover some of the amounts they will contribute.[38] If they cannot recover costs for participating in cleanup, then they will wait for a party to file a civil action against them to ensure they can sue for contribution against other PRPs.[39]

Most courts have drawn this line, holding “that costs may be recovered under § 107(a)[(4)(B)] notwithstanding that they may have been ‘compelled’ under an administrative order or settlement with the government where that order or settlement does not give rise to contribution rights under § 113(f)(3)(B).”[40] But if a PRP meets one of the requirements for suit under 113(f)(1) or (3)(B), it must proceed under that § 113 subsection.[41]

C. The Benefits of the Mutually Exclusive Approach

This mutually exclusive framework advances CERCLA’s goals by bringing all of the responsible parties to the settlement table, therefore ensuring responsible parties pay their fair share of the cleanup.[42] This framework promotes the private causes of action that Congress contemplated when it enacted SARA.[43] It does not allow a settling party to circumvent the contribution bar by bringing a § 107(a)(4)(B) action against another settling party for compelled costs pursuant to its settlement agreement. Moreover, the mutually exclusive framework does not allow a settling party to wait until its contribution claim is no longer ripe once the statute of limitations has run.

Although settling parties may be subject to § 107(a)(4)(B) cost recovery actions as a result of the mutually exclusive approach, settlements in most situations do not “‘resolve liability’ for response actions not yet completed or costs of responses not yet incurred.”[44] Thus, a cost recovery action that is permitted under the mutually exclusive approach against a PRP that has already settled or been subject to a civil action is for cleanup that the party did not yet resolve its liability for, and they may counterclaim for contribution under § 113(f)(1). Furthermore, settling PRPs may be subject to claims of contribution for settlements to which it was not a party. The idea is that by the end of response actions, each phase will have a settlement with possibly different PRPs. Through the exhaustion of contribution actions, each PRP will ultimately be responsible for their fair share, and thus fully reimbursing the entities cleaning up the contamination.

However, if instead of settling, a PRP decides to wait and see whether the United States, the State, or another entity brings an action against them, they risk the possibility of being subject to a recovery action for all costs incurred from a facility. As a result, they will bear the costs of: (1) the initial litigation; (2) the substantial judgment amount; and (3) the burden of seeking out other PRPs and bringing claims in contribution, until they are relieved of the inequitable dispersion of costs. This is the original intent of CERCLA.


CERCLA’s purpose is to facilitate the prompt cleanup of contaminated sites that pose a risk to health and welfare of the country. CERCLA’s success and integrity hinges on PRPs’ cooperation in voluntarily cleaning up sites, reimbursing the EPA for response costs, and sorting out amongst themselves the equitable allocation of the costs based on their responsibility. The mutually exclusive framework created by the United States’ Courts of Appeals encourages that cooperation. It maintains the liability structure that Congress contemplated when it adopted SARA, and ensures that the responsible parties at some point throughout a site’s cleanup will be allocated their share of the costs. In conclusion, circuits that have not yet heard these issues should adopt the mutually exclusive approach to maintain CERCLA through consistency and reliability.

[1] Gaba, supra note 62, at 142.

[2] Id.

[3] See Diamond X. Ranch LLC v. Atl. Richfield Co., 2016 U.S. Dist. LEXIS 114799, 12 (2016) ( “[A] party who may bring a contribution action for certain expenses must use the contribution action, even if a cost recovery action would otherwise be available.”) (quoting Whittaker Corp. v. United States, 825 F.3d 1002, 1007 (9th Cir. 2016)); see also Niagara Mohawk Power Corp. v. Chevron U.S.A., 596 F.3d 112, 118 (2d Cir. 2010); Hobart Corp. v. Waste Mgmt. of Ohio, Inc., 758 F.3d 757, 767 (6th Cir. 2014); Bernstein v. Bankert, 733 F.3d 190, 206 (7th Cir. 2013); Solutia, Inc. v. McWane, Inc., 726 F.Supp. 2d. 1316, 1342 (N.D. Ala. 2010); Morrison Enters., LLC v. Dravo Corp., 638 F.2d 594, 603 (8th Cir. 2011); Agere Sys. v. Advanced Envtl. Tech. Corp., 602 F.3d 204, 229 (3d Cir. 2010).

[4] See discussion infra Section III.B.

[5] Atl. Research Corp., 551 U.S. at 139; Agere Sys. v. Advanced Envtl., Tech. Corp., 602 F.3d 204, 220 (3d Cir. 2010); see Solutia, Inc. v. McWane, Inc., 726 F. Supp. 2d 1316, 1336 (N.D. Ala. 2010) (quoting Atlantic Research, 551 U.S. at 139.).

[6] 42 U.S.C. § 9613(f)(2).

[7] Pa. Dep’t of Envtl. Prot. v. Lockheed Martin Corp., 2015 U.S. Dist. LEXIS 10814 at *15–16 (2015) (citing United States v. Cannons Eng’g Corp., 899 F.2d 79, 85 (1st Cir. 1990)).

[8] Fla. Power Corp. v. First Energy Corp., 810 F.3d 996, 1001(6th Cir. 2015) (alterations omitted) (citing Hobart v. Waste Mgmt. of Ohio, Inc., 758 F.3d 757, 768 (6th Cir. 2014)).

[9] 42 U.S.C. § 9612(i) (2012); Lockheed Martin Corp., 2015 LEXIS 10814 at *16.

[10] See Lockheed Martin Corp., 2015 LEXIS 10814 at *17 (holding that the agreement was neither an administrative settlement nor judicially approved settlement because the agreement was made without following administrative procedures and no impartial arbiter determined whether the settlement amount was fair and reasonable); see CPC Int’l v. Aerojet-Gen. Corp., 759 F. Supp. 1269, 1283 (W.D. Mich. 1991) (stating that an “administrative or judicially approved” settlement must include hearings and public comment).

[11] Lockheed Martin Corp., 2015 LEXIS 10814 at *16.

[12]Id. at *18 (2015).


[14] Id. at *15; Gray, supra note 13, 175 (2016).

[15] Lockheed Martin Corp., 2015 LEXIS 10814 at *14; U.S. v. Aerojet General Corp., 606 F.3d 1142, 1149 (9th Cir. 2010); Gray, supra note 11, 175 (2016) (This benefit is limited as it only applies to “matters addressed in the settlement.”); see also 42 U.S.C. § 9613(f)(2) (2012).

[16] Lockheed Martin Corp., 2015 LEXIS 10814 at *15.

[17] Id.

[18] Id. at *18 (2015).

[19] Id. at *29 (2015).

[20] See Diamond X. Ranch LLC v. Atl. Richfield Co., 2016 U.S. Dist. LEXIS 114799, at *12 (2016) (quoting Whittaker Corp. v. United States, 825 F.3d 1002 (9th Cir. June 13, 2016)); Niagara Mohawk Power Corp. v. Chevron U.S.A., 5966 F.3d 112, 112 (2d Cir. 2010); Hobart Corp. v. Waste Mgmt. of Ohio, Inc., 758 F.3d 757, 767 (6th Cir. 2014); Bernstein v. Bankert, 733 F.3d 190, 206 (7th Cir. 2013); Solutia, Inc. v. McWane, Inc., 726 F. Supp. 2d 1316, 1342 (N.D. Ala. 2010); Morrison Enters., LLC v. Dravo Corp., 638 F.2d 594, 603 (8th Cir. 2011); Agere Sys., Inc. v. Advanced Envtl. Tech. Corp., 602 F.3d 204, 229 (3d Cir. 2010).

[21] ITT Indus. v. BorgWarner, Inc., 615 F.Supp.2d 640, 646–48 (W.D. Mich. 2009).

[22] See Whittaker Corp. v. United States., 825 F.3d 1002, 1011 (9th Cir. 2016)(holding that the plaintiff could only bring a contribution action for expenses it was found liable for in a prior action).

[23] See Whittaker Corp., 825 F.3d at 1009 (9th Cir. 2016) (holding that plaintiffs could recover costs with a cost recovery action for expenses separate from those which the plaintiff was found liable for in a prior action); Bernstein v. Bankert, 733 F.3d 190, 202–03 (7th Cir. 2012) (holding that plaintiffs could bring cost recovery action for expenses separate from those for which the plaintiffs had a right of contribution); NCR Corp. v. George A. Whiting Paper Co., 768 682, 690–92 (7th Cir. 2014) (holding that the plaintiff was required to bring all claims in contribution because each set of expenses was covered by an order triggering the right to contribution); Agere Sys., Inc. v. Advanced Envtl. Tech. Corp., 602, F.3d 204, 225 (3d Cir. 2010) (holding that a party who had been sued in a § 107(a) cost recovery action could bring a cost recovery action for expenses separate from the liability established by the prior suit, because § 113(f) had not been triggered for those separate costs and a contribution action was therefore unavailable for those costs it seeks).

[24] Whittaker Corp. v. United States, 825 F.3d 1002, 1011 (9th Cir. 2016).

[25] See discussion supra Section III.

[26] Agere Sys., Inc. v. Advanced Envtl. Tech. Corp., 602, F.3d 204, 225 (3d Cir. 2010).

[27] Id. at 21l.

[28] Id. at 225–26

[29] Id. at 212.

[30] Id.

[31] Id. at 225.

[32] Id.

[33] Id.

[34] Id.

[35] Id.

[36] Id. at 226 (3d. Cir. 2010) (citing Key Tronic Corp. v. United States, 511 U.S. 809, 819 n.13 (1994)).

[37] Id.

[38] Id.

[39] Id. (citations omitted).

[40] See Solutia, Inc. v. McWane, Inc., 726 F.Supp. 2d 1316, 1341 (N.D.A.L., 2010) (citing W.R. Grace & Co.Conn. v. Zotos Int’l, Inc., 559 F.3d 85, 90–91 (2d Cir. 2009) (holding that the plaintiff could bering § 107(a) claim based upon cleanup costs incurred pursuant to administrative settlement with state environmental agency that did not give rise to contribution rights under § 113(f)(3)(B), because agreement did not settle liability under CERCLA).

[41] PCS Nitrogen, Inc., v. Ross Dev. Corp. 104 F. Supp. 3d 729, 740 (D.S.C. 2015); Niagara Mohawk Power Corp. v. Chevron, (2d Cir. 2010); Hobart Corp. v. Waste Mgmt. of Ohio, Inc., 758 F.3d 757, 766 (6th Cir. 2014).

[42] A Bill to Extend and Amend the Comprehensive Environmental Response, Compensation, and Liability Act of 1980 and for Other Purposes: Hearings Before the Senate Committee on the Judiciary on S. 51, 99th Cong. 1, 52 (1985).

[43] See discussion supra Section I.B..

[44] Light, supra note 63, at 10791–800.


[i] See id. at 138–41 (describing the distinct differences between § 107 and § 113).

[ii] Jeffrey M. Gaba, The Private Causes of Action Under CERCLA: Navigating the Intersection of Section 107(a) and 113(f), 5 Mich. J. Envtl. & Admin. L. 117, 141 (2015); Gray, supra note 13, at 258 (2016).

[iii] Alfred R. Light, Avoiding the Contribution “Catch-22”: CERCLA Administrative Orders for Cleanup Are Civil Actions, 46 ELR 10791, 10791–92 (2016).

[iv] An American card game where the aim of the game is to discard all of your cards to get out of the game first, the last one holding a deck of cards is the loser.

[i] Elizabeth F. Mason, Comment, Contribution, Contribution Protection, and Nonsettlor Liability Under CERCLA: Following Laskin’s Lead, 19 B.C. Envtl. Aff. L. Rev. 73, 74–75 (1991).

[ii] 42 U.S.C. § 9604(a).

[iii] Id. § 9607(a).

[iv] Id. § 9606(a).

[v] Peter L. Gray, The Superfund Manual: A Practitioner’s Guide to CERCLA Litigation 255 (2016); Judy & Probst, supra note 7, at 193 (citing H.R. Rep. No. 1016, 96th Cong., 2d Sess., pt. 1, at 18 (1980), reprinted in 1980 U.S.C.C.A.N. (94 Stat.) 6119).

[vi] United States v. Chem-Dyne Corp., 572 F. Supp. 802, 805 (S.D. Ohio 1983) (citing 1980 U.S.C.C.A.N. (94 Stat.) 6119, 6119–20).

[vii] Gray, supra note 13, at 255; Judy & Probst, supra note 7, at 225.

[viii] Gray, supra note 13, at 256.

[ix] See Id. at 85 n.1, 86 n.2, 88 n.10 (listing the cases establishing CERCLA’s liability scheme).

[x] Judy & Probst, supra note 7, at 214; Luis Inaraja Vera, Compelled Costs Under CERCLA: Incompatible Remedies, Joint and Several Liability, and Tort Law, 17 Vt. J. Envtl. L. 394, 396 (2016); see also 42 U.S.C. § 9607(a) (providing the scope of those persons that may be held liable under CERCLA).

[xi] Vera, supra note 18, at 397.

[xii] United States v. Atl. Research Corp., 551 U.S. 128, 141 (2007); see Gray, supra note 11, at 257 n.3 (citing cases that found an implied right for contribution pursuant § 107(a) and federal law).

[xiii] The court in United States v. New Castle County, 642 F. Supp. 1258, 1262 (D. Del 1986) questioned whether CERCLA provided contribution rights and found a right to contribution under federal common law…In Wehner v. Syntex Agribusiness, Inc., 616 F. Supp. 27, 31 (E.D. Mo. 1985) the court that § 107(e)(2) implied a right of contribution. Look to Cooper Industries, 161-162, 125 S.Ct. 577 for a listing of these cases (if needed); Key Tronic Corp. v. United States, 511 U.S. 908, 816, also has listings of such cases.

[xiv] United States v. Atl. Research Corp., 551 U.S. 128, 140 (2007).

[xv] Judy & Probst, supra note 5, at 214; Vera, supra note 16, at 396.

[xvi] Richard H. Mays, Settlements with SARA: A Comprehensive Review of Settlement Procedures Under the Superfund Amendments and Reauthorization Act, 17 ELR 10101, 10102 (1987).

[xvii] Id. at 10102.

[xviii] 42 U.S.C. § 9613(f)(1) (2012) (emphasis added) (“Any person may seek contribution from any other person who is liable or potentially liable under [§ 107(a)] of this title, during or following any civil action under [§ 106] of this title or under [§ 107] of this title.”).

[xix] Mays, supra note 22, at 10102.

[xx] Vera, supra note 16, at 398.

[xxi] Gray, supra note 13, at 257.

[xxii] United States v. Atl. Research Corp., 551 U.S. 128, 131 (2007).

[xxiii] Id. at 132.

[xxiv] Id. at 133.

[xxv] Cooper Industries, Inc. v. Aviall Servs., Inc., 543 U.S. 157, 167–68 (2004).

[xxvi] Id.

[xxvii] Id. at 166.

[xxviii] Id.

[xxix] Id. at 166–67.

[xxx] Id. at 166.

[xxxi] Id. at 166–67.

[xxxii] United States v. Atl. Research Corp., 551 U.S. 128, 133 (2007).

[xxxiii] Id.; see, e.g., Metro. Water Reclamation Dist. v. N. American Galvanizing & Coatings, Inc., 473 F.3d 824, 835 (7th Cir. 2007) (“Nothing in subsection [§ 107(a)(4)](B) indicates that a potentially liable party . . .  should not be considered ‘any other person’ for purposes of a right of action.”).

[xxxiv] Atl. Research Corp, 551 U.S. at 133 (citations omitted).

[xxxv] E.I. DuPont de Demours & Co. v. United States, 460 F.3d 515, 543 (3d Cir. 2006), vacated, 551 U.S. 1129 (2007).

[xxxvi] Atl. Research Corp., 551 U.S. at 135–37.

[xxxvii] Id. at 138–41.

[xxxviii] Id. at 138–39 ( “[A] ‘tortfeasor’s’ right to collect from others responsible for the same tort after the tortfeasor has paid more than his or her proportionate share.”).

[xxxix] Id.

[xl] Id. at 139.

[xli] Id.

[xlii] See id. (explaining § 107, as opposed to § 113, must be used for party who incurs cleanup costs).

[xliii] Id.

[xliv] Id.

[xlv] Id.

[xlvi] Id.

[xlvii] Id.

[xlviii] Id. at 140.

[xlix] Id. (citation omitted).

[l] Id. at 140–41.

[li] Id. at 139 n.6.

[lii] Id.

[i] J.D. Candidate, 2018, Vermont Law School; Administrative Editor, Vermont Journal of Environmental Law. I would like to thank Martha Judy for her guidance and advice and the Vermont Journal of Environmental Laws Volume 19 Executive Board, for without them this article would not be possible.

[ii] Burlington N. & Santa Fe Ry. Co v. United States, 556 U.S. 599, 602.

[iii] 42 U.S.C. § 9607(a)(4)(B) (2012). Cost recovery is seen as the preferable cause of action because it has a longer statute of limitation and it provides the “opportunity for joint and several recovery.” Whittaker Corp. v. United States, 825 F.3d 1002, 1007 n.4 (9th Cir. 2016).

[iv] The contribution actions, under §§ 113(f)(1) and 113(f)(3)(B), allow parties to recover from other PRPs some of the costs they paid either pursuant to a CERCLA civil action or to “an administrative or judicially approved settlement” through equitable apportionment. 42 U.S.C. §§ 9613(f)(1), (3)(B).

[v] 42 U.S.C. § 9613(g)(3) (contribution actions are subject to a three-year statute of limitations); Whittaker Corp. v. United States, 825 F.3d 1002, 1013 (9th Cir. 2016).

[vi] United States v. Atl. Research Corp., 551 U.S. 140, 140 – 41 (2007).

[vii] Martha L. Judy & Katherine N. Probst, Superfund at 30, 11 Vt. J. Envtl. L. 191, 244–46 (2009) (explaining that after Atlantic Research Corp., the contribution-protection provision—provided to private entities in settlement agreements with the United States or States and to parties that have been subject to enforcement actions—have been called into question because uncertain whether private party cost recovery claims may be able to circumvent the contribution bar, dis-incentivizing settlements).

[viii] Luis Inaraja Vera, Compelled Costs Under CERCLA: Incompatible Remedies, Joint and Several Liability, and Tort Law, 17 Vt. J. Envtl. L. 394, 415–16 (2016).

Opportunities to Address Climate Change in the Next Farm Bill

By Sara Dewey,[1] Liz Hanson,[2] & Claire Horan[3]

This post is part of the Environmental Law Review Syndicate


The Farm Bill affects nearly every aspect of agriculture and forestry in the United States. Therefore, its next reauthorization offers an important opportunity to better manage the risks of climate change on farms, forests, and ranches by supporting resilience practices that also offer greenhouse gas (GHG) emission reductions.

Agriculture is vulnerable to the impacts of climate change, including rising temperatures, changes in rainfall and pest migration patterns, extreme weather events, and drought. In addition to being heavily affected by climate change, agriculture is also a significant contributor to climate change. Agricultural practices are responsible for about eight percent of U.S. GHG emissions.[4] Estimates of total food system emissions, which include the CO2 emissions from energy use and transportation, increase the agricultural industry’s proportion of U.S. GHG emissions to between 19 and 29 percent.[5]

To better align their practices with their long-term interests, farmers and ranchers can adopt practices that enhance their resilience, while also reducing GHG emissions, and increasing carbon sequestration. Many of these practices improve the long-term productivity and profitability of farms. For example, farmers are already adopting practices that reduce emissions or sequester carbon in the soil and in woody biomass while also improving productivity and resilience on their land.

This paper proposes a suite of practices that should be considered during the next authorization of the Farm Bill to improve on-farm efforts to adapt to and mitigate climate impacts. It is organized into four main sections. Part I provides background on the Farm Bill and the ways that the U.S. agricultural system contributes to GHG emissions. Part II provides an overview of opportunities for on-farm mitigation and adaptation. Many of the practices we recommend can reduce on-farm emissions and build a more resilient agricultural system. Part III identifies a set of metrics that we used to assess potential proposals. Lastly, Part IV summarizes how climate practices can be incorporated across titles and highlights three policy options.

I. Background

A. Agricultural Sources of GHG Emissions

Greenhouse gases trap heat in the atmosphere and contribute to increases in global temperatures. Although this a natural process, increased greenhouse gas emissions since the industrial revolution have increased atmospheric greenhouse gases to levels never before recorded. Agriculture, including raising crops and animals as well as resulting land use changes and farm equipment usage, is a source of three GHGs: methane (CH4), nitrous oxide (N2O), and carbon dioxide (CO2).[6]

Figure 1. GHG Profiles[7]

Globally, emissions from food systems are responsible for nearly a third of all GHG emissions.[8] Domestically, EPA’s Inventory of U.S. Greenhouse Gas Emissions and Sinks divides up agriculture-related emissions into different categories. N2O and CH4 emissions are categorized as “Agricultural,” and accounted for 8.3 percent of total greenhouse gas emissions in the United States in 2014.[9] In 2014, N2O emissions were 336 million metric tons of carbon dioxide equivalent (MMT CO2 Eq.); these emissions were caused primarily by soil management such as the use of synthetic fertilizers, tillage, and organic soil amendments.[10] Manure management, and biomass burning, also contribute to N2O emissions. CH4 emissions were 238 MMT CO2 Eq. and were produced by enteric fermentation during ruminant digestion (164 MMT CO2 Eq.), manure management (61 MMT CO2 Eq.), and the wetland cultivation of rice (12 MMT CO2 Eq.)[11]

CO2 emissions from agriculture-related land use changes and equipment usage are accounted for in the “Land Use, Land-Use Change, and Forestry” and the “Energy” categories, respectively. Estimates of total food system emissions, which include the CO2 emissions from energy use and transportation, increase the agricultural industry’s proportion of U.S. GHG emissions to between 19 and 29%.[12]

II. Strategies for Managing Climate Risk through Mitigation and Adaptation

Given agriculture’s contributions to GHG emissions that are contributing to climate change, which in turn affects agricultural productivity, it is appropriate to consider how climate change can be incorporated across the titles of the Farm Bill. The anticipated reauthorization in 2018 can play a critical role in addressing climate change in the United States by promoting practices that encourage mitigation and adaptation practices on farms.

Adopting new agricultural practices can be challenging, especially for small farmers or operations without access to large amounts of capital or information about adaptation opportunities. However, doing so will not only assist the U.S. farmers and ranchers confront shifting seasons, more severe storm events, new pests, drought, and other challenges,[13] it will also reduce the Farm Bill’s fiscal burden on taxpayers.[14] A number of land managers are already adopting strategies that not only reduce emissions or sequester carbon in the soil, but also have the important co-benefits of improving productivity and resilience.[15]

A. Mitigation Measures

Land managers can mitigate GHG emissions by offsetting current emissions, sequestering carbon, and/or preventing future emissions.[16] Figure 2 describes these strategies and the practices to achieve them.

First, land managers can reduce the GHG emissions of their farming practices in a number of ways. Practices such as conservation tillage reduce soil disturbance, and prevent some erosion, which can lower soil carbon loss. Precision agriculture strategies can reduce fertilizer inputs on cropland, which in turn reduces GHG emissions from fertilizer production and application.[17] Reincorporating livestock manure onto cropland as well as improved management of liquid manure using anaerobic digesters or other on-farm technology can reduce methane emissions from livestock waste by capturing it rather than emitting it.[18]

Second, land managers can sequester additional carbon through on-farm practices. Soil carbon can be increased by incorporating cover crops, including legumes, into crop rotations, reducing tillage, and agroforestry practices.[19] In addition, planting perennial crops or incorporating trees into farms through alley cropping, hedgerows, and riparian forest buffers can lead to long-term sequestration of carbon in woody biomass.

Finally, land managers can take steps to avoid future emissions. The most critical way to avoid new on-farm emissions is to avoid land conversion, which releases carbon that was previously sequestered in the soil and in woody biomass.

Figure 2. Practices for agricultural greenhouse gas mitigation[20]

B. Adaptation Measures

Adapting to a changing climate will require farmers, foresters, and ranchers to prepare for and respond to new risks, including extreme weather events, shifts in growing seasons, and different pests and plant diseases. Figure 3 provides an overview of the range of practices that farmers can undertake to adapt to climate change.

To make farming operations more resilient, farmers can enhance soil health, which will make agricultural systems better able to withstand extreme weather, drought, and erosion due to high winds or flooding.[21] Strategies for enhancing soil health include adjusting production inputs, timing of planting and soil amendments, cover crops, tillage, new crop species, and diversified crop rotations.[22]

Farmers can also take additional steps to make their farms more resilient to other climate risks. For example, to prepare for flooding, heavy rainfall, and other risks, farmers can implement resilient farm landscapes that include buffer strips and the return of marginal cropland to native vegetation. To prepare for new pests and diseases, farmers can diversify their crop selection and alter crop rotations. To adjust to changing seasons and a warming climate, farmers can plant different crops; crop scientists can also develop more heat- and drought-resistant crop varieties. Resilience planning is also important on the community level, as rural communities can ensure that new infrastructure investments supported by the Farm Bill, such as rural water and energy systems, are resilient to climate change effects.

Figure 3. Practices for agricultural adaptation to climate change[23]

C. Opportunities for Complementary Mitigation and Adaptation

Importantly, many on-farm practices can help with both climate adaptation and mitigation.[24] For example, improving soil health not only mitigates climate change, it also makes farms more resilient and better able to withstand the shifting, and at times extreme, conditions of a changing climate. Efficient fertilizer application will reduce GHG emissions while enhancing soil resilience. Similarly, cover cropping, diversified crops, and other practices that stabilize the soil will reduce GHG emissions from the soil while building soil health. It is important to note that the efficiency of these on-farm practices will vary by region, impacting the ways they can and should be implemented.[25]

Mitigation and adaptation strategies for agricultural systems often require long-term planning to strengthen “climate-sensitive assets,” such as soil and water, over time and in changing conditions.[26] Developing better regionally specific agricultural climate and conservation practice adoption data is required for this long-term planning to be successful. From those baseline data, regional efforts will be critical to identify mitigation opportunities, develop strategic adaptation planning, and implement enhanced soil and livestock management practices.[27]

III. Metrics for Prioritizing Reform Proposals

As the summary above indicates, there are many actions that can promote climate change mitigation or adaptation in agriculture. In addition, changes can be made to every Title of the Farm Bill that would promote one or more of these mitigation and adaptation strategies. Given this complexity, the uncertainties associated with quantitative estimates of the mitigation potential of different strategies, and the qualitative differences between mitigation and adaptation as goals, we developed a range of qualitative metrics that we used to analyze potential reforms. In particular, we considered:

  • Potential magnitude of climate impact: Priority was given to proposals that had proven climate benefits, did not require significant additional research, and targeted the largest sources of agricultural GHG emissions.
  • Co-benefits: Priority was given to proposals that could increase resiliency or economic benefits of farms.
  • Equity: Priority was given to programs that could benefit small and large farms in all regions.
  • Scalability: Priority was given to proposals that seemed replicable and applicable to farms across the country or where Climate Hubs could facilitate regional diversity.
  • Enforceability/Administrability: Priority was given to proposals that could be tied in with or build upon existing requirements or programs in the Farm Bill.
  • Feasibility: Feasibility considerations included ease of implementation technically, economically, and politically. Because any legislative change will need to be passed in Congress, political feasibility was determined to be one of the most important considerations. Accordingly, we prioritized proposals that seemed, based on stakeholder engagement, suitable for the next Farm Bill, given competing interests for funding and stakeholder sentiment towards climate action.

An analysis of these metrics is included throughout our recommendations. However, these should be considered as only a first step. While we have attempted to target the largest sources of GHG emissions, more detailed proposals will be required before there can be precise estimates of the potential for emission reductions. The USDA’s COMET-Farm, an online farm and ranch GHG accounting tool, can likely facilitate this effort.[28] Similarly, determining the economic feasibility of specific reform proposals has been difficult because of taxpayer subsidization, the uncertainty of how appropriations may be allocated, and the varying degrees of stringency that reforms could encompass (e.g. mandate vs. incentive). Finally, while previous Farm Bill reauthorizations can serve as a guide, the ongoing transitions at U.S. federal agencies engaged in Farm Bill programs will likely have impacts on the political feasibility of proposals that cannot be appropriately assessed at this time. For these reasons, we recommend that additional research measure the climate impact of proposals, outline the benefits and co-benefits for farmers and the public, articulate the administrability of the program, and gather stakeholder input and support for proposals.

IV. Pathways for Addressing Climate Change in the Farm Bill

To determine how the Farm Bill could better address climate change, we first categorized the range of mitigation and adaptation practices identified in Figures 2 and 3, above, in terms of their potential applicability to the Farm Bill. We then examined how these practices mapped onto the current titles in the Farm Bill. Finally, we assessed how the upcoming Farm Bill could better incentivize these actions across titles, with an eye toward win-win practices with both mitigation and adaptation benefits.

Figure 4 contains the range of possibilities we identified for addressing climate mitigation and adaptation by title. To fully assess the impact of each of these policy options – and its interaction with other policies and programs –requires additional research and outreach to stakeholders affected. We discuss in more detail below a set of recommendations that best fit our metrics, indicated by bold font in this table.

Figure 4. Options for Addressing Climate Change by Farm Bill Title

All of these areas for reform have the potential to advance climate-ready agricultural practices through the Farm Bill. Many of these areas for reform also have wide-ranging benefits beyond climate change mitigation or adaptation such as enhancing on-farm productivity and more efficiently using taxpayer dollars. We elected to focus on three recommendations we judged to be particularly important based on the metrics we established in Part III).

  • Recommendation 1: Incorporate climate measures into crop insurance and conservation compliance to better manage on-farm climate risks under Title II (Conservation) and Title XI (Crop Insurance).
  • Recommendation 2: Ensure the best available science and research—including the outcome of pilot programs—are incorporated into Farm Bill programs; support dissemination of downscaled climate data through USDA regional offices and land grant universities to develop agricultural climate mitigation and adaptation capacity under Title VII.
  • Recommendation 3: Advance manure management collection and storage methods, as well as biogas development under Title IX to mitigate GHG contributions from livestock.

Recommendation 1: Incorporate Climate into Crop Insurance and Conservation Compliance

 a. Reform crop insurance to incentivize climate risk management and eliminate   disincentives for adopting climate-friendly practices

Crop insurance, Title XI, makes government-subsidized crop insurance available to producers who purchase a policy covering losses in yield, crop revenue, or whole farm revenue. Farmers can select and combine several types of crop insurance policies: catastrophic coverage, “buy-up” coverage, and a supplemental coverage option for selected crops. USDA’s Risk Management Agency (RMA) sets insurance premium subsidy rates and develops specific contracts,[29] working with 18 insurance companies to administer the program.[30]

Crop insurance is deeply subsidized by the federal government, and it represents the single largest federal outlay in the farm safety net.[31] On average, taxpayers cover 62 percent of crop insurance premiums.[32] The insurance companies’ losses are reinsured by USDA, and the government also reimburses their administrative and operating costs.[33] The Congressional Budget Office anticipates that this program will cost taxpayers over $40 billion from 2016 to 2020.[34]

These subsidies disproportionately benefit large farms: while only about 15 percent of farms use crop insurance, insured farms account for 70 percent of U.S. cropland.[35] Small farmers struggle to utilize crop insurance because of the high administrative burden and challenges of insuring specialty crops.[36] In addition to clear equity concerns involving access to crop insurance, this situation is problematic from a climate perspective because larger farms are more likely to grow monocultures, which are both more vulnerable to pests and extreme weather events and can degrade soil health. Indeed, just four crops—corn, cotton, soybeans, and wheat—make up about 70 percent of total acres enrolled in crop insurance.[37]

The current loss coverage policies in the crop insurance program can discourage farmers from proactively reducing their risks by taking steps to enhance soil health and resilience. Because farmers with crop insurance are protected against losses incurred from impacts likely to increase with climate change, farmers may not be properly incentivized to respond to the changing conditions.[38] Some environmental organizations have even raised concerns that in response to the crop insurance transfer of risk, some farmers may be more willing to engage in unsustainable practices, such as aggressive expansion, irresponsible management, and use of marginal land.[39] In addition, farmers may make planting decisions based on the insurance program incentives rather than market-based signals.[40] In these ways, crop insurance can push farmers towards practices that pose risks to both their operations and taxpayer obligations.[41] It is therefore important that the crop insurance program better align farmers’ risk management incentives with the real and growing risks they face from climate change.

One way to achieve this objective is through incentivizing or requiring farmers to undertake actions to improve soil management and promote soil health. Some specific changes to the crop insurance program that could promote these practices include:

  • Incorporating climate projections to account for changing growing seasons and planting dates.
  • Providing insurance premium rebates for farmers who voluntarily undertake beneficial practices.
  • Incentivizing improved soil management practices, diversified crops, and manure management.
  • Adjusting the length of policies to better reflect the value added from changes that improve long-term soil health.
  • Writing soil health requirements into insurance policies.

More generally, changes to the crop insurance program that reduce the magnitude of the subsidy offered to farmers, such as setting a dollar-per-acre cap, could reduce the moral hazard that current policies create.[42] The methodology used to set premiums could also be adjusted to be based more on the projected frequency and intensity of events such as droughts and floods rather than on backward-looking data. RMA has started to incorporate climate-related risk metrics into annual rates by weighting recent loss experience more heavily, thereby more accurately reflecting the risks that growers face. However, it is important to consider future risks from climate change as well.

Requirements of the crop insurance program that act as disincentives to climate-friendly farming practices should be updated to account for growing climate risks farmers face. For example, RMA has guidelines in place about the termination of cover crops, because of concerns that these crops will scavenge water from the commodity crops.[43] This requirement can act as a disincentive to farmers’ adoption of cover cropping, a practice that builds the soil and reduces runoff in the non-growing season.[44] The next Farm Bill could specify that there should be no specific termination requirements for cover crops.

Insurance policies may also serve to incentivize some environmentally harmful practices, such as early and excess fertilizer application and cultivation of environmentally sensitive land.[45] Because early application maximizes crops’ uptake of nitrogen, it can increase yield in the short term, but it contributes to nitrous oxide emissions, unhealthy soils that become less able to fix nitrogen and must rely increasingly on fertilizer, and polluted runoff. In addition, synthetic fertilizers, which are made from non-renewable materials, including petroleum and potash, are produced at a huge energy cost.[46] Some studies have suggested that crop insurance may incent some farmers to convert highly erodible or wetlands to farmland.[47] Therefore, the next Farm Bill could also indicate this type of practice is not required to be eligible for crop insurance. This change could be complemented by an increase in the length of insurance policies, as discussed above, because insurance companies would benefit from the longer-term improvements in soil health.

b. Tie crop insurance to a new conservation compliance provision for building soil health for climate ready agriculture

Currently, in order to qualify for crop insurance, farmers must satisfy two conservation compliance requirements, the Wetland Conservation (“Swampbuster”) and Highly Erodible Land Conservation (“Sodbuster”) provisions.[48] These provisions ensure, respectively, that farmers do not convert a wetland or plant crops on highly erodible land or a previously converted wetland.[49] While these current conservation requirements are beneficial in addressing some climate impacts, adding a conservation compliance requirement directly targeted at climate-related practices would improve upon them.

With 70 percent of farmland in the crop insurance program, changes in conservation compliance through the next Farm Bill or through RMA’s policies can drive big climate change benefits. Under Title II, Congress could create an additional conservation compliance requirement for climate-friendly agricultural practices, which could either be required to obtain crop insurance or could make farmers eligible for rebates. The types of on-farm practices that could mitigate risk and enhance climate resilience include more precise irrigation and fertilizer application, reduced tillage of the soil, cover cropping, altering crop rotations, and building buffer strips and riparian buffers. Particularly beneficial practices for building resilient soil include cover cropping, diversified crop rotations, reducing tillage, and efficient irrigation.[50]

In addition, enforcement gaps have limited the success of the existing conservation compliance requirements. To make the mechanism effective, it will be important to establish simple and effective enforcement, for example by using remote sensing, and to ensure that Natural Resources Conservation Service (NRCS) offices have sufficient resources to carry out enforcement efforts.

First, these proposals could produce significant climate benefits from increasing soil health, in terms of both mitigation and adaptation. Reform of the crop insurance and conservation titles could also help address some of the equity issues that currently exist between small and large farms. Existing USDA programs, described in the next section, could help with scalability and administrability. Finally, in terms of feasibility, while any change may be difficult, our stakeholder engagement indicated that farmers are open to programs that target soil health, given the potential economic benefits to their farms. While the actual on-farm impacts will vary based on how the program is designed and constructed, building more resilient, healthy soil can help improve environmental outcomes and decrease the risk of crop loss.[51]

Recommendation 2: Ensure Best Available Science and Research Guides Farm Bill Programs

Agricultural practices that promote climate change mitigation and adaptation, including those described above, are often regionally specific in their implementation. For many new climate-ready practices to be included in conservation compliance or crop insurance, the USDA would need to account for this regional specificity. For example, the benefits of many of the on-farm practices that improve soil health, including more precise irrigation and fertilizer application, reduced tillage of the soil, and altering crop rotations, vary by region and soil type. In some areas, no-till methods may be infeasible; farmers who try to implement no-till in these areas would likely continue to till to some degree or after a short period of time, resulting in quick reversal of the achieved carbon sequestration benefits. Furthermore, the technical specificity of choosing among these practices and correctly implementing them requires guidance at a local level.

To address these types of knowledge gaps and to provide technical assistance to states and farmers, the USDA has created a range of programs, including Climate Hubs, which were established at public land-grant universities in 2014.[52] The Hubs deliver science-based knowledge, practical information, and program support for farmers to engage in “climate-informed decision-making” by farmers.[53]

Increasing funding in the 2018 Farm Bill in Title VII, the Research title, could solidify and expand USDA’s ability to administer and scale climate research and outreach efforts across all regions of the country. Additionally, creating systems to collect and analyze regional data on pilot programs and ensure best practices are adopted could assist long-term efforts to incorporate climate policies into Farm Bill programs.[54] For these reasons the Farm Bill should provide additional funding for climate research and monitoring, especially focused on regional resilience.

Recommendation 3: Address the Significant GHG Contributions of Livestock Management

Improving livestock management, especially manure management, is a significant opportunity for mitigating emissions of methane and achieving several co-benefits for the public and farmers. There is currently very little regulation of livestock manure management. Manure is sometimes stored—uncovered—in a single collection site, which causes the methane to be released directly into the atmosphere. In addition to being a major GHG emissions source, it can cause a range of considerable environmental harms.[55]

a. Require improved manure management, including the covering of lagoons

First, the upcoming Farm Bill could address manure management collection and storage methods. Practices can be improved through actions such as allowing livestock to roam,[56] covering manure lagoons, flaring the methane produced, or producing biogas for use. Simply covering a manure lagoon results in significant decreases in methane emissions, as well as decreased odors. Flaring is the combustion of methane, which yields water and carbon dioxide. Although flaring still emits GHGs, carbon dioxide is a less potent GHG than methane.

The Farm Bill could promote these practices either through incentives or mandates in the Conservation or Crop Insurance titles. For example, the Farm Bill could mandate or incentivize farmers with a threshold number of cattle, swine, or poultry cover manure and flare the produced methane to be eligible for crop insurance. Such a mandate would have the greatest impact at Concentrated Animal Feeding Operations (CAFOs), which may also be better able to bear the high capital costs associated with biogas production.

b. Pursue strategies to decrease methane emissions, including biogas and other on-farm renewable energy production

Second, the Energy Title could incentivize on-farm biogas. On farms, many different substrates may be used to produce biogas, including animal excrements (including that of cattle, swine, poultry,[57] and horse), food waste, milling by-products, and catch crops (such as clover grass on farms without livestock).[58] Farmers can realize substantial savings from biogas production, including through substituting biogas for other energy sources, through substituting digestate[59] for commercial fertilizers,[60] and by avoiding disposal and treatment of substrates (such as for waste-water treatment). Farmers may also be able to sell carbon offsets.[61] In addition, farmers producing biogas can avoid some of the worst problems with animal agriculture: farmers must do something with the manure, and its storage can produce strong odors,[62] unhealthy conditions for workers and families,[63] and pollution through runoff in the worst scenarios.[64]

Farmers have two main options for biogas use: (1) generation of electricity for on-site use or sale to the grid; and (2) direct use of biogas locally, either on-site or nearby.[65] Using the biogas to fuel a generator to produce electricity is considered the most profitable use for most farms.[66] Another use is to upgrade the biogas, then called biomethane, to be injected into the national natural gas pipeline network as a substitute for extracted natural gas.

Because farmers could benefit financially from on-farm use or the sale of biogas, the Farm Bill should continue and expand funding for the Rural Energy for America Program, which offers cost-sharing grants and loans for renewable energy improvements. [67] However, these programs are most likely to benefit large farms because anaerobic digesters are expensive and require a large and constant supply of substrate to produce a return on investment. We therefore suggest the Farm Bill also fund pilot programs to assist small farm communities to form cooperatives so that they are also able to utilize this technology and participate in the grant or loan program.

Even with the available grants and loans, farmers are still taking a substantial financial risk. USDA or land-grant universities should actively help communities or cooperatives with the planning and application process. Large farms or cooperatives who are unable or unwilling to operate and maintain anaerobic digesters themselves could hire a company to lease the equipment and manage the biogas production process.[68] USDA Rural Development Agencies could be a valuable liaison between biogas management companies and farmers.

CAFOs could be part of a voluntary program or required to use anaerobic digesters due to their greater contribution to climate change and other environmental harms. Because CAFOs are responsible for high levels of greenhouse gas emissions and because anaerobic digesters are economically feasible for large operations, there is reason to consider the benefits that could be achieved by requiring these practices for large CAFOs in the Farm Bill.

Livestock management is a critical area for addressing climate impacts, and biogas has the potential to be a win-win for farmers willing to invest in alternative energy production.


The U.S. agricultural system must evolve to mitigate climate change and adapt to the effects of a changing climate. Opportunities for climate change mitigation and adaptation exist across the Farm Bill titles, from bolstering climate resilient infrastructure in the Rural Development title to incentivizing sustainable forest management in the Forestry Title. Taking action on climate measures in the next Farm Bill reauthorization will help farmers better plan for changing conditions, protect taxpayers from increasing risks, and assist the United States in meeting its global climate commitments. The next Farm Bill should incorporate climate risk management provisions, and state and local actors should consider ways to support these efforts.

[1] J.D., Harvard Law School, Class of 2017.

[2] M.P.P. Candidate, Harvard Kennedy School, Class of 2018.

[3] J.D. Candidate, Harvard Law School, Class of 2018.

[4] EPA, Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990-2015, at ES-21 (2017).

[5] Research Program on Climate Change, Agriculture, and Food Safety, Food Emissions (2016),

[6] EPA, Inventory of U.S. Greenhouse Gas Emissions and Sinks: 1990 – 2014, at 5-1 (2016) [hereinafter EPA, Inventory],

[7] EPA, Overview of Greenhouse Gas Emissions [hereinafter EPA, Overview], The two to three percent of emissions unaccounted for are fluorinated gases, which are synthesized during industrial processes. Id.

[8] Natasha Gilbert, One-third of our Greenhouse Gas Emissions Come from Agriculture, Nature (Oct. 31, 2012),

[9] EPA, Inventory, supra note 7, at 5-1.

[10] Id.

[11] Id.

[12] Research Program on Climate Change, Agriculture, and Food Safety, Food Emissions (2016),

[13] See U.S. Dep’t of Agric., USDA Agriculture Climate Change Adaptation Plan 9 (2014) [hereinafter USDA, Adaptation Plan],; Louise Jackson & Susan Ellsworth, Scope of Agricultural Adaptation in the United States: The Need for Agricultural Adaptation, in The State of Adaptation in the United States (2012),

[14] For example, a recent report from the Office of Management and Budget and the Council of Economic Advisers estimates that the annual cost of the crop insurance program will increase by $4 billion per year in 2080 as a result of the impacts of climate change. OMB & CEA, Climate Change: The Fiscal Risks Facing the Federal Government 6 (Nov. 2016),; see also USDA, Adaptation Plan, supra note 14, at 9.

[15] U.S. Dep’t of Agric., Climate Change and Agriculture in the United States: Effects and Adaptation 126–27 (2013) [hereinafter USDA, Effects and Adaptation],

[16] M. McLeod et al., Cost-Effectiveness of Greenhouse Gas Mitigation Measures for Agriculture: A Literature Review, OECD Food, Agriculture and Fisheries Papers, No. 89, at 26 (2015).

[17] Peter Lehner & Nathan Rosenberg, Legal Pathways to Carbon-Neutral Agriculture, 47 Envtl. L. Rep. 10,845, 10,849 (2018).

[18] Id. at 19–21.

[19] For a more detailed review of how carbon sequestration can be increased in agriculture, see Daniel Kane, Nat’l Sustainable Agric. Coal., Carbon Sequestration Potential on Agricultural Lands: A Review of Current Science and Available Practices (2015),

[20] Adapted from P. Smith et al., Greenhouse Gas Mitigation in Agriculture, Philosophical Transactions of the Royal Society B, 363, 789–813 (2008).

[21] Alexandra Bot & José Benites, Food & Agric. Org. Of the United Nations, FAO Soils Bulletin 80, The Importance of Soil Organic Matter: Key to Drought-Resistant Soil and Sustained Food and Production 19 (2005),

[22] USDA, Effects and Adaptation, supra note 16, at 123; see also Nat’l Sustainable Agric. Coal., Climate Change and Agriculture Recommendations for Farm Bill Conservation Program Implementation 2 (2014),

[23] While these practices may generally lead to better resilience on farms, adaptation practices are highly region-specific.

[24] USDA, Effects and Adaptation, supra note 16, at 126–27 (2013).

[25] For example, in the Central Valley of California, an adaptation plan that included integrated changes in crop mix and altered irrigation, fertilization, and tillage practices, was found to be most effective for managing climate risk. Id. Along with the USDA Climate Hubs, the following organizations have undertaken projects related to regional agricultural adaptation research and planning: California Healthy Soils Initiative; Wisconsin Initiative on Climate Change Impacts; Southeast Florida Regional Climate Change Compact; The Mid-Atlantic Water Program; U.S. Midwest Field Research Network for Climate Adaptation.

[26] Id. at 126.

[27] Id.

[28] See COMET-Farm,

[29] U.S. Dep’t of Agric., About the Risk Management Agency,

[30] Dennis A. Shields, Cong. Research Serv., Crop Insurance Provisions in the 2014 Farm Bill 3 (2015).

[31] Id.

[32] Id.

[33] Dennis Shields, Cong. Research Serv., Federal Crop Insurance: Background 2 (2015).

[34] Cong. Budget Office, March 2016 Baseline for Farm Programs (2016),; see also Heritage Found., Addressing Risk in Agriculture (2016).

[35] U.S. Dep’t of Agric., Structure and Finances of U.S. Farms: Family Farm Report, 2014 Edition 32–33 (2014),

[36] Generally, the more diverse or specialized crops and livestock a farmer produces, the harder it is to obtain insurance. These policies are not designed to support small producers and the policies are administratively complex and burdensome for small farmers, with high premiums for small farmers. On the one hand, if small farmers used yield-based or revenue-based insurance policies, those farmers would need to purchase insurance for each crop, which requires producing a significant volume of each single crop to justify the paperwork and setting up a contracted purchase price from a processor. On the other hand, whole farm insurance policies base policies on average adjusted gross revenue of the farm, regardless of the variety of products the farmer grows. This type of policy is more appropriate for diversified farmers, but may still be too cumbersome for small farms to participate. See Jeff Schahczenski, Nat’l Sustainable Agric. Info. Serv., Crop Insurance Options for Specialty, Diversified, and Organic Farmers (2012),; Nat’l Sustainable Agric. Coal., Have Access Improvements to the Federal Crop Insurance Program Gone Far Enough?, NSAC’s Blog (July 28, 2016),

[37] Shields, Federal Crop Insurance: Background, supra note 35, at 1.

[38] Linda Prokopy et al., Farmers and Climate Change: A Cross-National Comparison of Beliefs and Risk Perceptions in High-Income Countries, 56 Envtl. Mgmt. 492, 497 (2015).

[39] Bruce Babcock, Environmental Working Group, Cutting Waste in the Crop Insurance Program 10 (2013).

[40] Id.

[41] C. O’Connor, NRDC Issue Paper 13-04-A, Soil Matters: How the Federal Crop Insurance Program Could Be Reformed to Encourage Low-risk Farming Methods with High-reward Environmental Outcomes (2013).

[42] See, e.g., Heritage Found., Addressing Risk in Agriculture (2016).

[43] NSAC, 10 Ways USDA Can Address Climate Change in 2016, NSAC’s Blog (Dec. 30, 2015),

[44] See Practical Farmers of Iowa, Cover Crops,

[45] USDA’s Economic Research Service found that “[l]ands brought into or retained in cultivation due to these crop insurance subsidy increases are, on average, less productive, more vulnerable to erosion […] then cultivated cropland overall. Based on nutrient application data, these lands are also associated with higher levels of potential nutrient losses per acre.” USDA Economic Research Service, Report Summary: Environmental Effects of Agricultural Land Use Change (Aug. 2006); see also Daniel Sumner and Carl Zulauf, The Conservation Crossroads in Agriculture: Insight from Leading Economists. Economic and Environmental Effects of Agricultural Insurance Programs, The Council on Food, Agricultural and Resource Economics (2012).

[46] See Stephanie Ogburn, The Dark Side of Nitrogen, Grist (Feb. 5, 2010), (“About one percent of the world’s annual energy consumption is used to produce ammonia, most of which becomes nitrogen fertilizer.”).

[47] See, e.g., Anne Weir and Craig Cox, Envtl. Working Grp., Crop Insurance: An Annual Disaster (2015).

[48] Sodbuster, 16 U.S.C. § 3811 et seq.; Swampbuster, 16 U.S.C. § 3821 et seq.

[49] See Nat. Res. Conservation Serv., U.S. Dep’t of Agric., Conservation Compliance Provisions,

[50] Id. at 7.

[51] O’Connor, Soil Matters, supra note 43, at 7.

[52] U.S. Dep’t of Agric. Climate Hubs, Mission and Vision,

[53] Id.

[54] The existing ARS LTAR system, which conducts longterm sustainability research, could be used to inform the regional best practices communicated in outreach efforts. See Agric. Research Serv., U.S. Dep’t of Agric., Long-Term Agroecosystem Research (LTAR) Network,

[55] For example, manure management practices can create a public nuisance for which neighbors have little recourse. In addition, runoff from agriculture is not adequately regulated under the Clean Water Act and results in pollution to the nation’s waterways. Every year a hypoxic zone, also called a dead zone, develops where the Mississippi River dumps pollution from Midwest livestock and fertilizers into the Gulf of Mexico. See Kyle Weldon & Elizabeth Rumley, Nat’l Agric. L. Ctr., States’ Right to Farm Statutes,; Ada Carr, This Year’s Gulf of Mexico “Dead Zone” Will Be the Size of Connecticut, Researchers Say, (Jun. 15, 2016),

[56] Farms where the cattle range freely do not release as much methane to the atmosphere because the less consolidated manure is more likely to be absorbed into the soil rather than anaerobically digested to produce methane.

[57] Using poultry manure as a substrate can be difficult because feathers and poultry litter can clog anaerobic digesters. See Donald L. Van Dyne & J. Alan Weber, Special Article, Biogas Production from Animal Manures: What Is the Potential?, Industrial Uses/IUS-4 20, 22 (Dec. 1994).

[58] SustainGas, Sustainable Biogas Production: A Handbook for Organic Farmers 38 (2013),

[59] Digestate is the solid that is left over after biogas has been produced. Digestate can be sold or used on farm as fertilizer. It smells better than manure, is free of harmful bacteria, and contains nitrogen in a form that is more bioavailable for crops.

[60] 40 organic farms in Germany, in a region without livestock, have found it worthwhile to cooperate in supplying and transporting clover grass up to 50 km to an AD because the digestate provides them with a flexible organic fertilizer. See SustainGas, supra note 60, at 28. They find that the digestate leads to higher quality for their food crops. Id. “Biogas has to serve food production via improved nutrient supply,” one farmer says. Id.

[61] If farmers can show that they have reduced their methane emissions, they may be able to sell the carbon offsets in exchanges such as the California GHG cap and trade market. See Cal. Air Resources Bd., Compliance Offset Protocol, Livestock Projects: Capturing and Destroying Methane from Manure Management Systems (2014),

[62] The odor-reducing benefits are viewed as especially desirable for poultry and swine farms.

[63] Biogas plants dispose of waste and sewage, making conditions healthier. Not only does the anaerobic digestion process remove pathogens, but because biogas production requires collecting manure at a central location, some unhygienic conditions are avoided. See Julia Bramley, et al., Tufts Department of Urban & Environmental Policy & Planning, Agricultural Biogas in the United States: A Market Assessment 122 (2011),

[64] Livestock manure generated at cattle yards and dairy farms can contaminate surface and ground water through runoff. Anaerobic digestion sanitizes the manure to a large extent, decreasing the risk of water contamination. Id.

[65] EPA, AgSTAR Handbook: A Manual for Developing Biogas Systems at Commercial Farms in the United States, 2d. ed. 2-5 (K.F. Roos et al. eds. Feb. 2004).

[66] Id. at. 3-1. For most farms, electricity comprises 70% to 100% of energy use. Id.

[67] U.S. Dep’t of Agric., Rural Energy for America Program Renewable Energy Systems & Energy Efficiency Improvement Loans & Grants,

[68] This model is frequently used for wind energy production. See Agric. Research Serv., U.S. Dep’t of Agric., Wind and Sun and Farm-Based Energy Sources, Agric. Res., Aug. 2006,

The Case for Cap-and-Trade: California’s Battle for Market-Based Environmentalism

By Theodore McDowell, J.D. 2017, University of Virginia School of Law

This post is part of the Environmental Law Review Syndicate. Click here to see the original post and leave a comment.

I. Introduction

The California Cap-and-Trade Program (“CAT”) is derived from the California Global Warming Solutions Act of 2006 (“Global Warming Act”), which requires the State to reduce its greenhouse gas (“GHG”) emissions to 1990 levels by 2020.[1] The California Air Resource Board (“CARB”) is the State regulatory agency responsible for the project.[2] In 2011, the CARB adopted cap-and-trade regulations and created the CAT to set limits on GHG emissions.[3] The first auctions for the CAT were held in 2012, and the program went into full effect on January 1, 2013.[4]

The CAT operates in two phases each year. First, a number of emission allowances are freely distributed to entities that fall under the purview of the program.[5] Second, the remaining allowances are auctioned off on a quarterly basis.[6] The free distributions are reduced annually, and eventually all the allowances will be distributed via auctions.[7] The program also permits carbon offsets to satisfy up to eight percent of an entity’s compliance obligations.[8] The ultimate objective is to create incentives for businesses to craft environmentally friendly industrial practices as the number of yearly allowances decreases over time.

The CAT also has an enormous scope, and it is the world’s second largest market-based mechanism designed to reduce GHG emissions.[9] This size makes the successful implementation of the program especially impressive. The success is due largely to a design structure that draws upon the shortcomings of previous cap-and-trade initiatives, such as the Regional Greenhouse Gas Initiative (“RGGI”) in the northeastern United States and the Emissions Trading System (“ETS”) in the European Union.

II. Lessons Learned from the Regional Greenhouse Gas Initiative

The CAT was not the first emissions marketplace in the United States. In 2009, the RGGI went into effect as a cap-and-trade marketplace for CO2 emissions in the following nine states: Connecticut, Delaware, Maine, Maryland, Massachusetts, New Hampshire, New York, Rhode Island, and Vermont.[10] However, the RGGI has been plagued with numerous shortcomings that have frustrated the performance of the initiative and which impart several lessons on how to more effectively design a cap-and-trade system.

A. Lesson 1: Cap-and-Trade Programs Need a Broad Scope

A key drawback of the RGGI is its limited scope. The program applies exclusively to CO2 emissions and only covers electrical power plants with the capacity to generate twenty-five or more megawatts.[11] Predictably, the results of the RGGI have been underwhelming, as only 163 facilities fall under the regulatory reach of the program.[12] Furthermore, CO2 emissions merely account for twenty percent of the GHG emissions in the nine participant states—a number that shrinks even further since the RGGI only regulates the electrical sector.[13] This narrowed scope has undermined the efficacy of the RGGI so drastically that Congress considers the program’s contribution to global GHG reductions to be “arguably negligible.”[14]

B. Lesson 2: Emission Forecasts Must Be Accurate

The second significant failing of the RGGI was that it overestimated the amount of CO2 emissions among the member states.[15] In fact, the RGGI set an initial emissions cap that was above actual emissions levels.[16] This was a gross oversight that stemmed from two key defects in the RGGI’s design.

First, the RGGI emission limits for the first cap period, which ran from 2009–2013, were based on emission estimations made in 2005.[17] Between 2005 and 2009, the amount of electricity generation in the member states decreased by thirty-six percent due to energy efficiency improvements and structural changes in energy generation portfolios.[18] Second, the RGGI distorted its emission forecasts by including all electrical power plants that had the capacity to generate twenty-five or more megawatts in its estimates.[19] Limiting the emission calculations to power plants that actually generated twenty-five or more megawatts would have produced more accurate projections.

These errors have been catastrophic for the initiative. The initial regulations had no effect on most businesses, which were already emitting below the inflated emissions cap.[20] Participation in the RGGI was therefore minimal, since many of the targeted businesses had no need to reduce emissions, purchase allowances, or generate offset credits.[21] Furthermore, because the RGGI does not limit the amount of allowances that can be “banked” and used in subsequent years, many companies have stored substantial amounts of these initial surplus allowances for future use.[22]

The administrators of the RGGI have taken extreme measures to try and remedy these miscalculations. Most notably, they implemented a “revised emissions cap,” running from 2014–2020, that slashes the emission limits by forty-five percent in an effort to match actual emission levels.[23] Such radical action would not have been necessary if the initial emissions cap had been more precise.

C. Lesson 3: Auctions Need Robust Price Floors

A final pitfall of the RGGI is its undervalued price floor for auctions. The reserve price has hovered around two dollars per allowance, despite being scheduled to increase according to the Consumer Price Index (“CPI”).[24] But the fact that auctioned allowances have been sold at prices exceeding five dollars indicates that businesses are willing to pay more.[25] The program therefore severely underappreciated the corporate demand for allowances and forfeited substantial potential earnings. Moreover, by greatly undervaluing the price floor, the RGGI administrators neglected to protect against suboptimal years when allowance prices have plummeted. A higher reserve price would have preserved the revenue generation capacity of the program, even during these off years.[26]

III. Lessons Learned from the European Union’s Emission Trading System

There are also numerous lessons to be learned from the deficiencies of the European Union’s ETS, which is the world’s largest market-based mechanism for reducing GHG emissions.

A. Lesson 1: Cap-and-Trade Programs Need Ambitious Initial Targets

At the conclusion of Phase I of the ETS, the “Learning Phase” that ran from 2005–2007, it was apparent that the initial targets for emission reductions were far too lenient.[27] Indeed, the lax regulations during Phase I only produced GHG reductions of three percent.[28] The EU was forced to compensate by crafting extreme targets for Phases II and III of the program, setting emissions goals of six percent below 2005 levels for Phase II and twenty-one percent below 2005 levels for Phase III.[29] If the EU had formulated a more ambitious target for Phase I rather than over-prioritizing the transition of members into the program, it would have avoided the need for these drastic adjustments.

B. Lesson 2: Allowances Must Be Apportioned Judiciously

Similar to the RGGI, the ETS grossly over-allocated emission allowances. In fact, ETS allowances initially exceeded the amount of actual emissions by four percent.[30] This miscalculation was devastating for Phase I of the ETS, as it enabled European businesses to emit 130 million tons more in GHGs than they had emitted prior to the implementation of the program.[31] This surplus destroyed the demand for allowances in the ETS marketplace, and auction prices fell precipitously.[32] The EU was forced to heavily reconfigure ETS allowance allocations to try and mitigate the damage caused by these initial overestimations, and it is still attempting to normalize the ETS marketplace.[33]

C. Lesson 3: Cap-and-Trade Programs Need Balanced Market Designs

The ETS has also been hamstrung by its inferior market design. Phase I of the program did not permit any allowances to be banked for future use.[34] Coupled with the initial over-allocation of allowances, this meant that most regulated entities possessed surplus allowances they had to expend by the year-end. This resulted in extreme downward price volatility at the conclusion of trading periods, as many companies attempted to dump the remainder of their emission allowances into the auctions.[35] The EU was once again forced to implement significant revisions to correct this oversight.[36] And while the ETS now permits allowances to be banked, the initial trading instability across Europe nearly destroyed the program.[37]

The EU also does not set a reserve price for ETS auctions, meaning there is no price protection for emission allowances.[38] This remains a gross oversight by the EU, as the lack of a price floor fails to account for the inevitable fluctuation of allowance prices due to changes in weather or energy price cuts. As a consequence, the ETS has lost significant revenue during periods of low auction demand where allowances have sold for pennies on the dollar, and the program will continue to be financially vulnerable until this design flaw is remedied.[39]

D. Lesson 4: Cap-and-Trade Programs Need Administrative Uniformity

Administrative inefficiencies have also plagued the ETS. The most glaring hole was the initial lack of a single registry for ETS participants.[40] Prior to 2012, each nation participating in the ETS had its own registry, which resulted in inconsistent regulation across the system.[41] The Danish registry, for example, failed to vet its registrants for two years.[42] The registry ultimately became so saturated with fraudulent companies that over ninety percent of account holders had to be deleted in 2010.[43] Even after the EU moved all participants into a single registry, the credibility lost among consumers during these initial years continues to plague the reputation of the program.

E. Lesson 5: Cap-and-Trade Programs Need Strong Cyber-Security

The final shortcoming of the ETS is that its cyber-security has been extremely assailable. “Phishing” has been one particularly vexing problem. The scam involves the creation and promotion of fake registries that solicit users to reveal their ETS identification codes. The “phishers” then use this information to carry out carbon trading transactions in legitimate registries. These deceptions have had severe economic ramifications, and as much as three million euros have been stolen in a single month.[44]

Hacking has been another key cyber-security issue for the ETS. Hackers have been able to infiltrate users’ computer systems and sell off all their allowances for immediate cash payments on the “spot market.”[45] Numerous companies have been crippled by this scam, and hackers have defrauded certain businesses of more than seven million euros worth of emission allowances.[46]

IV. The Success of the California Cap-and-Trade Program

When considering the numerous oversights of the RGGI and ETS programs, the success of the CAT is doubly impressive. This success is due to the balanced design of the CAT, which incorporates the strengths of the RGGI and ETS while mitigating their weaknesses.

A. Success 1: The CAT Has Precise Methods for Accurately Allocating Allowances

Both the RGGI and ETS erred by overestimating actual emission levels and allocating excessive allowances. The CARB avoided this mistake by crafting a precise allocation methodology that prevented surplus allowances from derailing the auction marketplace. Foremost, the CARB calculated California emission levels for the years immediately preceding the creation of the CAT to more accurately forecast future emissions. The CARB also narrowed the variability of its emissions estimates by only including emitters who had actually emitted 25,000 or more metric tons of CO2 or equivalents.[47] Emitters who merely had the capacity to emit beyond the 25,000 metric ton threshold were not included in the calculations. The greater accuracy of the CAT estimates was evidenced during the program’s first quarterly auction in 2012, where all twenty-three million allowances offered at the auction were purchased above the reserve price.[48]

B. Success 2: The CAT Began Ambitiously While Also Facilitating Transition

Another common error of the RGGI and ETS was that their design strategies over-prioritized transitioning members into their systems. The programs initially neglected to implement substantive emission reduction targets for fear of overwhelming participants, and they have subsequently instituted dramatic reforms to compensate. By contrast, the CARB recognized the need to balance the transition of members into the program against regulatory efficacy, lest one derail the other.

The CARB facilitated the transition of participants into the CAT by narrowing the scope of the first compliance period to only cover electrical and industrial sectors. It waited until the second compliance period to expand into the transportation and heating fuel sectors to provide companies time to adjust their business practices.[49] Yet the CARB also implemented considerable GHG reduction targets. The CARB initially set a 2020 reduction goal of seventeen percent below 2013 levels, which still eclipses the target of the RGGI.[50] Due to these ambitious benchmarks, the CAT has already produced “non-negligible” emission reductions and economic gains, with 2013 alone seeing GHG reductions of over a million and a half metric tons and statewide economic growth of two percent.[51] The CAT has benefitted greatly from such a stable infrastructure, and it remains on track to reach its ultimate emission reduction target by 2020.[52]

C. Success 3: The CAT Has a Broad Scope

The CARB also built off the mistakes of the RGGI by broadening the regulatory scope of the CAT. Because it only regulates CO2 emissions, the RGGI covers less than twenty percent of the GHG emissions generated across its nine participating states.[53] By contrast, the CAT emulates the ETS by also covering CO2 equivalents such as CH4, N2O and other fluorinated GHGs, resulting in more effective emission restrictions.[54] The CARB also recognized that the RGGI erred in solely regulating electrical power plants. Accordingly, the CARB extended CAT regulations into other sectors heavy in GHG emissions, such as industrial, transportation, and heating fuel sectors.[55] Because of this broader scope, the CAT already covers over 600 facilities in California, whereas the RGGI only reaches 163 facilities across nine states.[56] The CAT also covers more than eighty-five percent of California’s GHG emissions, which is almost four times the amount of GHG coverage under the RGGI.[57]

D. Success 4: The CAT Has a Balanced Market Design

The CAT also avoided the severe design blunders of the RGGI and ETS. Rather than undervaluing or ignoring auction price floors, the CARB instituted a strong reserve price of ten dollars in 2012, which has been set to increase each year thereafter by five percent (in addition to increases for inflation).[58] Allowances have consistently sold above these amounts, but the price floor has provided steady protection against downward price volatility during poor trading periods.[59] Moreover, the built-in mechanism for annual increases to the reserve price has ensured that the price floor continues to increase irrespective of CPI circumstances.[60]

The CAT further protects against precarious price drops by permitting allowances to be banked.[61] This avoids the price instability problems of the ETS by discouraging businesses from dumping surplus allowances into auctions at the end of trading periods. Nevertheless, the CAT imposes limits on the maximum amount of allowances that can be held by a business.[62] This circumvents the design flaw of the RGGI that allows businesses to bank an inordinate amount of allowances and eliminate any need to subsequently reduce emissions.[63]

The revenues generated by the CAT best demonstrate the success of its market design. The first auction raised more than $289 million, and the first compliance period generated $969 million in revenue for California.[64] Projections estimate that the CAT will generate two billion dollars or more per year as the program’s regulatory scope continues to scale upwards.[65]

E. Success 5: The CAT Has Strong Administrative and Security Practices

The CAT has also benefitted immensely from its efficient administration and strong security practices. Foremost, the CAT keeps a single registry for all its regulated entities, ensuring vigilant and orderly monitoring of all participants.[66] The cyber-security protocols of the CAT have been extremely successful as well.[67] To prevent hackers and phishers from infiltrating the program, CAT auctions take place over a four-hour window that is constantly supervised by state employees.[68] The bidders and supervisors remain undisclosed to the public, and all parties must surrender their electronic devices during the auction.[69] This “sealed bid” approach to the auctions has protected the CAT from the fraud and counterfeiting issues that tormented the RGGI and ETS.[70]

V. A Recent Legal Challenge: Are Cap-and-Trade Auctions Tax Programs?

Despite the success of the CAT, the program has faced serious legal obstacles. The principal challenge took place in the recent Morning Star Packing Company v. California Air Resources Board case, where the plaintiffs alleged that the auctions were unconstitutional and violated California law.[71] The chief contention was that the CAT constituted a tax on companies for emitting GHGs.[72] The plaintiffs argued that the statutory authorization of the CAT, the Global Warming Act, therefore fell under the purview of California’s Proposition 13, which requires legislators to pass by two-thirds vote “any act to increase state taxes for the purpose of increasing revenue.”[73] Because the Global Warming Act was not passed by a two-thirds vote, the plaintiffs asserted that the CARB exceeded its regulatory authority when it created the CAT.[74]

The dispositive issue in the case was whether the auctions were unconstitutional taxes or whether they were permissible regulatory fees placed on tradable commodities.[75] The Sacramento superior court ultimately upheld the CAT, concluding that emission allowances were tradable commodities in a marketplace.[76] The court considered several distinctions between taxes and regulatory fees, but the chief difference seemed to be that whereas the government sets tax prices, the market determined the auction price of the emission allowances.[77] Thus, the fact that the allowances had no value independent of the California regulatory scheme did not transform the auctions into a tax program, and the allowances remained tradable commodities.[78]

Yet the superior court ruling did not mark the end of the contentious litigation. The Morning Star decision was appealed to the Sacramento appeals court, which affirmed the lower court judgment by a two-to-one majority decision.[79] In turn, the appellate court ruling was appealed to the California Supreme Court, which ultimately declined to hear the case in June of 2017.[80] What should have been a resounding victory, however, was diminished by the fact that the State Supreme Court did not issue a written opinion on the program itself.[81] Nevertheless, the affirmation of the CAT provided market-based environmentalism with a new lease on life and has galvanized California policymakers and legislators.

VI. The Aftermath of Morning Star

The ramifications of the Morning Star have already been substantial in California. State legislators quickly capitalized on the State Supreme Court’s dismissal of the case by voting to extend the CAT an additional ten years through 2030.[82] The extension produced newfound confidence in environmentalism and revitalized the market economy surrounding the CAT – whereas previous quarterly auction sales had dropped sharply, the California government sold every emission permit offered in the August 2017 auction.[83]

Yet these successes have not been replicated on a national scale. This is somewhat perplexing, as the CAT provides a workable model upon which to base the creation of a federal cap-and-trade program. In particular, Congress could convincingly argue that the Morning Star case supports the notion that cap-and-trade programs deal with tradable commodities and do not constitute tax programs. Congress could therefore avoid having to rely on the Taxing and Spending Clause of the Constitution to justify the creation of an auction program and, instead, could derive its authority from the broader powers of the Commerce Clause.

The affirmation of Morning Star also provides strong persuasive reasoning for Congress to resolve the longstanding debate on whether emission allowances are “physical” (or “nonfinancial”) commodities, which are physically deliverable and consumable, or “financial” commodities that are satisfied through cash settlements.[84] Relying upon the Morning Star court’s description of allowances as being consumable and involving the physical transfer of title, Congress now has a strong basis for asserting, on the federal level, that allowances are physical commodities.[85] This would shield a federal cap-and-trade program from the administrative burdens of complying with the Commodity Exchange Act and other commercial regulations. [86]

Despite the reasoning provided by Morning Star, recent federal policy has demonstrated a marked shift away from the environmentalist approach espoused by the Obama Administration. The recent withdrawal of the Clean Power Plan, the Obama-era rule regulating greenhouse gas emissions, best evinces this change in protocol.[87] Indeed, with the Environmental Protection Agency consistently the choice target of President Trump’s proposed budget cuts, environmentalism on a national level has been placed in a precarious position.[88]

It remains to be seen whether this federal paradigm shift will take a toll on the CAT. It is certain, however, that the demise of the CAT would be the death knell for market-based environmentalism in the United States. Fortunately, the CAT has several contingency protocols to counteract market volatility. In particular, the CARB can hold unsold allowances off the market for at least nine months to compress the supply and force participants back to the auctions.[89] This foresight proved to be invaluable in the wake caused by the initial Morning Star appeal in 2016, during which time the May 2016 and August 2016 auctions only sold eleven percent and thirty-five percent, respectively, of the allowances offered.[90] The remedial mechanisms built into the CAT allowed administrators to re-stabilize the market, and the November 2016 auction resulted in the successful sale of eighty-nine percent of the offered allowances.[91] Nevertheless, these contingencies are merely stopgap solutions, and hesitation among market participants will likely resurface as Californian and national policy progress along their collision course. Until a clear and unified path towards environmentalism is forged across the nation, an ominous shadow will remain cast over the CAT.

 VII. Conclusion

The CAT has been a landmark initiative for environmentalism in the United States. Incorporating lessons from the RGGI and ETS, the program has struck a masterful balance in its market design and has produced significant environmental and financial gains for California. The affirming decision of the California judiciary and recent expansion of the program by the California legislature have been beacons of hope for cap-and-trade. Despite these successes, the future of the CAT remains in doubt, plagued by an uncertain socio-political climate where federal support for environmentalism has recently waned. And while the CAT has withstood previous legal and economic challenges, it is undeniable that the decisive battle for market-based environmentalism across the United States has begun.


[1] California Environmental Protection Agency, Assembly Bill 32 Overview,

[2] Id.

[3] California Cap-and-Trade Program Summary, Center for Climate and Energy Solutions (Jan. 2014),

[4] Id.

[5] Id. From 2013–2015, the program covered electrical and industrial power plants that emitted 25,000 or more metric tons of CO2 or equivalent gases per year. Since 2015, fuel distributors have also been covered.

[6] Id.

[7] Id.

[8] Id. Carbon offsets are greenhouse gas emission reductions that are credited to a company that funds or participates in an activity that reduces carbon footprints in the environment.

[9] Id.

[10] Lucas Bifera, Regional Greenhouse Gas Initiative, Center for Climate and Energy Solutions 1 (Dec. 2013),

[11] Jonathan Ramseur, The Regional Greenhouse Gas Initiative: Lessons Learned and Issues for Congress, Congressional Research Service 2 (Apr. 27, 2016),

[12] Id.

[13] Id. at 3.

[14] Id. at 19.

[15] Id. at 3–7.

[16] Id. at 4.

[17] Id. at 4–5.

[18] Id. at 5.

[19] See id.

[20] Id. at 4–5.

[21] Id. at 3­–7.

[22] Overview of RGGI CO2 Budget Trading Program, Regional Greenhouse Gas Initiative 6 (Dec. 2007),

[23] Ramseur, supra note 12 at 7–8.

[24] Id. at 8–12.

[25] Id.

[26] Id.

[27] Emissions Trading in the European Union: Its Brief History, Pew Center on Global Climate Change 1–2 (Mar. 2009),

[28] Id.

[29] Id.

[30] Tamra Gilbertson, Fraud and Scams in Europe’s Emissions Trading Systems, Climate & Capitalism, May 5, 2011,

[31] Id.

[32] Id.

[33] See id.

[34] Emissions Trading in the European Union, supra note 28 at 1–2.

[35] Id.

[36] Id.

[37] Id.

[38] Flawed Application of the Auction Reserve Price in the EU ETS, (Feb. 23, 2013),

[39] Gilbertson, supra note 31.

[40] Id.

[41] Id.; Union Registry, European Commission, (last visited Feb. 17, 2017).

[42] Gilbertson, supra note 31.

[43] Id.

[44] Id.

[45] Id.

[46] Id.

[47] California Cap-and-Trade Program Summary, supra note 4.

[48] Dana Hull, 13 Things to Know About California’s Cap-and-Trade Program, San Jose Mercury News (Feb. 22, 2013),

[49] California Cap-and-Trade Program Summary, supra note 4.

[50] Id.

[51] Dave Clegern, California greenhouse gas inventory shows state is on track to achieve 2020 AB 32 target, California Environmental Protection Agency (June 30, 2015),

[52] Id.; Michael Hiltzik, California’s cap-and-trade program has cut pollution. So why do critics keep calling it a failure?, L.A. Times (July 29, 2016),

[53] Ramseur, supra note 12 at 2.

[54] California Cap-and-Trade Program Summary, supra note 4.

[55] Id.

[56] Id.

[57] Id.; Emily Reyna, Four Reasons California Cap and Trade Had an Extraordinary First Year, Forbes (Jan. 14, 2014),

[58] California Cap-and-Trade Program Summary, supra note 4.

[59] Archived Auction Information and Results, California Environmental Protection Agency,

[60] California Cap-and-Trade Program Summary, supra note 4.

[61] Archived Auction Information and Results, supra note 60.

[62] California Cap-and-Trade Program Summary, supra note 4.

[63] Id.

[64] Hull, supra note 47; Michael Hiltzik, Emissions cap-and-trade program is working well in California, L.A. Times (June 12, 2015),

[65] Hiltzik, supra note 65.

[66] California Cap-and-Trade Program Summary, supra note 4.

[67] Laurel Rosenhall, Why hasn’t California’s cap and trade pollution program been the model for the U.S.?, L.A. Daily News (July 31, 2015),

[68] Id.

[69] Id.

[70] Id.; Gilbertson, supra note 31.

[71] Morning Star Packing Co., et al. v. California Air Resources Board, et al., Sacramento Superior Court, Case No. 34-2013-80001464 [hereinafter Morning Star Superior Court Ruling]. The case was consolidated and decided jointly with California Chamber of Commerce, et al. v. California Air Resources Board, et al., Sacramento Superior Court, Case No. 34- 2012-80001313. The joint decision is available at:

[72] Id. at 5.

[73] Id.

[74] Id.

[75] Id. at 11–14.

[76] Id. at 16–18.

[77] Id.; Allie Goldstein, Cap-and-Trade Is Not A Tax, California Court Says, Ecosystem Marketplace (Nov. 18, 2013),

[78] Goldstein, supra note 78.

[79] See generally Morning Star Appellate Decision.

[80] Dan Whitcomb, California Supreme Court Upholds Cap-and-Trade Law, CNBC (June 28, 2017),

[81] Id.; Chris Megerian, California Supreme Court Leaves in Place Decision Upholding Cap-and-Trade System, L.A. Times (June 28, 2017),

[82] Melanie Mason & Chris Megerian, California Legislature Extends State’s Cap-and-Trade Program in Rare Bipartisan Effort to Address Climate Change, L.A. Times (July 17, 2017),

[83] California Cap-and-Trade Program: Summary of Joint Auction Settlement Prices and Results, California Air Resources Board (Aug. 2017),; Chris Megerian, California Cap-and-Trade Program Gets Shot in the Arm with Strong Permit Auction, L.A. Times (Aug. 23, 2017),

[84] CFTC Glossary, United Statutes Commodity Futures Trading Commission,

[85] See generally Morning Star Superior Court Ruling.

[86] See, e.g., 7 U.S.C. § 1a(47)(B)(ii) (2012) (excluding from the definition of “swap” “any sale of a nonfinancial commodity or security for deferred shipment or delivery, so long as the transaction is intended to be physically settled”).

[87] Daniella Diaz et al., EPA Administrator Scott Pruitt Announces Withdrawal of Clean Power Plan, CNN (Oct. 10, 2017),

[88] Brady Dennis & Juliet Eilperin, EPA Remains Top Target with Trump Administration Proposing a 31 Percent Budget Cut, Washington Post (May 23, 2017),

[89] Hiltzik, supra note 53.

[90] Summary of Joint Auction Settlement Prices and Results, supra note 84.

[91] Id.

FERC Relicensing and Its Continued Role in Improving Fish Passage at Pacific Northwest Dams

By Skylar Sumner, a third-year J.D. student at Lewis & Clark Law School. 

This post is part of the Environmental Law Review Syndicate

I. Introduction

The history of the American west is inextricably intertwined with damming rivers.[1] Whether for navigation, irrigation, or hydroelectric power, nearly every American river has been dammed.[2] In fact, stretching back to the day the Founding Fathers signed the Declaration of Independence, determined Americans have finished an average of one large-scale dam every day.[3] Currently, there are at least 76,000 dams in this country.[4]

While these dams have vastly contributed to America’s efforts to settle the west, they have come with significant costs. Although these dams’ harms are varied,[5] one of the primary concerns among advocates in the Pacific Northwest is the dramatic impacts dams have on species of anadromous fish, particularly salmonids.[6] In the Columbia River basin, dams block salmon and steelhead migration to more than 55% of historically available spawning grounds.[7] Since many anadromous fish species in the Pacific Northwest are listed as either threatened or endangered,[8] the Endangered Species Act[9] (ESA) can be a valuable tool to induce voluntary dam removals by requiring the Federal Energy Regulatory Commission (FERC) to include costly fish passage upgrades in any relicensing proceeding.[10]

Northwest salmon advocates rejoiced in 2014 when, following a lengthy campaign from a coalition of tribal and environmental activist groups,[11] construction crews completed the largest dam-removal project in American history by removing both the Elwha and the Glines Canyon Dams.[12] Removing these dams started the process of restoring seventy miles of the Elwha River to natural flows that had not existed since construction of the dams first began in 1911.[13] Since the dams came down, the river’s ecological quality has improved at an astonishing rate.[14] In fact, salmon and steelhead populations in the Elwha River have already reached thirty-year highs.[15]

The tremendous success of freeing the Elwha cannot be overstated, but the dams required decades of activist toil to remove.[16] In contrast, removing the Little Sandy and Marmot dams from the Sandy River in Oregon was accomplished in only eight years.[17] There are certainly many core differences between these campaigns that help explain this discrepancy, but chief among these is the fact that Federal Power Act[18] (FPA) amendments incentivized the owner of the Little Sandy and Marmot dams to privately fund the removal, while the Elwha removal languished waiting on federal funding for over a decade.[19]

This Essay will discuss the statutory changes to the FERC relicensing process that have worked to improve fish passage at hydropower facilities in recent decades and will continue to fuel upgrades and dam removals in the future. Part II lays out an overview of the environmental requirements of FERC relicensing and analyzes the Bull Run Hydropower Project as an example of a successful dam removal that was prompted as a result of its owner pursuing relicensing. Part III then reviews the relicensing schedule for several dams in Oregon and Washington to discuss how these fish passage improvements will continue occurring for the foreseeable future.

II. FERC’s Current Statutory Requirements Will Improve Fish Passage at Hydroelectric Facilities.

The current regulatory process will—at least marginally—improve fish passage at many hydropower facilities in the near future as older dams apply for relicensing through FERC. Privately operated hydroelectric dams can only operate under a license from FERC.[20] For older dams, the cost of installing fish passage during the FERC relicensing process can exceed the cost of removal, thereby incentivizing the dam owner to opt for removal.[21] For dams that successfully obtain a license to continue operation, the current statutory relicensing framework requires FERC to include any recommended fish passage upgrades as mandatory conditions in the license.[22] Due to new environmental statutes and regulations passed during the lifetime of the preceding license, many hydroelectric dams in the Columbia River basin are likely to require passage upgrades.[23]

FERC is in the midst of a massive relicensing period.[24] The FERC relicensing process has had a tremendous impact on fish passage in the Columbia River basin in recent history, as both Oregon and Washington were included in FERC’s list of states requiring the most dam relicenses between 2005 and 2015.[25] As discussed below, absent a congressional amendment of the FPA, the FERC relicensing process will mandate fish passage upgrades at Northwest hydroelectric facilities for decades to come.

A. The FERC Licensing Process 

In 1920, Congress passed the FPA, authorizing the federal government to regulate private hydroelectric dams.[26] While older dams may have been constructed without a FERC license,[27] all dams must eventually obtain a license to continue operation.[28]

Initially, FERC only considered a dam’s power-generation potential when reviewing a license application, while ignoring the environmental impacts.[29] Then in 1986, Congress amended the FPA[30] to require FERC to include permit conditions protecting fish and wildlife.[31] Now, FERC licenses “require the construction, maintenance, and operation by a licensee at its own expense of such . . . fishways as may be prescribed by” the United States Fish & Wildlife Service or the National Oceanic and Atmospheric Administration (NOAA) Fisheries.[32] FERC cannot “modify, reject, or reclassify any prescriptions submitted by” those agencies.[33] If FERC disagrees with the fish passage conditions, FERC must either withhold the license or dispute the conditions before the relevant court of appeals.[34]

New FERC permits may last for a duration of up to fifty years.[35] Due to this timeframe, FERC will spend the foreseeable future considering relicensing applications for dams whose original permits were approved with minimal environmental consideration. For instance, FERC will review relicensing applications for dams that were approved without an Environmental Impact Statement (EIS) through 2020,[36] dams that were approved without wildlife permit conditions through 2036,[37] and dams that were approved prior to Endangered Species Act protections for anadromous fish through the 2040s.[38]

When owners of these dams apply for relicensing, modern environmental and endangered species protections will likely require project owners to significantly upgrade the dams’ fish passage facilities. FERC has proven willing to attach extremely costly fish passage conditions to its relicensing decisions, which can make removal the most cost-effective next step for hydroelectric dam operators.[39] For those dams that remain standing, new FERC licenses will still likely improve fish passage because relicensing will be conditioned upon upgrading fish passage to meet modern environmental and ESA requirements.[40]

B. Bull Run Hydropower System: An Example of How FERC Relicensing Provides Strong Incentive for Voluntary Dam Removal Settlement

The FERC relicensing process has proven to be an effective tool in persuading operators of large hydroelectric dams to negotiate effective and efficient dam removals that are entirely funded by the dam operators. Few cases highlight how well this process can facilitate dam removals better than the Marmot and Little Sandy dams of the Bull Run Hydropower Project. The Bull Run project is the gold-standard in dam removal for many reasons, including 1) it was entirely funded by the operator without predetermined cost caps;[41] and 2) the dams came out quickly, with minimal confrontation between the affected parties.[42]

Twenty-six miles east of Portland, Oregon the Bull Run River flows through the Mt. Hood National Forest.[43] The Bull Run River drains a 102 square-mile watershed and is almost entirely fed by rain and snowmelt from Mt. Hood.[44] As the main source of water for Portland, the Bull Run watershed provides tap water for nearly one-fifth of all Oregonians.[45] Development on the Bull Run began in the 19th century,[46] and the river became an important source of both water and electricity for the surrounding area.[47]

In 1912, Pacific Gas & Electric (PGE) completed the primary stage of one of the largest developments in the watershed: the Bull Run Hydropower Project.[48] To increase the powerhouse’s capacity, PGE constructed the Little Sandy Dam to divert water from the Little Sandy River to Roslyn Lake, the reservoir behind the project’s powerhouse.[49] The dam completely diverted the Little Sandy River 1.7 miles upriver from its confluence with the Bull Run River.[50] The dam blocked salmon migration upstream and decreased flows to the remaining salmon habitat downstream.[51]

The following year, PGE completed the Marmot Dam on the Sandy River.[52] This dam diverted water from the mainstem Sandy River to the Little Sandy upstream from the Little Sandy Dam, thereby increasing the capacity at Roslyn Lake.[53] The original Marmot Dam was a wood and sediment structure.[54] Unlike the Little Sandy Dam, the Marmot Dam did not block all salmon migration because the original structure included a fish ladder.[55] In 1989, PGE replaced the original Marmot Dam with a forty-seven foot concrete dam.[56]

The Bull Run Hydropower Project’s dams and diversions decreased fish runs in the Sandy River and Bull Run watersheds to 10%–25% of their historic runs.[57] PGE, the operator of the Marmot and Little Sandy Dams, operated four hydroelectric systems that would all require FERC relicensing in the early 2000s.[58] Due to the increasing burden of maintaining century-old dams, relatively low summer flows, and modern environmental regulations,[59] PGE determined that the Bull Run Hydropower System’s costs were simply insurmountable.[60] PGE chose to voluntarily surrender its FERC license.[61] After negotiating a settlement agreement with all affected parties,[62] FERC granted PGE’s petition to surrender its license in 2004.[63] Because of the inclusive settlement process,[64] public support for the final project was high, and PGE obtained all necessary environmental permits to move forward with the dam removal in only eighteen months.[65]

On July 24, 2007, engineers began the process of removing the Marmot Dam by setting off explosives to crack the concrete face.[66] The process ended that October with the breach of a temporary diversion dam built just upstream.[67] At the time, this was the largest dam removed in the Pacific Northwest, both in terms of height and trapped sediment.[68] The Sandy River recovered much more rapidly than expected, with migrating coho salmon reported swimming past the old dam site just one day after engineers completed the removal process.[69] The Little Sandy Dam was removed the following summer.[70]

An important takeaway from the Bull Run Hydropower Project’s removal is that, under the right circumstances, environmental conditions placed on FERC relicensing approvals can act as a tremendous hammer to force dam removals. In fact, PGE decided to pursue settlement negotiations before it even received the final fish passage requirements.[71] Preliminary estimates were enough for PGE to determine that the Bull Run system would not be economical. The Bull Run removal process shows just how effectively the FERC regulatory process can trigger rapid dam removals with minimal delays and no public funding.

III. The Glut of Pending and Upcoming License expirations Will Require FERC to Revisit Fish Passing in the Pacific Northwest for Several Decades. 

Because of the fifty-year lifetime of its licenses, FERC is currently in the process of relicensing the final pre–National Environmental Policy Act[72] (NEPA) hydroelectric dams.[73] Several dams in both Washington and Oregon are still operating under such licenses.[74] Although the relicensing process has proceeded slowly, one certainty is that fish passage upgrades will be a mandatory condition for almost any new FERC license. This Part discusses a few dams in both Northwest states that are scheduled for relicensing in the coming decades and provides contemporary examples of the fish passage upgrades that FERC has already required at Northwest dams in recent years.

A. Washington Dam Relicensing

FERC currently licenses fifty-five privately operated hydroelectric dams in Washington.[75] Two of these dams—Sullivan Lake and Packwood Lake—were licensed prior to the mandatory environmental review process codified in NEPA.[76] The Packwood Lake dam, for example, was last licensed in July 1960.[77]

Packwood’s initial license was set to expire in 2010, but the dam has been operating under annual interim permits while working to determine what mandatory conditions will attach to the final new license.[78] As part of this relicensing process, Energy Northwest—the operator of Packwood Dam—has had to cooperate with NOAA Fisheries to determine the impact that the dam’s continued operation will have on listed species.[79] NOAA Fisheries found that three listed species were likely to be affected by the dam’s operation: Lower Columbia River Chinook, coho, and steelhead.[80] To mitigate these harms, Energy Northwest has built an exclusionary screen to keep migrating salmonids out of the channel leading to the powerhouse,[81] but more expansive requirements may be included before FERC can issue the final license.[82]

Along with the pre-NEPA dams, FERC also oversees seventeen dams that are operating under licenses issued prior to the Electric Consumers Protection Act and, as such, did not require any wildlife considerations.[83] These dams will be pursuing relicensing through the 2030s, which will inevitably mandate new fish passage conditions, thereby improving salmonid accessibility to spawning grounds.[84]

B. Oregon Dam Relicensing

Of the twenty-five actively licensed dams in Oregon,[85] there are three dams operating under pre-NEPA licenses: the Klamath, Hell’s Canyon, and Carmen-Smith dams.[86] The greatest fish-passage improvements will occur in the Klamath River, where PacifiCorp—the dams’ owner—has agreed to remove four huge dams by 2020, opening up 570 miles of riparian habitat for returning salmon.[87] Under the agreement, PacifiCorp will provide $200 million for the removal, and the state of California will fund up to an additional $250 million by selling general obligation bonds.[88]

On top of this monumental dam removal, the Carmen-Smith dam near Eugene, Oregon also agreed to significant improvements for salmon in order to relicense.[89] The Carmen-Smith license was issued in 1959 and expired in 2008.[90] As part of its relicensing effort, the Eugene Water and Electricity Board (EWEB) entered into a settlement agreement with sixteen other parties consisting mainly of government agencies, Native American Tribes, and environmental organizations.[91] This agreement included extensive salmonid habitat enhancements and a fish passage–system upgrade.[92] However, a precipitous decline in utility prices triggered a renegotiated agreement, and the fish passage upgrade was replaced with a trap-and-haul system to transport the fish around the dam’s powerhouses.[93] The parties submitted this amended agreement to FERC in 2016.[94] However, should NOAA Fisheries find this trap-and-haul system insufficient to protect the listed species, then EWEB could still be required to install the original fish passage upgrades.[95]

In addition, FERC oversees seven additional dam licenses that were approved prior to the Electronic Consumers Protection Act.[96] The last of these licenses expires in 2039.[97]

IV. Conclusion 

Dam removals have become much more common in recent decades, and FERC relicensing has played a large role by requiring expensive fish-passage upgrades as a mandatory condition of an extended operating license. This uptick in FERC-triggered removals was caused by the fact that many of the last dams to be licensed without any environmental oversight have sought relicensing in the past decade. While almost all the pre-NEPA dams have been relicensed at this point, FERC relicensing will continue to trigger fish passage upgrades at facilities that were originally licensed before FERC started attaching mandatory wildlife considerations in 1986. Organizations operating dams in the Pacific Northwest that were licensed prior to these wildlife conditions will be pursuing relicensing through 2039.

In some cases—like the Little Sandy and Marmot Dams in Oregon—the economic cost of the Electronic Consumers Protection Act’s fish passage requirements will exceed the benefit of continued operation and make removal the more cost-effective option. In most other cases, the new FERC license will still mandate fish passage upgrades like installing a fish-ladder or implementing a trap-and-haul system. Through either dam-removal or upgrades, these FERC conditions will improve fish-passage at hydroelectric dams throughout the Pacific Northwest.

[1] U.S. Army Corps of Eng’rs, Water in the U.S. American West 6 (2012).

[2] Id. at 14.

[3] Address, Bruce Babbitt, Sec’y of the Interior, Remarks at the Ecological Society of America Annual Meeting (Aug. 4, 1998),

[4] Heinz Center, Dam Removal: Science and Decision Making 3 (2002) (the list referenced here has not been updated since 2001 due to post-9/11 security concerns).

[5] See Christopher Scoones, Let the River Run: Strategies to Remove Obsolete Dams and Defeat Resulting Fifth Amendment Taking Claims, 2 Seattle J. Envtl. L. 1, 2 (2012).

[6] See Laurie A. Weitkamp, A Review of the Effects of Dams on the Columbia River Estuarine Environment, With Special Reference to Salmonids 6 (1994).

[7] John Harrison, Dams: Impacts on Salmon and Steelhead, N.W. Power and Conservation Council (2008),

[8] See, e.g., Wash. State Recreation and Conservation Office, Salmon Species Listed Under the Federal Endangered Species Act (2009),

[9] Endangered Species Act of 1973, 16 U.S.C. §§ 1531–1544 (2012).

[10] Margaret B. Bowman, Legal Perspectives on Dam Removal, 52 BioScience 739, 741 (2002).

[11] Julia Guarino, Tribal Advocacy and the Art of Dam Removal: The Lower Elwha Klallam and the Elwha Dams, 2 Am. Indian L. J. 114, 130–31 (2013).

[12] Elwha River Restoration: Freeing a River, Nat’l Park Serv., (last visited Sept. 30, 2017).

[13] Lower Elwha Klallam Tribe, Timeline of the Elwha River Dams & Removal Efforts, (last visited Sept. 30, 2017).

[14] Lynda V. Mapes, Elwha: Roaring Back to Life, Seattle Times (Feb. 13, 2016), (Scientists have been “amazed at the speed of change under way in the Elwha.”).

[15] Id.

[16]Lower Elwha Klallam Tribe, supra note 13.

[17] Michael C. Blumm & Andrew B. Erickson, Dam Removal in the Pacific Northwest: Lessons for the Nation, 42 Envtl. L. 1043, 1069–71.

[18] 16 U.S.C. §§ 791–825.

[19] Philip M. Bender, Restoring the Elwha, White Salmon, and Rogue Rivers: A Comparison of Dam Removal Proposals in the Pacific Northwest, 17 J. Land Res. & Envtl. L. 189, 228 (1997).

[20] 16 U.S.C. § 797(e) (2012).

[21] See, e.g., Blumm, supra note 17, at 1053–54 (discussing the relicensing process for the Elwha and Glines Canyon dams).

[22] 16 U.S.C. § 811.

[23] See infra notes 36–38 and accompanying text.

[24] 2007 was the peak year for hydroelectric relicensing. Applications for New Licenses (Relicenses), Fed. Energy Reg. Commission (Aug. 15, 2017),

[25] Id.

[26] 16 U.S.C. § 797(e).

[27] Congress did not authorize the federal government to license private dams built before June 10, 1920. Id.

[28] Id.

[29] Federal Power Act, Hydropower Reform Coalition (2017),

[30] Electric Consumers Protection Act of 1986, Pub. L. No. 99-495, 100 Stat. 1243 (codified at 16 U.S.C. § 791a).

[31] 16 U.S.C. § 803(j).

[32] Id. § 811.

[33] Am. Rivers v. Fed. Energy Regulatory Comm’n, 201 F.3d 1186, 1210 (9th Cir. 1999).

[34] Id.

[35] 16 U.S.C. § 799.

[36] National Environmental Policy Act of 1969, 42 U.S.C. §§ 4321–4347. NEPA was signed into law in 1970. What is the National Environmental Policy Act?, Envtl. Protection Agency, (last visited Sept. 30, 2017).

[37] Wildlife considerations were required in the Electricity Consumers Protection Act, enacted in 1986. 16 U.S.C. § 803(j).

[38] For example, Oregon coastal coho salmon were not listed until 1998. See, e.g., ESA Chronology for Oregon Coast Coho, Nat’l. Oceanic & Atmospheric Admin. Fisheries (last visited Sept. 30, 2017).

[39] For example, FERC would have required PacifiCorp to spend over $30 million on fish passage upgrades to relicense the Condit Dam, so PacifiCorp chose to remove the dam at a cost of approximately $17 million. David H. Becker, The Challenges of Dam Removal: The History and Lessons of the Condit Dam and Potential Threats from the 2005 Federal Power Act Amendments, 36 Envtl. L. 812, 826–27 (2006).

[40] 16 U.S.C. § 811.

[41] Blumm, supra note 17, at 1070.

[42] Becker, supra note 39, at 832 n.135.

[43]Bull Run Watershed, City Portland, (last visited Sept. 30, 2017).

[44] Id.

[45] Janie Har, Bull Run Watershed: Journey to the Source of Portland’s Copious, Constant Water, Oregonian (Aug. 13, 2010),

[46] The City of Portland first diverted water from the Bull Run in 1894. Andrew Theen, From Bull Run to Mount Tabor: The History of Portland’s Open Reservoirs (Timeline), Oregonian (Dec. 17, 2014),

[47] Bull Run: The Town That Time Forgot, PDX Hist. (Oct. 28, 2016),

[48] The main powerhouse was completed in 1912. The Century-Old Bull Run Powerhouse Finds New Life, Thanks to 3 Portland Preservationists, Oregonian (Dec. 6, 2012),

[49] Blumm, supra note 17, at 1067.

[50] Id.

[51] Blumm, supra note 17, at 1067.

[52] Id.

[53] Id.

[54] Id.

[55] Id.

[56] Id. at 1067–68.

[57] Id. at 1068.

[58] Of PGE’s four hydroelectric systems, the Bull Run project was the smallest. Julie A. Keil, Bull Run Decommissioning: Paving the Way for Hydro’s Future, Hydro Rev. (Mar. 1, 2009),

[59] The Bull Run system affected fish passage, temperature pollution, and river flows; several threatened fish species also migrated to the rivers. Id.

[60] This is understandable when you consider the fact that PGE would have had to upgrade two century-old dams just to continue electricity production at a single powerhouse. Id.

[61] Fed. Energy Regulatory Comm’n, Draft Environmental Impact Statement: Bull Run Project (2003).

[62] There were a total of twenty-two parties in the settlement. Id. PGE also agreed to pay all costs for the removal in the settlement, thereby circumventing the arduous process of securing federal funding. Blumm, supra note 17, at 1070.

[63] Portland Gen. Elec., Turbidity Management Plan: Bull Run Hydropower Project 1 (2005).

[64] Most notably, the nearest city—Sandy, Oregon—was a party to the settlement. Becker, supra note 39, at 832 n.135 (2006).

[65] Id.

[66] Marmot Dam, Oregon’s Largest Dam, Is Being Removed: Salmon and Wildlife Habitat and Public Recreation to Benefit, Horizon Int’l Sols. Site, (last visited Sept. 30, 2017).

[67] Jon Major et al., Initial Fluvial Response to the Removal of Oregon’s Marmot Dam, 89 Eos 241, 241 (2008).

[68] Id.; Charles Podolak & Jon Major, An Example of One River’s Response to a Large Dam Removal (2016),

[69] Elizabeth Brink, Feeding a Hungry River, 23 World Rivers Rev. 6, 6 (2008).

[70] Id.

[71] Blumm, supra note 17, at 1069.

[72] National Environmental Policy Act of 1969, 42 U.S.C. §§ 4321–4370h (2012).

[73] See supra notes 35–36 and accompanying text.

[74] Fed. Energy Regulatory Comm’n, Active Licenses (2017), (available for download) [hereinafter Active Licenses].

[75] Id.

[76] Id.

[77] Id.

[78] Fed. Energy Regulatory Comm’n, Pending Licenses, Relicenses and Exemptions (2017), (available for download).

[79] See, e.g., Nat’l Oceanic & Atmospheric Admin., Endangered Species Act Section 7 Formal Consultation, and Manguson-Stevens Fishery Conservation and Management Act Essential Fish Habitat Consultation for the License for Construction, Post-Construction Monitoring and Evaluation of a Tailrace Barrier at Packwood Lake Hydroelectric Project (FERC Project No. 2244) (2007), Since the last license was issued before Congress passed NEPA in 1970, these environmental reviews were never conducted before. Fed. Energy Regulatory Comm’n, Hydropower Primer: A Handbook of Hydropower Basics 20 (2017),

[80] Nat’l Oceanic & Atmospheric Admin., Packwood Lake Hydroelectric Project, (last visited Sept. 30, 2017).

[81] Id.

[82] See 16 U.S.C. § 811 (2012); see also supra notes 32–34.

[83] See Active Licenses, supra note 74, see also supra notes 30–31 and accompanying text.

[84] Active Licenses, supra note 74.

[85] Id.

[86] Id.

[87] See David N. Allen, The Klamath Hydroelectric Settlement Agreement: Federal Law, Local Compromise, and the Largest Dam Removal Project in History, 16 Hastings W.-N.W. J. Envtl. L. & Pol’y 428, 431–33 (2010).

[88] Id. at 459.

[89] Carmen-Smith Hydroelectric Project, Eugene Water & Electricity Bd., (last visited Sept. 30, 2017).

[90] Active Licenses, supra note 74.

[91] Christian Hill, EWEB Backs Deal to Save $80 Million on Dam Relicensing, Reg.-Guard, Nov. 2, 2016, at A1.

[92] Carmen-Smith Hydroelectric Project, supra note 89.

[93] Hill, supra note 91. For an illustration of the system, see Carmen-Smith Project: Upstream Fish Passage, Eugene Water & Electricity Bd., (last visited Sept. 30, 2017).

[94] Carmen-Smith Hydroelectric Project, supra note 89.

[95] Hill, supra note 91.

[96] Active Licenses, supra note 74.

[97] Id.

[ELRS] With Energy Law Federalism Under Construction, State Policymaking May Be Delayed

By John Bullock, Executive Editor, Harvard Environmental Law Review.*

This post is part of the Environmental Law Review Syndicate


As the public has become more aware of the intense connection between the practices of electric utilities and greenhouse gas emissions, interested groups have shone a brighter spotlight on the regulation of utilities in the United States. Some have called on the Federal Energy Regulatory Commission (“FERC”) to take on a more environmentally conscious role when exercising their authority to set wholesale rates.[1] While FERC still hasn’t explicitly taken environmental considerations into wholesale rate setting, it has taken steps to continue to ensure reliability as the nation’s energy portfolio composition shifts.[2]

Generally, under the Federal Power Act, FERC has jurisdiction over sales of electricity for resale in interstate commerce (wholesale sales), electricity transmission, and practices “affecting” rates.[3] The Supreme Court recently authorized a construction of FERC’s jurisdiction in FERC v. Electric Power Supply Association (“EPSA”) to include practices that “directly affect” wholesale rates.[4] This decision was seen as good for clean energy, as it removed barriers for demand response resources[5] to compete in the wholesale market in the short-term, while allowing FERC to have more regulatory flexibility in the long-term.[6]

At the state level, legislators and regulatory bodies generally retain the authority to set retail rates, maintain and site local facilities, and to establish resource portfolios.[7] There are a wide range of potential policies that can be used to foster clean energy, including feed-in tariffs,[8] renewable portfolio standards,[9] rebates for renewables,[10] a carbon tax,[11] a ban on carbon imports and new coal plant construction,[12] and net-metering policies.[13] A majority of states in the country have passed some form of a renewable portfolio standard mandating a certain percentage of the state’s electricity come from renewable resources.[14] These policies can originate in the state legislature or can come from the state utility regulator directly.[15] These state policies use several different regulatory tools, from market-based incentives like renewable energy credits to other state law mechanisms such as long-term power purchase agreements or mandated utility-owned renewable generation.

Some of these state clean energy policies have recently been challenged or are currently being challenged in the federal courts on preemption and dormant commerce clause grounds.[16] Challenges to these policies typically allege that the state programs are either preempted by the Federal Power Act, or are an impermissible intrusion into Congress’s exclusive power to regulate interstate commerce.

The Court, by authorizing an expansion of FERC’s jurisdiction in EPSA, and by failing to clarify the preemption analysis under the Federal Power Act in another recent case, Hughes v. Talen Energy Marketing LLC, may have inadvertently created considerable uncertainty about the extent of federal and state authority—or at least failed to remedy existing uncertainty. More thorough discussions on the shifting approach to the division of state and federal authority in energy law can be found elsewhere.[17] This Article will instead offer some speculation about the impacts of EPSA and Hughes on state policymaking.

FERC v. EPSA and Hughes v. Talen Energy Marketing

In Federal Energy Regulatory Commission v. Electric Power Supply Ass’n, the Supreme Court upheld FERC’s assertion of jurisdiction by allowing it to regulate practices that “directly affect” wholesale rates.[18] At issue in EPSA was whether FERC had authority to regulate demand response transactions (where a provider contracts with consumers to reduce energy consumption), or whether those transactions should be classified as “retail sales.”[19] The Federal Power Act grants FERC jurisdiction over practices affecting rates, and in EPSA, the Court adopted a D.C. Circuit test that cabined that authority to practices “directly affecting” rates.[20] After adopting the directly affecting test, the Court found that FERC had jurisdiction over demand response practices, that the rule did not impermissibly tread into authority reserved to the states, and that FERC did not act arbitrarily and capriciously in its decision to compensate electricity users at the same rates as electricity generators.

Whereas EPSA dealt primarily with the extent of FERC’s jurisdiction under the Federal Power Act, Hughes v. Talen Energy tackled the separate but related issue of whether a state program was preempted under the Federal Power Act.[21] The case was on review from the Fourth Circuit, where the appellate court found that a Maryland program was preempted both as a matter of field preemption (because FERC “occupies the field” of setting wholesale rates), and also as a matter of conflict preemption (because rates under Maryland’s program conflicted with FERC approved rates).[22] On review, the Supreme Court affirmed the lower court’s ruling, albeit on narrow grounds, finding that the Maryland program “impermissibly intrude[d] upon the wholesale electricity market, a domain Congress reserved to FERC alone.”[23]

One could argue that the Supreme Court narrowed the scope of the Fourth Circuit holding. For example, the Court distinguished between contracts-for-differences (which was the regulatory mechanism that Maryland deployed to encourage new natural gas plant development) and other more traditional long-term power purchase agreements.[24] However, in other ways, the Court’s opinion is actually more ambiguous—the Court does not clarify whether the correct analytical approach here should be conflict, field, or another form of preemption analysis,[25] and two Justices wrote concurring opinions to advocate for their distinct approaches.[26]

Because the opinion only addressed a narrow set of situations, the court did little if anything to address whether any other state regulatory mechanisms designed to encourage renewable deployment would be preempted under the Federal Power Act, and specifically limited their holding to Maryland’s program.[27] The decision provides no guidance on how to analyze these state law regulatory programs unless they contain contracts-for-differences that are pegged to a FERC-approved wholesale price, as Maryland’s program did. Therefore, the case is unlikely to act as a prophylactic to the litigation that is ongoing in the lower courts.[28] It makes one wonder why the Supreme Court took the case in the first place—there was no circuit split after the Fourth Circuit’s decision, and the Court failed to use the case as an opportunity to instruct the lower courts.

Putting Hughes and EPSA Together:
Examining Impacts on State Regulatory Authority

Combining the holding from EPSA with Hughes along with some of the more archaic language in previous energy preemption cases provide ample fuel for challenges to state renewable energy policies. Simply, if the Federal Power Act draws a jurisdictional “bright-line,”[29] or if “[i]t is common ground that if FERC has jurisdiction over a subject, then the States cannot have jurisdiction over the same subject,”[30] then any practice that “directly affects” wholesale rates should be exclusively within FERC’s jurisdiction. This could result in effectively shrinking state regulatory authority after EPSA and Hughes.

Still, the extent of practices that come within FERC’s “affecting” jurisdiction is unknown, and it may be that FERC must first exercise this jurisdiction over a particular practice before it has a preemptive effect. However, this doesn’t prevent litigants from making those arguments in the lower courts to invalidate clean energy programs, and Hughes may stand as a missed opportunity to clarify the scope of preemption under the Federal Power Act.

In fact, litigants are already citing Hughes and EPSA to challenge state clean energy programs. On October 2016, the Coalition for Competitive Energy filed a challenge to the New York Public Service Commission’s Clean Energy Standard in the Southern District of New York.[31] The Clean Energy Standard was issued in August,[32] and set a target for New York to obtain fifty percent of their electricity from renewable resources by 2030.[33] In addition to continuing New York’s renewable energy credit program,[34] the Clean Energy Standard included a requirement that load-serving entities purchase Zero-Energy Credits that correlate with electricity generated by nuclear facilities.[35] Coalition for Competitive Energy is challenging this specific program (the zero-emissions credits) in their complaint, alleging that it “operates within the area of FERC’s exclusive jurisdiction” and should therefore be preempted.[36] The petition cites EPSA to argue that “[s]tate actions that ‘directly affect the wholesale rate’” are invalid.[37]

Additionally, the Second Circuit recently granted Allco’s request for an injunction to prevent state officials from conducting a clean energy request for purchase (“RFP”) in Connecticut.[38] The decision did not enjoin state officials in Massachusetts and Rhode Island who are also participating in the RFP.[39] While the Second Circuit did not disclose their reasoning when it granted the injunction,[40] Allco’s petition for injunction pointed to Hughes when arguing that the program was preempted under the Federal Power Act.[41]

While it may seem that uncertainty in the preemption context is a net loss for individuals concerned about an accelerated transition to clean energy, climate advocates may also weaponize Hughes in other contexts to argue that other state polices that prop up coal and natural gas plants are preempted by the Federal Power Act. For example, the Ohio Public Utilities Commission recently attempted to use power-purchase agreements—which can sometimes be a tool to generate procure renewables[42]—to subsidize coal plants in the state.[43] The proposal was blocked by FERC before it could take effect,[44] but the program could have been challenged under Hughes if it remained in place.

Both examples citing to Hughes show challenges to state energy programs that operate outside of FERC-approved markets, unlike the Maryland program at issue in Hughes where the parties adjusted the FERC-approved rate.[45] Perhaps the biggest challenge going forward for clean energy advocates will be how to distinguish state programs that do not advance climate goals (like the Maryland program at issue in Hughes) from those that do (such as the program at issue in Allco), when both often use the exact same regulatory tools.

The Supreme Court may return to the question of the extent of federal and state authority under the Federal Power Act sometime within the next few years. It could reach one of several conclusions. It may reaffirm past language about the “bright-line” between federal and state regulatory authority—confirming that EPSA represented an expansion of FERC’s power and a simultaneous restriction on state authority. It may endorse some form of concurrent jurisdiction, as it did in the Natural Gas Act context in Oneok Inc. v. Learjet, Inc.,[46] and if it does, it may then decide how to restructure the preemption analysis under this concurrent jurisdictional model. It may establish some method of floor preemption,[47] or alternatively, it may leave the preemption decision up to the federal agency,[48] as it does in some other contexts.[49] Also, the Court may simply leave the resolution of these issues up to the lower federal courts.


Regardless of the approach the court takes, the fact that all of these questions remain open and unresolved currently creates considerable legal uncertainty for state regulators that are trying to update and craft effective clean energy laws. States are already testing the boundaries of their authority in many instances,[50] and many may continue to do so despite these new uncertainties. Further, it may be impossible to disaggregate the influence that legal uncertainty is having on state regulators from other influences such as political pressures. I would assume state legislators and regulators—some that are designing state laws to ensure their compliance with the Clean Power Plan—would likely prefer clarity on what regulatory mechanisms they are allowed to use without running afoul of the Supremacy Clause. Hughes thus represents a missed opportunity, and the recent power trio of Oneok, EPSA, and Hughes may shortly turn into a quartet.

* J.D. Candidate, Harvard Law School. The author would like to thank Ari Peskoe, Senior Fellow in Electricity Law at the Harvard Environmental Law Program Policy Initiative, and Robin Smith and Nate Bishop for their help and advice. Any mistakes or omissions are the author’s own.

[1] See, e.g., Christopher Bateman and James T.B. Tripp, Towards Greener FERC Regulation of the Power Industry, 38 Harv. Envtl. L. Rev. 275 (2014) (arguing that consideration of environmental consequences by FERC is permissible under the Federal Power Act); Joel B. Eisen, FERC’s Expansive Authority to Transform the Electricity Grid, 49 U.C. Davis L. Rev. 1783, 1788 (2016) (arguing that under recent case law, FERC may now include environmental considerations into wholesale rates so long as those considerations “directly affect” those rates); Steven Weissman & Romany Webb, Berkeley Center for Law, Energy & the Environment, Addressing Climate Change Without Legislation: Volume 2, How the Federal Energy Regulatory Commission Can Use Its Existing Legal Authority to Reduce Greenhouse Gas Emissions and Increase Clean Energy Use (2014), (arguing that FERC can add the cost of carbon when setting the prices in the wholesale market).

[2] Order No. 1000, Transmission Planning and Cost Allocation by Transmission Owning and Operating Public Utilities, 136 FERC ¶ 61,051, 76 Fed. Reg. 49841 (Aug. 11, 2011) (codified at 18 C.F.R. § 35) (requiring regional transmission planning to consider state and local public policy requirements); Order No. 745, Demand Response Compensation in Organized Wholesale Energy Markets, 134 FERC ¶ 61,187, 76 Fed. Reg. 16657 (Mar. 24, 2011) (codified at 18 C.F.R.§ 35.28(g)(1)(v)) (allowing demand response providers to bid into the wholesale market).

[3] New York v. FERC, 535 U.S. 1, 6–7 (1996).

[4] 136 S. Ct. 760, 773 (2016).

[5] FERC defines demand response as “a reduction in the consumption of electric energy by customers from their expected consumption in response to an increase in the price of electric energy or to incentive payments designed to induce lower consumption of electric energy.” 18 C.F.R. § 35.28(b)(4) (2015).

[6] See Joel B. Eisen, FERC v. EPSA and the Path to a Cleaner Energy Sector: Introduction, 40 Harv. Envtl. L. Rev. Forum 1, 7–8 (2016) (“In the long run, this concise, broad jurisdictional standard gives FERC considerable flexibility to promote a cleaner, more efficient Smart Grid.”).

[7] See 16 U.S.C. § 824 (a)–(b) (2016); New York, 535 U.S. at 19–25 (“FERC has recognized that the states retain significant control over local matters . . . [including] generation and transmission siting . . . [and] authority over utility generation and resource portfolios”) (citing Order No. 888, Promoting Wholesale Competition Through Open Access Non-discriminatory Transmission Services by Public Utilities, 75 FERC ¶ 61,080, 61 Fed. Reg. 21540, 21,626 n.543, n.544 (May 10, 1996) (codified at 18 C.F.R. § 35 and § 385)).

[8] See generally Toby Couture and Karlynn Cory, National Renewable Energy Laboratory, State Clean Energy Policies Analysis (SCEPA) Project: An Analysis of Renewable Energy Feed-in Tariffs in the United States (2009),

[9] See generally David Hurlbut, National Renewable Energy Laboratory, State Clean Energy Practices: Renewable Portfolio Standards (2008),

[10] See generally Eric Lantz and Elizabeth Doris, National Renewable Energy Laboratory, State Clean Energy Practices: State Renewable Rebates (2009),

[11] The State of Washington considered a carbon tax in a 2016 ballot initiative. See, Initiative Measure No. 732 (filed March 29, 2016)

[12] Minn. Stat. § 216H.03, subd. 3(2) and (3) (2007) (“no person shall . . . (2) import or commit to import from outside the state power from a new large energy facility that would contribute to statewide power sector carbon dioxide emissions; or (3) enter into a new long-term power purchase agreement that would increase statewide power sector carbon dioxide emissions.”)

[13] See generally Edison Electric Institute, Solar Energy and Net Metering (2016),

[14] Jocelyn Durkay, “State Renewable Portfolio Standards and Goals,” National Conference of State Legislatures (July 27th, 2016) (reporting that “Twenty-nine states, Washington, D.C. and three territories have adopted an RPS, while eight [additional] states have set renewable energy goals”).

[15] See Public Service Commission of N.Y., Order Adopting a Clean Energy Standard (Aug. 1 2016).

[16] See, e.g., North Dakota v. Heydinger, 825 F.3d 912 (8th Cir. 2016) (of three separate opinions, two held that Minnesota statute was preempted by the Federal Power Act); Rocky Mountain Farmers Union, et al., v. Richard W. Corey, 730 F.3d 1070 (9th Cir. 2013); Energy and Environmental Legal Institute v. Epel, 793 F.3d 1169 (10th Cir. 2015); Allco Finance Ltd. v. Klee, 805 F.3d 89, 95–96 (2d Cir. 2015) (rejecting plaintiff’s argument that solar contracts approved by the state regulator were preempted by the Public Utilities Regulatory Policies Act); see also Harvard Environmental Law and Policy Institute, State Power Project: Examining State Authority in Interstate Electricity Markets, (2016).

[17] Jim Rossi, The Brave New Path of Energy Federalism, 95 Tex. L. Rev. (forthcoming 2016).

[18] EPSA, 136 S.Ct. at 773.

[19] The D.C. Circuit found that FERC’s regulation of demand response transactions impermissibly intruded outside of FERC’s authorized jurisdiction under the Federal Power Act. EPSA v. FERC, 753 F.3d 216, 222 (D.C. Cir. 2014).

[20] EPSA, 136 S.Ct. at 774 (citing Calif. Independent System Operator v. FERC, 372 F.3d 395, 403 (D.C. Cir. 2004)).

[21] Hughes v. Talen Energy Marketing LLC, 136 S. Ct. 1288 (2016).

[22] PPL Energy Plus, LLC v. Nazarian, 753 F.3d 467 (4th Cir. 2014).

[23] Hughes, 136 S. Ct. at 1292.

[24] Id. at 1299 (“But the contract at issue here differs from traditional bilateral contracts in this significant respect: The contract for differences does not transfer ownership of capacity from one party to another outside the auction.”).

[25] Id. at 1297 (“A state law is preempted where Congress has legislated comprehensively to occupy an entire field of regulation, leaving no room for the States to supplement federal law,” as well as “where, under the circumstances of a particular case, the challenged state law stands as an obstacle to the accomplishment and execution of the full purposes and objectives of Congress” (citations omitted).

[26] Id. at 1300 (Sotomayor, J., concurring) (clarifying that the purpose of the Federal Power Act should serve as the “ultimate touchstone” for the preemption analysis and the Court should resist “talismanic” preemption vocabulary); id. at 1301 (Thomas, J., concurring) (stating that he would not rest his holding on principles of implied-preemption).

[27] Id. at 1299 (“Our holding is limited: We reject Maryland’s program only because it disregards an interstate wholesale rate required by FERC. We therefore need not and do not address the permissibility of various other measures States might employ to encourage development of new or clean generation, including tax incentives, land grants, direct subsidies, construction of state-owned generation facilities, or re-regulation of the energy sector. Nothing in this opinion should be read to foreclose Maryland and other States from encouraging production of new or clean generation through measures untethered to a generator’s wholesale market participation.”).

[28] See supra note 16 and accompanying text.

[29] Federal Power Commission v. Southern Cal. Edison Co., 376 U.S. 205, 215–216 (1964) (“Congress meant to draw a bright line easily ascertained, between state and federal jurisdiction. . .”). But see Oneok, Inc. v. Learjet, Inc., 135 S. Ct. 1591, 1601 (2015) (describing the clear division between federal and state authority in the Natural Gas Act context as a “Platonic ideal”); FERC v. EPSA, 136 S.Ct. 760, 780 (2016) (“The [Federal Power Act] makes federal and state authority complementary”); Hughes v. Talen Energy Marketing, LLC., 136 S.Ct. 1288 (2016) (Sotomayor, J., concurring) (“the Federal Power Act, like all collaborative federalism statutes, envisions a federal-state relationship marked by interdependence”).

[30] Mississippi Power & Light Co. v. Mississippi ex. rel. Moore, 487 U.S. 354, 377 (1984) (Scalia, J., concurring). The majority opinion also acknowledges “FERC has exclusive authority to determine the reasonableness of wholesale rates. . .” Id. at 355.

[31] Complaint, Coalition for Competitive Energy v. Zibelman (S.D.N.Y. filed Oct. 19, 2016) (No. 1:16-cv-08164),

[32] Public Service Commission of New York, Order Adopting a Clean Energy Standard (Aug. 1 2016),

[33] Id. at 6.

[34] Id. at 38.

[35] Id. at 45.

[36] Complaint at 5, Coalition for Competitive Energy v. Zibelman, (S.D.N.Y. filed Oct. 19, 2016) (No. 1:16-cv-8164),

[37] Id. at 11.

[38] Order Granting Preliminary Injunction, Allco Finance Ltd. v. Klee (2d. Cir. 2016) (No. 16-2946).

[39] See id.

[40] See id.

[41] Petition for Injunction at 2, Allco Finance Ltd. v. Klee, No. 16-2946 (2d. Cir. 2016) (No. 16-2946).

[42] Cf. American Council on Renewable Energy, Renewable Energy in Massachusetts (2014), (“In February 2014, the state approved 12 long-term power purchase agreements with four Massachusetts utilities for 409 MW of wind projects in Maine and New Hampshire”).

[43]See In the Matter of the Application of Ohio Electric Company, Case No. 14-1297-EL-SSO (Pub. Util. Comm’n of Ohio 2016),

[44] Gavin Bade, FERC Blocks Ohio Power Plant Subsidies for AEP and FirstEnergy, Utility Dive (Apr. 28, 2016),

[45] Hughes v. Talen Energy Marketing LLC, 136 S. Ct. 1288, 1299 (2016).

[46] 135 S. Ct. 1591, 1599 (instructing that for preemption under the Natural Gas Act, the appropriate inquiry is to examine the target at which state law “aims”).

[47] Jim Rossi and Thomas G. Hutton, Federal Preemption and Clean Energy Floors, 91 N.C. L. Rev. 1283 (2013).

[48] See Rossi, supra note 17 at 65 (stating that whether state programs are preempted may be left to FERC, as opposed to a case-by-case determination by the judiciary).

[49] See generally Brian Galle & Mark Seidenfeld, Administrative Law’s Federalism: Preemption, Delegation and Agencies at the Edge of Federal Power, 57 Duke L. J. 1933 (2008).

[50] See supra note 16 and accompanying text.

[ELRS] Pipelines, Protests and General Permits

By Samantha L. Varsalona, Staff Member, Georgetown Environmental Law Review

This post is part of the Environmental Law Review Syndicate


The Dakota Access Pipeline (DAPL) has become a contentious topic in recent months. The controversy centers around Dakota Access, LLC[1], a subsidiary of Energy Transfer Crude Oil Company, LLC, and the Standing Rock Sioux Tribe of North and South Dakota[2] (the Tribe or Sioux), a federally-recognized Indian tribe. The Tribe’s reservation, Standing Rock Indian Reservation, is half a mile upstream from where DAPL’s crude oil pipeline would cross the Missouri River underneath Lake Oahe in North Dakota.[3] While much of the recent media attention surrounding Dakota Access and the Tribe has focused on the destruction of the Tribe’s ancestral burial grounds, the underlying issue can be traced back to the nationwide permits issued by the Army Corps of Engineers (the Corps) in 2012.[4] More specifically, this article will examine Nationwide Permit 12 (NWP 12), which was one of the fifty NWPs issued by the Corps in 2012[5] and is at the heart of the current legal battle between Dakota Access and the Tribe.


The Tribe and environmentalists alike raised concerns about the potential health and environmental consequences of oil spills, being that the Missouri River “provides drinking water for millions of Americans and irrigation water for thousands of acres of farming and ranching lands.”[6] Besides the Tribes concern about the proximity of the pipeline to their reservation, they were also concerned about the pipeline disrupting sacred ancestral burial grounds and places of cultural significance to the Sioux people.[7] In particular, the Sioux have traditionally placed significance on the convergence of the Missouri and Cannonball Rivers because their ancestors gathered at that location to peacefully trade with other tribes.[8] Ironically, this is not the first time the Army Corps of Engineers (the Corps) or the federal government has taken the Tribe’s land in particular location without their consent. In 1958 the Corps dredged the sacred Cannonball river to construct the Oahe Dam, which created the man-made Lake Oahe that now covers the confluence of the two rivers and is the future site of DAPL.[9] The Oahe Dam not only destroyed a site of spiritual significance to the Sioux, but also flooded nearly fifty-six thousand acres of Standing Rock Reservation and over one hundred four thousand acres on the Cheyenne River Reservation.[10] Overall, the construction of the Oahe Dam destroyed more Indian land than any other public works project in America.[11] Nonetheless, the Tribe continues to use the banks of the Missouri River for “spiritual ceremonies, and the River, as well as Lake Oahe, plays an integral role in the life and recreation of those living on the reservation.”[12] With that poignant history in mind, it comes as no surprise that the Tribe would fight so vehemently against DAPL which would obviously affect both the Missouri River and Lake Oahe.

Fearing, once again, the possibility of sacred burial grounds being destroyed, the Tribe pursued legal action against the Army Corps of Engineers (Corps), the federal agency that approved DAPL’s permits, in hopes of being granted an injunction that would block DAPL’s construction of the pipeline.[13] The outcome of the suit, decided September 9th by the D.C. District Court, held that the Corps had sufficiently followed federal law in approving the pipeline.[14] Minutes after the court’s decision came down, the Department of Justice, the Department of the Army and the Department of the Interior issued a joint statement temporarily halting the work.[15]

The future of DAPL underneath Lake Oahe is still unclear and it will, more than likely, continue to be a political hot potato for months to come. In its simplest form, the conflict comes down to the permitting process and the Corp’s alleged failure to adequately consult the Tribe before issuing the permit.[16] The permit granted to DAPL is a type of general permit known as Nationwide Permit 12 (NWP 12) and has caused considerable controversy in the past several years.

Nationwide Permits

Although one might logically assume that a crude oil pipeline traversing thousands of miles across the United States would require an extensive federal appraisal and permitting process, that assumption would be incorrect. Domestic oil pipelines require no general approval from the federal government.[17] For example, DAPL needed almost no federal permitting of any kind because “99% of its route traversed private land.”[18] However, when construction activity occurs in waters of the United States, meaning in federally regulated waters such as Lake Oahe, the Corps needs to permit the activity under the Clean Water Act (CWA) or the Rivers and Harbors Act or sometimes both.[19]

Section 404(e) of the CWA has been the provision primarily used by the Corps to issue general permits.[20] Nationwide permits (NWP) are a type of general permit that are issued or reissued every five years by the Corps headquarters[21], whereas regional permits are issued by an individual Corps District for a specific geographical area.[22] NWPs authorize small-scale activities that are “similar in nature and result in no more than minimal individual and cumulative adverse environmental effects.”[23] Because NWPs pre-approve categories of activities upfront, there is considerably less federal involvement upon commencement of an individual project. Indeed, in most cases project proponents can commence their activities without ever notifying the Corps.[24] Some of the NWPs, including NWP 12, require the project proponent to submit a Pre-Construction Notification (PCN) to the relevant Corps District Engineer who then confirms whether or not the proposed activities qualify for NWP authorization.[25] If the District Engineer determines that the proposed activity qualifies, he/she then issues a verification letter to the project proponent. It is important to note that the District Engineer is merely verifying that the activity is one that was already pre-authorized by the Corps when they promulgated the NWP reissuance.[26]

NWPs are designed to streamline the permitting process and are often considered to be more cost-efficient and cost-effective for both the Corps and the individual or business seeking the permit.[27] Although NWPs can have important benefits when used for their intended purpose, some of the NWPs, NWP 12 in particular, are often used by the oil and gas industries as a way to fast-track the permitting process by avoiding project-specific environmental review and by skirting around a more comprehensive public participation process.[28] The oil and gas industries circumvent stricter federal regulations by evading the National Environmental Policy Act’s (NEPA) “hard look” review which requires federal agencies to analyze the environmental consequences of all “major Federal actions significantly affecting the quality of the human environment.”[29] If the federal action is one that would significantly affect the environment, the level of federal involvement and regulation is substantially elevated.[30]Although NEPA review applies only to major federal actions and imposes obligations only on federal agencies, “it is well-settled that ‘federal involvement in a non-federal project may be sufficient to federalize the project for purposes of NEPA.’”[31] In other words, it is possible for the Corps to have “sufficient control and responsibility”[32] over a project to warrant them having authority to control portions of a project that would normally be out of their jurisdiction. The district engineer makes the determination as to whether the scope of the Corps involvement warrants them to federalize the entire project.[33] For example, if a pipeline spans 100 miles and 40 miles of the project fall within federal control, the district engineer can determine the scope of the project gives the Corps sufficient control to warrant federalizing all 100 miles of the project, even if the other 60 miles were done by private action.[34]

NWP 12

The Corps renewed fifty nationwide permits on February 21, 2012 and they will expire on March 19, 2017.[35] The Corps, however, has no intention of letting these NWPs expire and on June 1, 2016 they proposed to reissue the NWPs and published the proposed rules in the Federal Register to solicited public comments.[36] The renewal included NWP 12, which covers “construction, maintenance, repair and removal of utility lines . . . provided the activity does not result in the loss of greater than 1/2 acre of waters of the United States for each single and complete project.”[37] The Corps defined NWP 12 to include “pipeline[s] for the transportation of gaseous, liquid, liquescent, or slurry substance, and any cable, line, or wire. . . .”[38] Accordingly, the construction of a pipeline may qualify for NWP 12 as long as the construction is a single and complete project and does not result in a loss greater than 1/2 acre of jurisdictional waters. At this point NWP 12 seems innocuous enough, however the conflict arises over the Corps defining a single and complete project as,“[the] portion of the total linear project proposed or accomplished by one owner/developer . . . that includes all crossings of a single water of the United States (i.e., a single waterbody) at a specific location.[39]

The effect of this definition is that it allows each water crossing to be verified under NWP 12 separately, essentially creating many “single and complete projects” along one proposed route.[40] In other words, the Corps allows pipeline proponents to “stack” NWP 12 hundreds, if not, thousands of times along a single pipeline.[41] For instance, TransCanada’s Gulf Coast Pipeline, which is the bottom half of the Keystone XL Pipeline, is 485 miles long and crosses United States waters 2,227 times, meaning the it “crosse[d] . . . waters about once every 1150 feet.”[42] The Corps verified the Gulf Coast Pipeline under NWP 12, even though NWP 12 was used 2,227 times in the process.[43] Another example is the Corps’ verification of Enbridge’s Flanagan South Pipeline under NWP 12 despite the pipeline traversing 27 miles of federal land, and crossing waters of the United States 1,950 separate times.[44] The Corps is essentially allowing project proponents to piecemeal the pipeline into separate smaller projects, which is seemingly inconsistent with NEPA.[45] What is perhaps more extraordinary is the Corps defines a single and complete non-linear project as requiring the project to have independent utility[46], which is defined as the project having the ability to be “constructed absent other projects in the project area.”[47] Not only does the definition of single and complete non-linear project require independent utility, it also specifically states “[s]ingle and complete non-linear projects may not be “piecemealed . . . .”[48] It is bewildering why Corps distinguishes so drastically between linear and non-linear projects, especially when considering linear projects that cannot function independently are, by their very nature, neither “single” nor “complete.”

The Corps justifies the expansive nature of NWP 12 by requiring the project proponent to submit a PCN to the Corps District Engineer (DE).[49] The DE will then review the PCN and determine if the proposed action “will result in more than minimal individual or cumulative adverse environmental effects or may be contrary to the public interest.”[50] On its face, requiring the DE to perform an extra layer of review may alleviate concerns about the open-ended nature of NWP 12. However, the review is based solely on the discretion of the DE and whether he/she determines there will be cumulative effects.[51] The PCN verification of the Gulf Coast Pipeline is an example of the considerable amount of discretion granted to the Corps. The Gulf Coast Pipeline passes through three Corps’ districts; Galveston, Fort Worth, and Tulsa and even though all three districts issued verification letters, none of the letters “provide a reasoned basis for any cumulative impacts analysis.”[52] As District Judge Martinez’s dissent points out, the verification letters issued by the three districts attempted to circumvent the analysis by “simply stat[ing] the legal standard and then recit[ing] that it made a ‘determination’ that such criteria were satisfied.”[53] Even though the DE and the Corps provided no specific findings as to why authorizing the use of NWP 12 2,227 times wouldn’t have a cumulative effect, the Tenth Circuit Court of Appeals approved the Corps use of discretion in verifying NWP 12.[54]

As seen above, the Corps definition of “single and complete” essentially allows the project proponent to segment the pipeline into smaller projects, which, in turn, allows the Corps to treat the project as not significant enough to warrant them having “control and responsibility”[55] over the entire project.[56] The approval of the Gulf Coast Pipeline is an example of how easily NWP 12 can be manipulated. Judge Martinez’s dissent challenges the Corps conclusion that its’ involvement did not warrant them to have sufficient control and responsibility and he asserted that “[c]onsidering the number of permits [2,227] issued by the Corps . . . it is patently ludicrous for Appellees to characterize the Corps’ involvement in the subject project as minimal . . . .”[57]

NWP 12 and DAPL

The malleability of NWP 12 is seen, once again, in its application permitting the Dakota Access Pipeline.[58] DAPL is not similar to the Gulf Coast Pipeline and Flanagan South Pipeline in the sense that the Corps didn’t seemingly abuse its authority by granting the use of NWP 12 thousands of times, rather the application of NWP 12 in DAPL’s context is offensive in the sense that it approved the pipeline even though the Tribe alleged it was not adequately consulted[59] as required under Section 106 of the National Historic Preservations Act (NHPA).[60]

Section 106, also known as the “stop, look, and listen” provision[61] requires “[f]ederal agencies takes into account the effects of their undertakings on historic properties and afford the Council a reasonable opportunity to comment on such undertakings.”[62] Meaning, the Corps are required to consider, prior to the reissuance of the NWPs, the effects of the permits on properties of cultural and historical significance.[63] This would have required the Corps to consult with the Tribe before they reissued the NWPs in 2012. Additionally, the consultation can’t just be a rubber stamping process, it “must recognize the government-to-government relationship between the Federal Government and Indian tribes.”[64]

The Corps claimed, and District Court Judge Boasberg agreed, that the Corps “made a reasonable effort to discharge its duties under NHPA prior to promulgating NWP 12” and that “the Corps’ effort to speak with those it thought be concerned was sufficient . . . .”[65] This “reasonable effort” to consult the Tribe included the Corps sending a notification letters containing information pertaining to its proposed NWPs, as well as the Corps holding listening sessions and workshops with tribes, and eventually the Corps sending letters to the Tribe inviting them to begin consultations.[66] The Advisory Council on Historic Preservation (ACHP), the federal agency that promulgates the regulations used to implement Section 106[67], wrote five letters[68] to the Corps questioning the adequacy of the tribal consultations. The EPA and Department of Interior also wrote letters to the Corps questioning their use of NWP 12 and the adequacy of tribal consultations.[69] The ACHP’s final letter states that it believes the “findings made by the Corps are premature, based on an incomplete identification effort, which was not sufficiently informed by the knowledge and perspective of consulting parties . . . .”[70] Despite all the objections from the Tribe and three other federal agencies, the Corps and Judge Boasberg emphasize that the Corps’ efforts were reasonable “given the nature of the permit.”[71] In other words, because NWP 12 is broad and over inclusive then apparently the Corps’ consultation requirements can be viewed in the same way.


This article has attempted to highlight a fundamental problem with how the United States permits domestic oil pipelines. The controversy surrounding the Dakota Access Pipeline has the potential to have both negative and positive implications. The most obvious potentially negative consequence is that the Sioux Tribe may, once again, lose sites of cultural significance at the hands of the U.S. government. However, a positive outcome that has emerged from this whole fiasco is that it has created a national dialog regarding not only nationwide permits and pipelines, but more importantly, how we, as citizens, view and understand the rights of Native Americans.

[1]           Energy Transfer, Overview, (last visited Oct. 10, 2016).

[2]           Indian Entities Recognized and Eligible To Receive Services From the United States Bureau of Indian Affairs, 80 Fed. Reg. 1942-02, 1946 (Jan. 14, 2015).

[3]           Standing Rock Sioux Tribe v. Army Corps of Engineers, No. 16-1534, 2016 WL 4734356, at *6 (D.D.C. Sept. 9, 2016).

[4]           Reissuance of Nationwide Permits, 77 Fed. Reg. 10,184 (Feb. 21 2012).

[5]           Id.

[6]           David Archambault II, Taking a Stand at Standing Rock, N.Y. Times (Aug. 24, 2016),

[7]           Standing Rock Sioux Tribe, 2016 WL 4734356, at * 6.

[8]           Id.

[9]           Id.

[10]         Michael L. Lawson, Dammed Indians: The Pick-Sloan Place and the Missouri River Sioux, 1944-1980, 50-52 (1994).

[11]         Id. at 50.

[12]         Standing Rock Sioux Tribe, 2016 WL 4734356, at * 6.

[13]         Complaint for Declaratory and Injunctive Relief at 1, Standing Rock Sioux Tribe, 2016 WL 4734356 (Jul. 27, 2016) (No. 1:16-cv-01534), 2016 WL 4033936.

[14]         Standing Rock Sioux Tribe, 2016 WL 4734356, at *26.

[15]         Dep’t of Justice, Joint Statement from the Department of Justice, the Department of the Army and the Department of the Interior Regarding Standing Rock Sioux Tribe v. U.S. Army Corps of Engineers (2016).

[16]         Standing Rock Sioux Tribe, 2016 WL 4734356, at *1.

[17]         Id.

[18]         Id. at *7.

[19]         Id. at *1.

[20]         33 U.S.C. § 1344(e)(1) (2012).

[21]         Id.

[22]         U.S. Army Corps of Eng’rs, About national and regional permits, (last visited Oct. 22, 2016).

[23]         Reissuance of Nationwide Permits, 77 Fed. Reg. at 10,186.

[24]         33 C.F.R. § 330.1(e)(1) (2013).

[25]         Reissuance of Nationwide Permits, 77 Fed. Reg. at 10,184.

[26]         Id. at 10,185.

[27]         See generally Eric Biber, The Permit Power Revisited: The Theory and Practice of Regulatory Permits in the Administrative State, 64 Duke L.J. 133 (2014).

[28]         Industry attorneys and environmental consulting firms have highlighted the strategic benefits of utilizing NWP 12 as a way to construct pipelines with minimal federal regulatory interference. See, Robert E. Holden, E&P Wetlands Compliance Strategy: Nationwide Permits, Law360 (Oct. 9, 2014); John Kusnier, What Pipeline Companies Should Consider When Planning Projects, North American Oil & Gas Pipelines, (July 19, 2013); Lowell M. Rothschild, The Importance Of Keystone To NWP 12, Law360 (Aug. 29, 2012)

[29]         Citizens Alert Regarding the Env’t v. EPA, 259 F.Supp.2d 9, 15 (D.D.C. 2003).

[30]         For a more detailed discussion of NEPA and its statutory goals, see Robertson v. Methow Valley Citizens Council, 490 U.S. 332 (1989).

[31]         Citizens Alert Regarding the Envt, 259 F.Supp.2d at 15 (quoting Macht v. Skinner, 916 F.2d 13, 18 (D.C. Cir. 1990)).

[32]         33 C.F.R. Part 325, app. B (7)(b)(2) (2013).

[33]         Id.

[34]         Id. § 7(b)(3).

[35]         Reissuance of Nationwide Permits, 77 Fed. Reg. at 10,184.

[36]         Dep’t of Defense, Proposal To Reissue and Modify Nationwide Permits (2016)

[37]         U.S. Army Corps of Eng’rs, Decision Document: Nationwide Permit 12 (2012), [hereinafter Nationwide Permit 12].

[38]         Id. at 1.

[39]         U.S. Army Corps of Eng’rs, 2012 Nationwide Permits, Conditions, and Definitions, with corrections (2012), (emphasis added) [hereinafter 2012 Nationwide Permits, Conditions, and Definitions].

[40]         Sierra Club, Comments on the U.S. Army Corps of Engineers’ Proposal to Reissue and Modify Nationwide Permit 12, (2016),

[41]         Id.

[42]         Sierra Club v. Bostick, 539 Fed. Appx. 887, 898 (10th Cir. 2013) [hereinafter Gulf Coast Pipeline].

[43]         Id.

[44]         Sierra Club v. Army Corps of Eng’rs, 803 F.3d 31, 39 (D.C. Cir. 2015).

[45]         See 40 CFR § 1508.25(a) (2010) (requiring connected and cumulative actions to be analyzed together unless they would have independent utility).

[46]         2012 Nationwide Permits, Conditions, and Definitions, at 45.

[47]         Id. at 43.

[48]         Id. at 45.

[49]         Nationwide Permit 12, at 2.

[50]         Reissuance of Nationwide Permits, 77 Fed. Reg. at 10260.

[51]         33 C.F.R. § 330.1(d).

[52]         Gulf Coast Pipeline, at 900.

[53]         Id.

[54]         Id. at 896.

[55]         See 33 C.F.R. pt. 325, app. B (7)(b).

[56]         See generally, Lindsay M. Nelson, The Gulf Coast Pipeline: A Stealthy Step Toward the Completion of the Keystone XL Pipeline Project, 44 Cap. U. L. Rev. 429 (2016).

[57]         Gulf Coast Pipeline, at 899 (emphasis added).

[58]         Standing Rock Sioux Tribe, 2016 WL 4734356, at *1.

[59]         Complaint for Declaratory and Injunctive Relief at 36-8, Standing Rock Sioux Tribe, 2016 WL 4734356 (Jul. 27, 2016) (No. 1:16-cv-01534), 2016 WL 4033936.

[60]         See generally, 36 C.F.R. § 800.2 (2016).

[61]         Standing Rock Sioux Tribe, 2016 WL 4734356, at *2.

[62]         36 C.F.R. § 800.1(a) (2016).

[63]         Standing Rock Sioux Tribe, WL 4734356, at *2.

[64]         Quechan Tribe of Fort Yuma Indian Reservation v. U.S. Dep’t of the Interior, 755 F. Supp. 2d 1104, 1108-9 (S.D. Cal. 2010).

[65]         Standing Rock Sioux Tribe, 2016 WL 4734356, at *19 (emphasis added).

[66]         Id. at *18-9.

[67]         Standing Rock Sioux Tribe, 2016 WL 4734356, at *1.

[68]         The Advisory Council on Historic Preservation, Dakota Access Pipeline Project 1 (May 19, 2016).

[69]         Environmental Protection Agency, Additional Comments on Dakota Access Pipeline Draft Environmental Assessment (March 11, 2016); Department of the Interior, Letter to the Corps (March 29, 2016).

[70]         The Advisory Council on Historic Preservation, Dakota Access Pipeline Project at 1.

[71]         Standing Rock Sioux Tribe, 2016 WL 4734356, at *19.

[ELRS] Science And Deference: The “Best Available Science” Mandate is A Fiction in the Ninth Circuit

By Elizabeth Kuhn, Associate Editor, Lewis & Clark, Environmental Law*

This post is part of the Environmental Law Review Syndicate


Many recent decisions by the Ninth Circuit[1] have required the court to review agency actions under the Administrative Procedure Act[2] (APA) arbitrary or capricious standard.[3] The Supreme Court has held that the arbitrary or capricious standard is a “highly deferential” standard of review, though the inquiry must nonetheless “be searching and careful.”[4] Furthermore, the agency’s decision is “‘entitled to a presumption of regularity,’ and [the Court] may not substitute [its] judgment for that of the agency.”[5] For purposes of this discussion, it is important to note that “traditional deference to the agency is at its highest where a court is reviewing an agency action that required a high level of technical expertise.”[6]

In cases where a petitioner is challenging an agency action under the Endangered Species Act[7] (ESA) the court will usually be tasked with reviewing whether the action was arbitrary or capricious in light of the ESA’s “best available science” mandate.[8] The ESA requires an agency to insure that its actions will not jeopardize the continued existence of any endangered species,[9] and the best available science mandate requires the agency to utilize the best available scientific data to inform its no jeopardy review.[10] Challenges to an agency action as arbitrary and capricious for failing to utilize the best available science must show that the agency ignored the relevant available science.[11]

Given the heightened level of deference for decisions based on science and the low standard of what constitutes the best available science, the ESA mandate rarely threatens to invalidate an agency’s decision.[12] In fact, none of the Ninth Circuit cases in the last year that have considered the issue have substantively evaluated an agency decision under the best available science mandate.[13] Rather, the agencies were given heightened deference to make their own decisions as to what constituted best available science.[14] This leaves us to wonder whether the ESA’s best available science mandate serves as a purposeful requirement in the Ninth Circuit.

The APA and the Arbitrary and Capricious Standard

The APA provides the standard for judicial review of an agency decision. Specifically, section 10 addresses judicial review and provides:

To the extent necessary to decision and when presented, the reviewing court shall decide all relevant questions of law, interpret constitutional and statutory provisions, and determine the meaning or applicability of the terms of an agency action.[15]

Section 10 further establishes the arbitrary and capricious standard by stating that the reviewing court shall “hold unlawful and set aside agency action, findings, and conclusions found to be … arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.”[16]

The APA’s arbitrary and capricious standard of review, however, is only applied when the governing legislation does not set forth its own standard of review.[17] There are several examples of legislation that utilize the APA as a default,[18] but key to this commentary is the fact that the ESA also relies on the APA as its default standard of review.

Meaning of Arbitrary and Capricious

Based on the text of the applicable legislation, it is easy to know when the arbitrary and capricious standard will be applied as the governing standard of review. However, in addition to understanding when the standard of review will be applied, it is helpful for both agencies and courts to have the same understanding of what is meant by “arbitrary and capricious.”

Congress did not define precisely what it meant by “arbitrary and capricious” within the text of the APA.[19] Instead, courts have looked to the terms’ ordinary meaning for a definition.[20] For example, Black’s Law defines arbitrary as a decision “founded on prejudice or preference rather than on reason or fact.”[21] Additionally, capricious is defined as “unpredictable or impulsive behavior” or “contrary to the evidence or established rules of law.”[22]


The arbitrary and capricious standard of review is a very narrow standard of review that requires the reviewing court to assume a deferential posture such that the court may not simply substitute its judgment for that of the agency.[23] Although the court’s deference must be at its highest when reviewing agency decisions relying on technical expertise, the reviewing court still has an affirmative obligation under the APA to ensure the agency exercised sound judgment and made a reasonable decision based on its available information.[24] Thus, in its review the court must walk a fine line between substituting its judgment for that of the agency and simply affirming agency decision making because it was the decision of the agency.

The U.S. Supreme Court has somewhat defined this line by stating that courts are only to determine if the agency considered the “relevant factors” and if the agency made a “clear error of judgment,” rendering its actions arbitrary and capricious.[25] Because terms such as “clear error of judgment” do not provide a clear standard, the Supreme Court articulated four specific scenarios for when agencies’ actions are considered arbitrary and capricious:

  1. The agency “relied on factors which Congress has not intended it to consider.”
  2. The agency “entirely failed to consider an important aspect of the problem.”
  3. The agency “offered an explanation for its decision that runs counter to the evidence before the agency.”
  4. The agency offered an explanation “so implausible that it could not be ascribed to a difference in view or the product of agency expertise.”[26]

These rules provide clarity to both courts and agencies because they set forth a specific standard for determining whether an agency has acted arbitrarily and capriciously.

Best Available Science Under the Endangered Species Act


The Endangered Species Preservation Act of 1966[27] (ESPA) was the first environmental statute to impose a requirement to utilize science in environmental decisions made by an administrative body.[28] The statute required the Secretary of the Interior to make determinations as to which species were at risk of extinction and directed the secretary to consult with relevant scientists in creating the list of endangered species.[29] The ESPA did not require the ultimate listing decisions to rest on the scientific information, but Congress intended the consultations to provide the foundation for the listings.[30]

The “best available science” requirement was later introduced in the Endangered Species Conservation Act of 1969[31] and remained largely unchanged in the current ESA.[32] However, Congress neither defined “best available science” nor provided instruction as to how to apply the requirement in either the 1969 Act or the current 1973 Act.[33] It has been suggested that the term “best available science” was not further defined in either the 1969 or 1973 statutes because Congress simply intended to continue the ESPA requirement to seek input from scientists prior to making listing decisions.[34]

What is Required Under the Best Available Science Mandate?

Without an explicit statutory definition or guidelines of how to apply the best available science mandate, we are forced to rely on judicial opinions interpreting the ESA to ascertain what is required by the mandate. Two distinct guidelines emerge from looking at these opinions: (1) an agency cannot ignore relevant available data and (2) an agency does not have an obligation to generate new data, even if only relatively weak data is available.[35]

The Ninth Circuit has repeatedly held that an agency “cannot ignore available biological information.”[36] Put more specifically, the agency “must not disregard available scientific evidence that is in some way better than the evidence it relies on.”[37] Furthermore, the court has held that an agency is not necessarily in noncompliance with the best available science mandate if it disagrees with or discredits the available scientific data.[38] For example, in Kern County Farm Bureau v. Allen[39] (Kern) the court rejected Kern’s argument that the United States Fish & Wildlife Service (FWS) violated the best available science mandate by misinterpreting three studies. In Kern, the fact that the FWS cited the studies and did not ignore them was enough to comply with the best available science mandate.[40] Therefore, a challenger must specifically point to relevant data that was omitted from consideration to sustain a claim that an agency failed to utilize the best available science.[41]

Although the Ninth Circuit has required an agency to utilize the best scientific data available, the court has also held that the mandate “does not… require an agency to conduct new tests or make decisions on data that does not yet exist.”[42] This holding is consistent with other circuits that have addressed this issue.[43] For example, the D.C. Circuit has held that an agency must utilize the best scientific data available, not the best scientific data possible.[44]

This approach has been met with criticism because agencies are allowed to rely on data that is weak or inconclusive when it is the only data available.[45] Because few data are available for many endangered species,[46] there exists the possibility that many decisions regarding endangered species will be made with little to no scientific data in support. If that were the case, the purpose of consulting scientific data prior to making a decision would be entirely undermined.

Application of the Best Available Science Mandate Under the Current Endangered Species Act

The best available science mandate is triggered any time an agency contemplates an action that might impact an endangered species. Section 7(a) of the ESA requires the agency to “insure that any action authorized, funded, or carried out by such agency is not likely to jeopardize the continued existence of any endangered or threatened species or result in destruction or adverse modification of the habitat of such species.”[47] Section 7(a) further requires that in fulfilling the requirements under the section the agency “shall use the best available scientific and commercial data.”[48]


The deference afforded to agencies in review of science-based decisions raises doubt as to whether the best available science mandate actually operates as a substantial requirement to an agency proposing an action under section 7. The Ninth Circuit in particular has held that when the analysis of an agency decision requires a high level of technical expertise, the court “must defer to the informed discretion of the responsible federal agencies.”[49] In fact, it is common practice across the circuits to give an “extreme degree” of deference to decisions founded on the scientific or technical expertise of an agency.[50]

Ninth Circuit Deference on Matters of Science

A Substantive Mandate in 2005

In 2005 the Ninth Circuit decided Pacific Coast Federation of Fishermen’s Ass’ns v. Bureau of Reclamation[51] (Pacific Coast) and breathed life into the best available science mandate. Prior to this decision, many courts had used deference to avoid upholding the substantive mandate requiring agencies to insure against jeopardy.[52] In Pacific Coast, the Ninth Circuit inserted itself into the Klamath Basin conflict.[53] The conflict stemmed from the National Marine Fisheries Service (NMFS) issuing a biological opinion (BiOp) requiring the Bureau of Reclamation (BOR) to limit diversion of water from the Klamath River for irrigation purposes because this diversion would jeopardize the continued existence of the endangered suckerfish and coho salmon.[54] This closure resulted in significant agricultural losses, as 2001 also saw record drought.[55]

After the drought of 2001, the Departments of the Interior and Commerce commissioned the National Research Council (NRC) to perform a “scientifically rigorous peer review” of whether the BiOp was consistent with available scientific information.[56] The conclusion of the NRC study questioned the validity of the 2001 BiOp.[57] The study found that “the 2001 BiOp’s drastic halting of water diversions was not scientifically supported,” but the study did not offer comment as to the minimum water levels necessary to maintain the endangered fish.[58]

In 2002, BOR prepared a long-range biological assessment and proposed a new flow regime that would vary the river flow by “water year type.”[59] The NMFS concluded that the BOR’s proposed actions would jeopardize the continued existence of coho salmon, and it issued a new BiOp that developed a reasonable and prudent alternative (RPA) to replace the BOR proposal.[60] That RPA was the subject of Pacific Coast.

The Northern District of California found that the short-term measures of the RPA were not arbitrary and capricious.[61] On appeal to the Ninth Circuit, the Court did not grant the customary heightened deference to the agency’s decision.[62] Rather, the Court engaged in a “careful and searching” review of the BiOp, stating that the agency “is obligated to articulate a rational connection between the facts found and the choices made.”[63] Specifically, the court found that

Although . . . the agency believed that the RPA would avoid jeopardy to the coho, this assertion alone is insufficient to sustain the BiOp and the RPA. The agency essentially asks that we take its word that the species will be protected if its plans were followed. If this were sufficient, the NMFS could simply assert that its decisions were protective and so withstand all scrutiny.[64]

Therefore, the Ninth Circuit found the authorized short-term measures of the Bi-Op to be arbitrary and capricious.[65]

This decision marked an important step in making the ESA’s best available science requirement a substantive mandate. Despite the deference due to the agency, the court looked substantively at the BiOp to find that it could not insure against jeopardy. This case sent a message that an agency could not rely on heightened deference to avoid judicial review of its actions.

Clarification of the Arbitrary and Capricious Standard in 2008

In 2008, the Ninth Circuit sought to “clarify some of [its] environmental jurisprudence” by hearing en banc Lands Council v. McNair (Lands Council III).[66] The court felt a need for uniformity because Ecology Center, Inc. v. Austin[67] “defied well-established law concerning the deference [the court] owe[s] to agencies and their methodological choices.”[68] Additionally, the court likely wanted to address the fact that “in recent years, [the Ninth Circuit’s] environmental jurisprudence has, at times, shifted away from the appropriate standard of review and could be read to suggest” that judges should sit on the bench and “act as a panel of scientists.”[69]

The en banc review resulted in a reversal of the preliminary injunction initially granted by the Ninth Circuit in The Lands Council v. McNair (Lands Council II)[70] and the overruling of Ecology Center.[71] Lands Council III overruled Ecology Center’s instruction that courts may suggest how an agency is required to validate its scientific methodology.[72] In Ecology Center, the court required the Forest Service to “demonstrate the reliability of its scientific methodology or the hypothesis underlying the Service’s methodology with on the ground analysis,”[73] but the court in Lands Council III concluded that the Forest Service may use a particular analysis “if it deems it appropriate or necessary, but it is not required to do so.”[74] In other words, as long as “there is a reasonable scientific basis to uphold the legitimacy of [the] modeling,” the courts are required to give deference to the agency and uphold its model.[75] Therefore, Lands Council III significantly reigned in the court’s ability to question how agencies justify scientific methodology.

In addition to precluding courts from prescribing the means by which an agency validates its scientific methodologies, Lands Council III also established that courts do not have the authority to choose which scientific studies support agency actions.[76] If the agency considered the scientific evidence available to it, courts must defer to the agency’s interpretations of that scientific evidence.[77]Therefore, because the Forest Service considered many different studies, the court in Lands Council III explicitly deferred to the agency’s interpretation of the scientific evidence.[78]

Finally, Lands Council III overruled Ecology Center’s requirement that an agency must present every scientific uncertainty in the evidence used to inform a decision.[79] Consequently, an agency no longer bears “the burden to anticipate questions that are not necessary to its analysis, or to respond to uncertainties that are not reasonably supported by any scientific authority.”[80] The Ninth Circuit only requires that an agency “acknowledge and respond to comments by outside parties that raise significant scientific uncertainties and reasonably support that such uncertainties exist.”[81]

Thus, the en banc court established three rules to guide Ninth Circuit jurisprudence when using the arbitrary and capricious standard of review for an agency’s use of science:

  1. Courts may not prescribe the specific means by which an agency must validate methodologies.
  2. Courts may no longer choose between which scientific studies support an agency’s action, so long as the agency provides an explanation for its conclusion.
  3. An agency no longer needs to address every scientific uncertainty surrounding the science it uses to support its position. The agency only needs to “acknowledge” and “respond” to the claims by parties raising and supporting that “significant scientific uncertainties” exist.[82]

Current Cases

Pacific Coast marked what commentators believed was a change toward a more substantive science requirement.[83] However, a decade later it does not appear as though the Ninth Circuit has continued down the Pacific Coast path of reducing the deference it affords to agencies when reviewing compliance with the best available science mandate. Rather, the Ninth Circuit has stayed consistent with the “rules” issued by the Lands Council III en banc court. However, the three cases decided by the Ninth Circuit in 2015 reviewing the best available science requirement under the ESA[84] show that heightened agency deference is rendering the science mandate utterly meaningless.

In Alliance for the Wild Rockies v. Bradford,[85] the Ninth Circuit issued a memorandum opinion affirming that the United States Forest Service (USFS) did not violate the ESA by concluding that its Grizzly Project would not likely adversely affect the grizzly bear population.[86] The court noted that USFS met the requirements of the ESA by consulting the Wakkinen Study when making its determination.[87] The court also noted that its review of the scientific judgments and technical analyses made within an agency’s field of expertise should be at its most deferential.[88] Therefore, the court concluded that USFS had complied with the ESA’s best available science mandate.[89]

In Center for Biological Diversity v. United States Fish & Wildlife Service,[90] the Center for Biological Diversity (CBD) brought suit against the FWS challenging the FWS’s decision to sign a memorandum of agreement (MOA) for groundwater pumping based on conclusions reached in its biological opinion (BiOp).[91] CBD sued for declaratory and injunctive relief against the FWS alleging, among other things, that the BiOp failed to meet the best available science standard set forth by §7 of the ESA.[92]

Specifically, CBD argued that the foundation of the BiOp’s no jeopardy finding was based on expediency not on science.[93] CBD attempted to support its argument by pointing to the fact that the conservation measures’ flow reduction triggers were negotiated and not biologically based.[94] The Ninth Circuit noted that the ESA does not require FWS to design or plan its projects using the best science possible.[95] Rather, “once action is submitted for formal consultation, the consulting agency must use the best scientific and commercial evidence available in analyzing the potential effects of that action on endangered species in its biological opinion.”[96] Therefore, the court concluded that negotiated terms do not of themselves prove that the BiOp analysis failed to utilize the best available science.[97]

Additionally, CBD argued that the BiOp’s conclusions should not be given deference because the FWS failed to address concerns raised by its own scientists regarding the effectiveness of the MOA’s conservation measures.[98] The Ninth Circuit explained that CBD’s claim failed as there was no evidence supporting a conclusion that FWS scientists’ concerns were supported by better science than the science used in the BiOp, or that FWS disregarded better scientific information than the evidence FWS relied upon.[99] Thus, the Ninth Circuit concluded that CBD was unable to prove that the no jeopardy conclusion in the BiOp was arbitrary or capricious for failing to utilize the best available science.[100]

In Cascadia Wildlands v. Thrailkill,[101] Cascadia Wildlands (Cascadia) brought action seeking to enjoin the Douglas Fire Complex Recovery Project (Recovery Project), which authorized salvage logging of roughly 1,600 acres of fire-damaged forest.[102] In approving the Recovery Project, the Medford District of the Bureau of Land Management relied on a biological opinion issued by the FWS.[103] This biological opinion concluded that the Recovery Project was not likely to result in jeopardy to the Northern Spotted Owl species or in destruction or adverse modification of the critical habitat.[104] Cascadia claimed the FWS biological opinion failed to comply with requirements of the ESA because the FWS did not apply scientific data to the opinion.[105]

As to the no jeopardy conclusion, the court found that the record supported that the FWS relied on several surveys to reach its conclusion and gave the agency deference that the data it used was the best available scientific data.[106] With regard to the effects on the habitat, the court found that the FWS utilized several lengthy scientific reports regarding pre-fire and post-fire habitats to support the conclusion in its biological opinion.[107] Furthermore, the court noted that a reviewing court cannot substitute its judgment for that of the agency when the agency used adequate and reliable data.[108]

Cascadia also argued that the FWS’s 2011 Northern Spotted Owl Recovery Plan constituted the best available science and that the FWS was required to follow it.[109] The court rejected this argument. The court stated that recovery and jeopardy are two distinct concepts.[110] The court noted that a Recovery Project does not necessarily need to promote or bring about a long-term recovery of the species.[111] Rather, the biological opinion should and does focus on the Recovery Project’s ability to conserve the habitat so as not to have a detrimental effect on the species population.[112]

The court ultimately concluded that Cascadia failed to show that the FWS did not utilize the best available scientific information when issuing its biological opinion that the Recovery Project would not jeopardize the Northern Spotted Owl or its critical habitat.[113] Therefore, the Ninth Circuit affirmed the district court’s denial of the preliminary injunction to enjoin the Recovery Project.[114]

These three cases illustrate that the Ninth Circuit is still affording agencies heightened deference in scientific judgments and technical analyses. The court appears to look merely at whether the agency consulted scientific data prior to making decisions without reviewing the adequacy of the scientific data. Therefore, the ESA’s best available science mandate can be easily satisfied and will be subject to little scrutiny in the Ninth Circuit.


When reviewing scientific decisions based on agency expertise, the standard practice across the circuits is to afford deference to the agency unless is it shown that the agency ignored relevant scientific data when making its decision.[115] Unfortunately, this practice leaves little recourse for petitioners seeking to hold an agency accountable for substantiating its decision. As it stands now, the best available science requirement is satisfied as long as the agency considers the available data.[116] The agency is free to disagree with the data, discredit the data, or rely on weak or inconclusive data if it is the only data available.[117] As long as the agency articulates the rationale between the data and the decision made, the court will uphold the agency action.[118] This means that as long as an agency communicates a justification for its decision, the justification itself will more than likely not be reviewed by the court.

In 2005, the Ninth Circuit substantively reviewed an agency decision and found the agency relied heavily on unstated assumptions rather than scientific evidence.[119] Had the court simply given deference to the agency’s conclusion because it articulated a justification for its decision, the court would have failed to notice that the agency was not actually basing that decision on science. Pacific Coast exemplifies the need for substantive review of agency decisions, even though the court does not like to assume the role of technical expert.[120]

Although the Ninth Circuit demonstrated in Pacific Coast that it was willing to substantively review agency decisions relating to science, the court has since shifted back to the more customary deferential approach. As the three 2015 cases show, the Ninth Circuit is reluctant to substitute its judgment for that of an agency with regard to science and as a result affords agencies great deference when reviewing decisions based on the agency’s scientific expertise.

It is unclear why the Ninth Circuit has shifted back to the deferential standard of review. Perhaps it is because Congress has remained silent on the science standard for over three decades, or perhaps the court is reluctant to proceed differently than the other circuits. Whatever the reason, it is clear that until courts engage in substantive review of agencies’ scientific decisions or Congress establishes an explicit standard of the type and quality of scientific data required, the best science available mandate will continue to operate as a fiction in the review of agency decisions.

* J.D. Candidate 2017, Lewis & Clark Law School. Please send correspondence to

[1] E.g., Cascadia Wildlands v. Thrailkill, 806 F.3d 1234 (9th Cir. 2015); Ctr. for Biological Diversity v. U.S. Fish & Wildlife Serv., 807 F.3d 1031 (9th Cir. 2015); All. for the Wild Rockies v. Bradford, 601 Fed. App’x 488 (9th Cir. 2015) (mem.).

[2] Administrative Procedure Act, 5 U.S.C. §§ 551–559, 701–706, 1305, 3105, 3344, 4301, 5335, 5372, 7521 (2012).

[3] 5 U.S.C. § 706(2)(A) (requiring a reviewing court to uphold agency action unless it is “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.”).

[4] Marsh v. Or. Nat. Res. Council, 490 U.S. 360, 378 (1989); San Luis & Delta-Mendota Water Auth. v. Jewell, 47 F.3d 581 (9th Cir. 2014).

[5] Jewell, 47 F.3d at 601. (quoting Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 415–16 (1971)).

[6] Marsh, 490 U.S. at 377.

[7] Endangered Species Act of 1973, 16 U.S.C. §§ 1531–1544 (2012).

[8] See cases cited supra note 2.

[9] 16 U.S.C. § 1536(a)(2) (2012).

[10] Id.

[11] See Bldg. Indus. Ass’n of Superior Cal. v. Norton, 247 F.3d 1241, 1246 (D.C. Cir. 2001).

[12] Katherine Renshaw, Leaving the Fox to Guard the Henhouse: Bringing Accountability to Consultation Under the Endangered Species Act, 32 Colum. J. Envtl. L. 161, 187 (2007).

[13] See cases cited supra note 2.

[14] See cases cited supra note 2.

[15] 5 U.S.C. § 706 (2012).

[16] Id.

[17] E.g., Al-Fayed v. Cent. Intelligence Agency, 254 F.3d 300, 304 (D.C. Cir. 2001) (“The APA, however, ‘provides a default standard of judicial review . . . where a statute does not otherwise provide a standard.’”).

[18] The National Forest Management Act (NFMA), 16 U.S.C. §§1600–1687 (2012), and the National Environmental Policy Act of 1969 (NEPA), 42 U.S.C. §§ 4321–4370h (2012), are other examples of legislation that rely on the APA as a default standard of review.

[19] See 5 U.S.C. §706(2)(A) (2012).

[20] See Fed. Commc’ns Comm’n v. Fox Television Stations, Inc., 556 U.S. 502, 516 (2009) (stating that the arbitrary and capricious standard is satisfied so long as the Commission’s action was not arbitrary or capricious in the ordinary sense); United States v. Locke, 471 U.S. 84, 95 (1985) (deference to the supremacy of the Legislature, as well as recognition that Congressmen typically vote on the language of a bill, generally requires us to assume that the legislative purpose is expressed by the ordinary meaning of the words used).

[21] Black’s Law Dictionary 112 (9th ed. 2009).

[22] Id. at 224.

[23] Marsh v. Or. Nat. Res. Council, 490 U.S. 360, 377–78 (1989); see U.S. Postal Service v. Gregory, 534 U.S. 1, 6–7 (2001).

[24] See Marsh, 490 U.S. at 377–78.

[25] Id.

[26] Motor Vehicle Mfrs. Ass’n of U.S., Inc. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43 (1983).

[27] Pub. L. No. 89-669, 80 Stat. 926 (1966).

[28] Holly Doremus, Listing Decisions Under the Endangered Species Act, Why Better Science Isn’t Always Better Policy, 75 Wash. U.L.Q. 1029, 1042 (1997).

[29] § 1(c), 80 Stat. at 926.

[30] Doremus, supra note 15, at 1042.

[31] Pub. L. No. 91-135, 83 Stat. 275 (1969).

[32] See 16 U.S.C. § 1536(a)(2) (2012).

[33] See § 3(a), 83 Stat. at 275; 16 U.S.C. § 1532 (2012).

[34] Doremus, supra note 15, at 1043.

[35] Renshaw, supra note 12, at 167.

[36] Conner v. Burford, 848 F.2d 1441, 1454 (9th Cir. 1988); see also San Luis & Delta-Mendota Water Auth. v. Locke, 776 F.3d 971, 995 (9th Cir. 2014).

[37] San Luis & Delta-Mendota Water Auth., 776 F.3d at 995.

[38] See id.

[39] 450 F.3d 1072 (9th Cir. 2006).

[40] Id. at 1081.

[41] See id.

[42] San Luis & Delta-Mendota Water Auth., 776 F.3d at 995.

[43] E.g., Am. Wildlands v. Kempthorne, 530 F.3d 991, 998–99 (D.C. Cir. 2008).

[44] Bldg. Indus. Ass’n of Superior Cal. v. Norton, 247 F.3d 1241, 1246 (D.C. Cir. 2001).

[45] Renshaw, supra note 12, at 169.

[46] Id.

[47] 16 U.S.C. § 1536(a)(2) (2012).

[48] Id.

[49] Selkirk Conservation All. v. Fosgren, 336 F.3d 944, 954 (9th Cir. 2003).

[50] City of Waukesha v. U.S. Envtl. Prot. Agency, 320 F.3d 228, 247 (D.C. Cir. 2003); see also Maine v. Norton, 257 F. Supp. 2d at 389 (“The court must defer to the agency’s expertise, particularly with respect to decision-making which involves a high level of technical expertise.”); A.M.L. Int’l, Inc. v. Daley, 107 F. Supp. 2d 90, 102 (D. Mass. 2000) (“Indeed, a reviewing court must afford special deference to an agency’s scientific expertise.”).

[51] 426 F.3d 1082 (9th Cir. 2005).

[52] Renshaw, supra note 12, at 187.

[53] See Pacific Coast, 426 F.3d 1082.

[54] See id. at 1087.

[55] See id.

[56] J.B. Ruhl, The Battle Over Endangered Species Act Methodology, 34 Envtl. L. 555, 584–85 (2004).

[57] Pacific Coast, 426 F.3d at 1087.

[58] Renshaw, supra note 12, at 188.

[59] Pacific Coast, 426 F.3d at 1088.

[60] Id.

[61] Id. at 1089.

[62] See id.

[63] Id. at 1091.

[64] Id. at 1092.

[65] Id.

[66] 537 F.3d 981, 984 (9th Cir. 2008).

[67] 430 F.3d 1057 (9th Cir. 2005).

[68] Lands Council III, 537 F.3d at 991.

[69] Id. at 998.

[70] 494 F.3d 771 (9th Cir. 2007), rev’d en banc, 537 F.3d 981 (9th Cir. 2008).

[71] Lands Council III, 537 F.3d 990–94.

[72] Id. at 990.

[73] Ecology Center, 430 F.3d at 1064.

[74] Lands Council III, 537 F.3d 991–92.

[75] Id. at 992.

[76] Id. at 994–95.

[77] Id. at 995.

[78] Id. at 996.

[79] Lands Council III, 537 F.3d at 1001.

[80] Id.

[81] Id.

[82] Id. at 992–94, 1001 ; see also Ryan G. Welding & Michael E. Patterson, Maintaining the Ninth Circuit’s Clarified Arbitrary and Capricious Standard of Review for Agency Science After Lands Council v. McNair, 31 Pub. Land & Resources L. Rev. 55, 79–80 (2010).

[83] See Renshaw, supra note 12.

[84] See cases cited supra note 2.

[85] 601 Fed App’x 488 (9th Cir. 2015).

[86] Id. at 490.

[87] Id. (“The Forest Service relied on the Wakkinen Study, which is the best available science, and the Fish & Wildlife Service concurred in the Forest Service’s determination.”).

[88] Id.

[89] Id.

[90] 807 F.3d 1031 (9th Cir. 2015).

[91] Id. at 1035.

[92] Id.

[93] Id. at 1048.

[94] Id.

[95] Id.

[96] Id.

[97] Id.

[98] Id.

[99] Id. at 1049–50.

[100] Id. at 1049–51.

[101] 806 F.3d 1234 (9th Cir. 2015).

[102] Id. at 1235–36.

[103] Id.

[104] Id. at 1236.

[105] Id. at 1238–41.

[106] Id. at 1241–42.

[107] Id. at 1242.

[108] Id. at 1243.

[109] Id. at 1243–44.

[110] Id.

[111] Id.

[112] Id.

[113] Id. at 1244.

[114] Id.

[115] See cases cited supra note 50.

[116] See Conner v. Burford, 848 F.2d 1441, 1454 (9th Cir.1988); see also San Luis & Delta-Mendota Water Auth. v. Locke, 776 F.3d 971, 995 (9th Cir. 2014).

[117] See, e.g., San Luis & Delta-Mendota Water Auth., 776 F.3d at 995.

[118] Ctr. for Biological Diversity v. U.S. Fish & Wildlife Serv., 807 F.3d 1031, 1043 (9th Cir. 2015).

[119] See Pacific Coast, 426 F.3d 1082.

[120] See Marsh v. Or. Nat. Res. Council, 490 U.S. 360, 377 (1989.