Lessons from the Clean Air Act: Building Durability and Adaptability into US Climate and Energy Policy 1108421520, 9781108421522

Table of contents :
Contents
List of Figures
List of Tables
List of Contributors
Acknowledgments
1 Introduction
2 The Clean Air Act’s National Ambient Air Quality Standards
3 Stationary Sources, Movable Rules
4 Leveraged Federalism and the Clean Air Act
5 Promoting Environmental Quality through Fuels Regulations
6 The Clean Air Act’s Use of Market Mechanisms
7 Conclusion
Index

Citation preview

lessons from the clean air act Building Durability and Adaptability into U.S. Climate and Energy Policy Climate and energy policy needs to be durable and flexible to be successful, but these two concepts often seem to be in opposition. One venerable institution where both ideas are apparent is the Clean Air Act, first passed by the United States Congress in 1963, with amendments in 1970 and 1990. The Act is a living institution that has been hugely successful in improving the environment. It has programs that reach across the entire economy, regulating various sectors and pollutants in different ways. This illuminating book examines these successes – and failures – with the aim to offer lessons for future climate and energy policymaking in the US at the federal and state level. It provides critical information to legislators, regulators and scholars interested in understanding environmental policymaking. Ann Carlson is the Shirley Shapiro Professor of Environmental Law at the University of California, Los Angeles School of Law. She is the faculty co-director of the Emmett Institute on Climate Change and the Environment and has written extensively on climate change, air pollution and energy law. Dallas Burtraw is the Darius Gaskins Senior Fellow at Resources for the Future. Burtraw has worked to promote efficient control of air pollution and has written extensively on electricity industry regulation and environmental outcomes.

Lessons from the Clean Air Act building durability and adaptability into u.s. climate and energy policy Edited by

ANN CARLSON University of California, Los Angeles, School of Law

DALLAS BURTRAW Energy and Climate Program, Resources for the Future

American Academy of Arts and Sciences, Cambridge, Massachusetts

University Printing House, Cambridge cb2 8bs, United Kingdom One Liberty Plaza, 20th Floor, New York, ny 10006, USA 477 Williamstown Road, Port Melbourne, vic 3207, Australia 314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre, New Delhi – 110025, India 79 Anson Road, #06–04/06, Singapore 079906 Cambridge University Press is part of the University of Cambridge. It furthers the University’s mission by disseminating knowledge in the pursuit of education, learning, and research at the highest international levels of excellence. www.cambridge.org Information on this title: www.cambridge.org/9781108421522 doi: 10.1017/9781108377195 © American Academy of Arts and Sciences, 2019 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2019 Printed and bound in Great Britain by Clays Ltd, Elcograf S.p.A. A catalogue record for this publication is available from the British Library. Library of Congress Cataloging-in-Publication Data names: Carlson, Ann E., editor. | Burtraw, Dallas, editor. title: Lessons from the Clean Air Act : building durability and adaptability into US climate and energy policy / edited by Ann Carlson, Dallas Burtraw. description: Cambridge [UK] ; New York, ny : Cambridge University Press, 2018. identifiers: lccn 2018045900 | isbn 9781108421522 (hardback) subjects: lcsh: United States. Clean Air Act. | Air–Pollution – Law and legislation – United States. | Air – Pollution – Law and legislation – Compliance costs – United States. | Air quality management – United States. | Climatic changes – Government policy – United States. | Energy policy – United States. classification: lcc kf3812 .l47 2018 | ddc 344.7304/642– dc23 LC record available at https://lccn.loc.gov/2018045900 isbn 978-1-108-42152-2 Hardback isbn 978-1-108-43266-5 Paperback The views expressed in this volume are those of the contributors and do not necessarily reflect the views of the Officers and Members of the American Academy of Arts and Sciences. Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

To Bob Fri (1935–2014), our intellectual leader on this project and our dear friend

Contents

page viii

List of Figures List of Tables

ix

List of Contributors

x xii

Acknowledgments 1

Introduction Ann Carlson and Dallas Burtraw

2

The Clean Air Act’s National Ambient Air Quality Standards A Case Study of Durability and Flexibility in Program Design and Implementation William Boyd

3

4

5

1

15

Stationary Sources, Movable Rules Intransigence and Innovation under the Clean Air Act Hannah J. Wiseman

57

Leveraged Federalism and the Clean Air Act The Case of Vehicle Emissions Control Barry G. Rabe

113

Promoting Environmental Quality through Fuels Regulations Lessons for a Durable Energy and Climate Policy Joseph E. Aldy

159

6

The Clean Air Act’s Use of Market Mechanisms Eric M. Patashnik

201

7

Conclusion Ann Carlson and Dallas Burtraw

225

Index

239 ii

Figures

2.1 2.2 2.3 2.4 3.1 5.1 5.2 6.1 6.2

NAAQS framework page 28 Current NAAQS 30 NAAQS review process 32 Progress in reducing criteria pollutants 48 Selected criteria pollutant emissions from selected stationary sources 92 RFG and low-volatility fuel markets in the United States 167 Structure of nested requirements under RFS2 173 SO2 emissions from Acid Rain Program sources, 1980–2009 209 Ozone season NOx emissions from all NOx Budget Trading Program sources 211

iii

Tables

3.1 4.1 4.2 4.3 4.4 5.1 5.2

Major CAA Provisions for Stationary Sources, Excluding SIPs and the Acid Rain Program page 81 Political Leadership and Partisan Control of US Federal and State of California Executive Branches, 1966–2016 118 EPA Decisions on California Waiver Requests, 1968–2017 131 US Federal and California Light-Duty Vehicle Emission Standards, 1966–2010 142 Average Age of Motor Vehicles in Operation in the United States, 1969–2015 145 Outline of the Boutique Fuels Program 166 Volume Requirements under RFS1 and RFS2 172

i

Contributors

Joseph E. Aldy is Associate Professor of Public Policy at the John F. Kennedy School of Government at Harvard University, University Fellow at Resources for the Future, Faculty Research Fellow at the National Bureau of Economic Research and Senior Adviser at the Center for Strategic and International Studies. Aldy served as the Special Assistant to the President for Energy and Environment in the White House in 2009–10. William Boyd is Professor of Law at the UCLA School of Law and Professor at the UCLA Institute of the Environment and Sustainability. He teaches and conducts research in the areas of energy law and regulation, climate change law and policy and environmental law. Dallas Burtraw is the Darius Gaskins Senior Fellow at Resources for the Future. Burtraw has worked to promote efficient control of air pollution and has written extensively on electricity industry regulation and environmental outcomes. Ann Carlson is the Shirley Shapiro Professor of Environmental Law at the UCLA School of Law. She is Faculty Codirector of the Emmett Institute on Climate Change and the Environment and has written extensively on climate change, air pollution and energy law. Eric M. Patashnik is the Julis-Rabinowitz Professor of Public Policy and Professor of Political Science at the Watson Institute for International and Public Affairs at Brown University and Director of the Master of Public Affairs Program. He is the author of several books, including Reforms at Risk: What Happens after Major Policy Changes Are Enacted (2008). Barry G. Rabe is the J. Ira and Nicki Harris Family Professor of Public Policy at the Gerald R. Ford School of Public Policy at the University of Michigan. Rabe is

List of Contributors

xi

a political scientist who studies federalism and is the author, most recently, of Can We Price Carbon? (2018). Hannah Wiseman is the Attorneys’ Title Professor at the Florida State University College of Law. Her research focuses on choice of governance level and effective regulation of risk in the areas of energy and environmental law.

Acknowledgments

This book is the culmination of a multiyear study convened by the American Academy of Arts and Sciences as part of its Alternative Energy Future initiative. We are grateful to the Alfred P. Sloan Foundation, whose support not only allowed the Academy to commission this research but also allowed us to organize several meetings of the study group to refine our thinking and identify connections among our various case studies. These meetings were particularly critical to our work given the strong interdisciplinary nature of this project. We also thank the Kresge Foundation for supporting a series of workshops to discuss our conclusions with scholars and policymakers across the country. We are grateful to the other members of our research team – Joseph E. Aldy, William Boyd, Eric M. Patashnik, Barry G. Rabe and Hannah J. Wiseman – who devoted countless hours to research and writing and patiently responded to several rounds of review and requests for revisions. We are also indebted to Academy members Maxine Savitz and Granger Morgan, cochairs of the Alternative Energy Future project, and project advisors Barbara Kates-Garnick and Judson Jaffe for generously sharing their advice and expertise. Myron Gutmann, former Associate Director for Social, Behavioral, and Economic Sciences at the National Science Foundation, encouraged us to pursue this study and funded an initial workshop in February 2013 to shape its design. We thank the participants in that workshop, as well as those who attended a workshop in Washington, DC, in March 2017 to discuss our preliminary analysis. Their many comments and suggestions had a tremendous influence on our final conclusions. The Academy provided extraordinary organizational support for this study. We are especially grateful to John Randell, John E. Bryson Director of Science, Engineering, and Technology Programs at the Academy, for his strong facilitation of this project, as well as his colleagues Alison Leaf, Greg Savageau, Zackory Burns, Shalin Jyotishi and Rachel Johnson. Kristen McCormack and Amelia Keyes from Resources for the Future provided critical analytical support and offered substantial ii

Acknowledgments

xiii

feedback on the chapter drafts. Matthew Schneider, Simon Vickery and Christi Zaleski provided important research support at the beginning of the project, and their thorough analysis of the prior scholarship, legislation and key judicial decisions pertaining to the Clean Air Act provided a strong foundation for this study. Above all, we owe an incalculable debt of gratitude to the late Robert W. Fri, who chaired the Academy’s Alternative Energy Future project from 2010 until his death in October 2014. Bob was the intellectual driver behind this research, and his vision and spirit have continued to guide this project. We and countless others benefited from his extraordinary wisdom and leadership over the course of his long career. This book is dedicated to his memory.

1 Introduction Ann Carlson and Dallas Burtraw

Despite the recent retrenchment of the United States from its commitments under the Paris Agreement, the USA is indispensable to international efforts to address climate change. We remain the second largest emitter of greenhouse gases in the world.1 Historically, we are the largest emitter.2 In order to contribute to global efforts to limit temperature increases to 2˚C or less, the United States will need to transform its electricity and transportation sectors in the coming decades, with other parts of the economy to follow.3 Electricity and transportation in combination emit more than half the country’s carbon dioxide, the most common greenhouse gas.4 Their transformation will need to occur over the next four and a half decades (roughly the life of the forty-nine-year-old Clean Air Act) and will require massive shifts in the fuels we use to power our businesses, industry, government and homes and to propel our cars, trucks, trains, ships and planes.5 In designing public policy, analysts often focus on criteria such as efficiency and distributional fairness. These are important attributes of good policy design, 1

2

3

4

5

In 2014, US carbon dioxide (CO2) emissions constituted 15 percent of global CO2 emissions. China, the largest polluter, emitted 30 percent. See Global Greenhouse Gas Emissions Data, in GREENHOUSE GAS EMISSIONS, https://www.epa.gov/ghgemissions/global-greenhouse-gas-emissions-data. See Mengpin Ge et al., 6 Graphs Explain the World’s Top 10 Emitters, WORLD RESOURCES INSTITUTE BLOG (Nov. 25, 2014), http://www.wri.org//blog/2014/11/6-graphs-explain-world%E2%80%99s-top-10emitters. Article 2 of the Paris Agreement, the international document that resulted from the 21st United National Framework Convention on Climate Change Conference of the Parties, commits the signatories to “[h]olding the increase in the global average temperature to well below 2˚C above preindustrial levels and pursuing efforts to limit the temperature increase to 1.5˚C above pre-industrial levels, recognizing that this would significantly reduce the risks and impacts of climate change.” See United Nations, Paris Agreement, 2015, http://unfccc.int/files/essential_background/convention/appli cation/pdf/english_paris_agreement.pdf. See U.S. EPA, GREEENHOUSE GAS INVENTORY REPORT: 1990–2014, https://www3.epa.gov/climate change/ghgemissions/usinventoryreport.html. For analyses of potential pathways to achieving 80 percent GHG emissions reductions in the United States by 2050, see ENERGY AND ENVIRONMENTAL ECONOMICS, PATHWAYS TO DEEP DECARBONIZATION IN THE UNITED STATES.

2

Ann Carlson and Dallas Burtraw

and in the climate and energy context we have seen significant attention paid to them. Our focus in this book is, nevertheless, different. Our claim is that despite the recent pause in US efforts to reduce emissions, the United States will need to return to aggressively regulating its greenhouse gas emissions. And in doing so, we will need well-designed energy policy that will need to be not only efficient and fair but also durable in order to achieve substantial reductions in greenhouse gas emissions by midcentury. Most centrally, we will need durable energy policy to motivate the substantial private-sector investment in long-lived energy infrastructure that will be necessary to transform our energy system. Durability alone, however, will not suffice. We also need our policies to be adaptable – to incorporate and respond to new scientific, technological and economic information. Without policy that endures and yet evolves, we will not achieve the long-term transformation in our energy systems that is crucial to stabilizing the earth’s temperatures. Finally, policies to reduce greenhouse gases also need to be as flexible as possible. Flexibility eases the burden of compliance by drawing on emitter knowledge and experience to help determine how best to reduce emissions and by reducing the economic costs of doing so.6 Addressing climate change will be the most expensive and far-reaching environmental effort in history. Flexible policy is likely to improve cost-effectiveness, which can, in turn, speed up emissions reductions, improve the fairness of climate policy and contribute to its political durability. The Clean Air Act (CAA) is currently the central policy tool for regulating greenhouse gases (GHGs) from the energy sector.7 Its use is in many respects by default: efforts to pass comprehensive federal climate legislation have failed, and given the current political climate, these efforts are moribund.8 And the Environmental Protection Agency’s (EPA) regulation of GHGs under the CAA is the result not of an affirmative choice by the agency but in response to a series of lawsuits that have forced EPA to regulate emissions from a variety of sources.9 The Trump administration’s posture toward regulating GHGs under the CAA is dramatically different from its predecessor under President Obama. It includes rolling back the Clean Power Plan, which 6

7

8

9

See U.S. EPA, Building Flexibility with Accountability into Clean Air Programs, in CLEAN AIR ACT OVERVIEW, https://www.epa.gov/clean-air-act-overview/building-flexibility-accountability-clean-airprograms. See U.S. EPA, Air Pollution: Current and Future Challenges, in CLEAN AIR ACT OVERVIEW, https:// www.epa.gov/clean-air-act-overview/air-pollution-current-and-future-challenges. The only bill to pass a single house of the US Congress was the American Clean Energy and Security Act, also known as Waxman-Markey for its authors. See Center for Climate and Energy Solutions, The American Clean Energy and Security Act, https://www.c2es.org/federal/congress/111/acesa. See, e.g., Massachusetts v. EPA, 549 U.S. 497 (2007), St. NY et al. v. EPA, Docket No. 06–01322 (D.C. Cir. Sept 13, 2006).

Introduction

3

was the Obama administration’s program to regulate the electricity sector and a centerpiece of the nation’s climate policy submitted as part of its commitment under the Paris Agreement.10 A recent proposal, not yet finalized, would also freeze combined GHG emissions and fuel economy standards for passenger automobiles and revoke California’s authority to issue its own standards.11 Yet, unless Congress eliminates EPA authority to regulate GHGs, the agency will nevertheless be legally required under the CAA to issue new sets of regulations.12 The CAA is, in many respects, a remarkable statute. Over its almost fifty-year history, the CAA has resulted in large reductions in harmful air pollutants. These reductions have occurred across all areas of the country, across a wide range of pollutants and from a huge number of sources, including in the electricity and transportation sectors, which must be at the heart of long-term climate and energy policy.13 In many respects, the CAA has been at once durable – in that it has continued to produce reductions in air pollution over the course of its long life and has long outlasted the political coalition that led to its adoption; it has been adaptable – in its ability to respond over time to changes in economic, technological and scientific information; and it has been flexible – through the use of incentivebased regulation in lieu of prescriptive regulations in many cases, which has enabled greater pollution reductions at lower cost. To give two examples, the CAA is responsible for the virtually complete elimination of lead over a two-decade period from the transportation sector, leading to widespread public health benefits.14 This was accomplished through durable but evolving regulations that became increasingly stringent in response to an emerging scientific consensus about the effects of lead15 and ultimately employed a flexible market-based approach to achieve the virtual elimination of lead in gasoline.16 Second, the CAA is responsible for dramatic reductions in fine particulate matter,17 again through regulations that adapted in 10

11

12

13 14

15

16

17

See A Review of the Clean Power Plan, 82 Fed. Reg. 16329 (April 4, 2017) (EPA announcement that it will reconsider the CPP), https://www.federalregister.gov/documents/2017/04/04/2017–06522/reviewof-the-clean-power-plan. U.S. EPA, THE SAFER AND AFFORDABLE FUEL EFFICIENT VEHICLES PROPOSED RULE FOR MODEL YEARS 2021–2026, https://www.epa.gov/regulations-emissions-vehicles-and-engines/safer-and-affordable-fuelefficient-vehicles-proposed. For an overview of the regulatory requirements that the Massachusetts v. EPA decision triggered, see Ann Carlson, An Ode to the Clean Air Act, 30 J. LAND USE L 119 (2014). See U.S. EPA, OUR NATION’S AIR: STATUS AND TRENDS THROUGH 2015 (2016). See U.S. EPA, Progress Cleaning the Air and Improving People’s Health, in CLEAN AIR ACT OVERVIEW, https://www.epa.gov/clean-air-act-overview/progress-cleaning-air-and-improving-peoples-health. See Herbert L. Needleman, The Removal of Lead from Gasoline: Historical and Personal Reflections, 84 ENVTL. RES. 20, 20–34 (1999). See Robert W. Hahn & Gordon L. Hester, Marketable Permits: Lessons for Theory and Practice, 16 ECOLOGY L. Q. 361, 380–91 (1989). From 2000 to 2015, the national average of PM10 particle pollution decreased by 36 percent, and the national average of PM2.5 levels decreased by 37 percent. See U.S. EPA, Particulate Matter (PM10) Trends, in AIR TRENDS, https://www.epa.gov/air-trends/particulate-matter-pm10-trends; U.S. EPA,

4

Ann Carlson and Dallas Burtraw

response to new scientific information. The harmful effects of particulate matter were not widely recognized until the 1990s, after the last major amendments to the CAA.18 Once the harmful effects became clear, EPA exercised its mandate to protect human health by naming fine particulate matter as a regulated pollutant and then implementing reductions through regulations that encompassed both flexible and prescriptive approaches.19 Both because of its centrality in environmental regulation and because the CAA is now being used as the principal means for regulating US GHG emissions, this book makes the CAA its centerpiece. Our aim is to determine whether and how the CAA has exhibited durability, adaptability and flexibility – or failed to do so – and to draw lessons for what the CAA can tell us about how policymakers can incorporate mechanisms to ensure a long-lasting but evolutionary and cost-effective approach to cutting GHGs to almost zero. We believe our analysis is helpful not only at the federal level but also for state policymakers focused on regulating carbon emissions from their transportation and electricity sectors and ultimately buildings and industry. Each of the next five chapters focuses on important features of the CAA: the National Ambient Air Quality Standards (NAAQS), the regulation of stationary sources, the regulation of mobile sources, the regulation of transportation fuels, and the use of market mechanisms across a number of the CAA’s programs. In each chapter we set forth and describe in detail the principal regulatory mechanisms used in the program. We then evaluate each program to determine whether and how, through its regulatory means, each has been durable across the life of the statute yet nimble enough to respond to changes in new information about air pollutants, technologies to reduce them and science about the health and welfare effects of air pollution exposure. Moreover, we consider whether the regulations allow for flexibility in implementation and whether the regulatory approach has either enhanced or inhibited compliance. In this introductory chapter, we set forth and define a series of concepts and features that each of the subsequent five case studies uses to frame and analyze the program on which the chapter focuses. We also provide a general overview of the major provisions of the CAA in order to put each of the chapters into context. We also begin to tease out common themes and issues that emerged from the chapters but save for the conclusion a lengthy discussion of the lessons we have collectively learned from the CAA about how to design durable yet adaptable and flexible energy and climate policy.

18 19

Particulate Matter (PM2.5) Trends, in AIR TRENDS, https://www.epa.gov/air-trends/particulate-matterpm25-trends. See National Ambient Air Quality Standards for Particulate Matter, 62 Fed. Reg. 38,652 (July 18, 1997). Id.

Introduction

5

We begin with a discussion of the central concepts on which we focus: durability, adaptability and flexibility. We describe why we believe these attributes are so crucial to climate and energy policy and what we mean in using the terms.

1.1 why durability, adaptability and flexibility? The transportation and electricity sectors, as well as buildings and industrial facilities, are infrastructure intensive and involve large capital investments that often last many decades. We have power plants in the United States that were built over seventy years ago.20 The average age of our nuclear fleet is approaching forty years.21 Cars built today can last for 150,000 miles or more with a typical age of 15 years,22 heavy-duty trucks may stay on the road for thirty years and our rail cars, ships and airplanes remain in operation for decades.23 In order to incentivize innovations in all these capital-intensive industries to produce low or zero GHG emissions by midcentury, it seems almost selfevident that we need stable, durable policies across multiple decades. Stability and durability in policy will provide the signal necessary to investors and innovators to develop technologies and systems that can help accomplish our long-term emissions goals. Stability and durability will also reduce the attendant risk that accompanies investments in the research and development of these technologies. By stability and durability, however, we do not mean that a policy must be fixed for the next forty-five years. Instead, by durability, we mean a policy framework that continues to accomplish the objectives for which it was adopted. In political terms, we argue – borrowing from the work of our colleague Eric Patashnik – that a durable policy is one that remains effective after the coalition that led to its adoption no longer exists or no longer holds the reins of power. The policy, in other words, outlasts its initial supporters. And the policy must accomplish its goals even in the face of changes in scientific knowledge, in technological innovation, and in economic change.24 20

21

22

23

24

See Steven Mufson, Vintage U.S. Coal-Fired Power Plants Now an “Aging Fleet of Clunkers,” WASH. POST, June 13, 2014, https://www.washingtonpost.com/business/economy/a-dilemma-with-aging-coalplants-retire-them-or-restore-them/2014/06/13/8914780a-f00a-11e3-914c-1fbd0614e2d4_story.html? utm_term=.58b4def10fce. See U.S. Energy Info. Admin., How Old Are U.S. Nuclear Power Plants, and When Was the Newest One Built?, in FREQUENTLY ASKED QUESTIONS (June 21, 2017), http://www.eia.gov/tools/faqs/faq.php? id=228&t=21. National Highway Transportation and Safety Administration, Final Regulatory Impact Analysis, Corporate Average Fuel Economy for MY 2017-MY 2025 Passenger Cars and Light Trucks (August, 2012). California Environmental Protection Agency Air Resources Board, Draft Supporting Information for Technology Assessments: Truck and Bus Sector Description VI-2 (2016). ERIC PATSHNIK, REFORMS AT RISK: WHAT HAPPENS AFTER MAJOR POLICY CHANGES ARE ENACTED? (2008).

6

Ann Carlson and Dallas Burtraw

In order to last and maintain its effectiveness for multiple decades, a durable policy must also, then, include mechanisms to adapt to new information about science, technology and economics. We do not, in other words, mean by durability that a policy should remain the same across many years. The world will not remain static over the next several decades; indeed, the rapidity with which the electricity and transportation sectors are transforming both technologically and economically is already outpacing predictions of just a few years ago. But the methods to adapt policy to new information must, we believe, be themselves predictable in order to provide clear signals to regulated industries that policies will change. Put a different way, we need durable yet evolving policy that – through its adaptive mechanisms – is predictable. We recognize that developing policy that exhibits both durability and adaptability may seem contradictory. Yet predictability through regularized administrative processes that allow an agency to adapt to new information can harmonize the two concepts. By anticipating the need for policy to adapt within the domain of an expert agency, the CAA has in many respects provided for this regularized process. Regulated parties are on notice that the CAA will adapt, but only with significant lead time, with an opportunity for broad public participation and with the input of sophisticated scientific and technical experts. This attention to process, we believe, is one key to the CAA’s success: if a policy changes in a way that is not predictable or is not in accord with legal and administrative processes, it is unlikely to provide the stable, durable signal that investors need to make long-term capital commitments. Although one might expect Congress to provide this adaptive response, historically, legislative guidance has occurred only rarely. The modern CAA has been significantly amended only twice since its adoption in 1970 and not at all since 1990.25 Flexibility adds a third element to many successful policies. Major environmental transformation involves the turnover of industrial, commercial and residential capital that takes years. Assuming that the entire economy is not going to shut down until that happens, it is meaningful to ask: what entities are going to make which investments and when? This is a challenging question for regulators, who typically cannot observe the distribution of emissions reduction opportunities and their costs from among a large number of entities that must comply with regulation. Regulators want to get the biggest bang for the buck out of regulations that impose costs on the economy but often lack complete knowledge about where to find the biggest bang. In response to this challenge, the CAA has indeed adapted, through trials in regional offices and ultimately through national programs, to find ways to introduce flexible approaches to 25

See U.S. EPA, Evolution of the Clean Air Act, in CLEAN AIR ACT OVERVIEW, https://www.epa.gov/ clean-air-act-overview/evolution-clean-air-act.

Introduction

7

regulation. A flexible approach will provide incentives to a group of regulated entities to reduce emissions, rewarding individual entities that can reduce the most and operate most cleanly while calibrating the overall effort to meet environmental goals. We see in the following chapters that flexibility has been a key ingredient of some of the most successful regulations under the CAA. However, we also see cases where it has undermined environmental outcomes and threatened the durability of some initiatives. One example in the CAA that ties these three concepts together is the Acid Rain Trading Program, which incorporated an emissions cap that declined in two phases combined with emissions trading among regulated entities.26 This innovative, flexible approach was the political lynchpin to the Acid Rain Program because it promised to achieve the scientifically informed environmental goal at less cost than would mandated specific measures at specific power plants.27 However, once the second phase was completed, the program essentially collapsed because it failed to continue to adapt to new scientific and economic information and technological changes.28 The five CAA case studies that follow have identified a number of means to incorporate new information into long-term policy. Not every section of the CAA contains all these mechanisms, although some apply to the entire Act. Instead, our authors discuss the ways in which a variety of these mechanisms promote – and in some cases undermine – flexible, adaptable, yet durable policy. These mechanisms include requirements that EPA periodically engage in the revision of standards;29 cooperative federalism arrangements with shared responsibility between the federal government and states;30 formal procedures and processes that push EPA to carry out its statutory duties, including citizen suits,31 petition processes and notice and comment;32 expansive definitions in the Act – such as air pollutant – that allow the agency to regulate new pollutants;33 technologybased standards that by definition incorporate evolution into them (terms such as best and lowest);34 automatic tightening of targets, such as in the two phases of

26

27 28

29 30 31 32 33 34

See U.S. EPA, Acid Rain Program, in CLEAN AIR MARKETS, https://www.epa.gov/airmarkets/acid-rainprogram. Id. See Juha Siikama¨ki et al., The U.S. Environmental Protection Agency’s Acid Rain Program, in RESOURCES FOR THE FUTURE (Nov. 2012), http://www.rff.org/files/sharepoint/WorkImages/Download/ RFF-Bck-AcidRainProgram.pdf. 42 U.S.C. § 7408 (2012). 42 U.S.C. § 7410 (2012). 42 U.S.C. § 7604 (2012). 42 U.S.C. § 7607 (2012). 42 U.S.C. § 7602 (2012). 42 U.S.C. § 7479 (2012).

8

Ann Carlson and Dallas Burtraw

the Acid Rain Trading Program;35 and the incorporation of sophisticated personnel, including scientists and economists, into EPA and its advisory boards.36 In some instances, however, programs in the CAA have either failed to adapt, as is the case of the Acid Rain Trading Program in its failure to require reductions in sulfur dioxide beyond the two stages included initially in the program’s provisions,37 or adapted too frequently, as in the case of the Renewable Fuel Standard with its requirement that EPA establish a new target each year for the amount of renewable fuels that refiners must produce.38 This apparent excess of adaptability resulted from the inability of Congress to anticipate technologically feasible goals. Our authors have also examined the interactions among a variety of branches of government, including Congress, the executive (principally EPA), the judiciary and the states in pushing the evolution of air pollution policy. One notable finding is that Congress played a crucial role in the first two decades after the CAA passed in amending portions of the Act that either weren’t working well or were too ambitious. A key example was recognizing that establishing health-based standards for hazardous air pollutants, as required by the 1970 CAA, was simply not working. Instead, Congress replaced the healthbased approach with a technology-based one in the 1990 amendments, leading to a flurry of successful regulatory activity. The lack of congressional involvement in adapting to new circumstances in the subsequent two and one-half decades (with the exception of amendments to fuels programs) has hampered EPA efforts to address remaining pollution problems. A notable example of a policy that would have benefited from congressional clarification is EPA’s attempt to regulate interstate air pollution by using flexible market mechanisms. Had Congress clarified the agency’s ability to do so EPA could have designed better programs, implemented more quickly and cost effectively and with less court involvement. Another important finding of several of the chapters is the key role states have played in pushing the evolution of the statute. The best-known example is California’s unique role in regulating emissions from mobile sources and fuels and decisions by other states to adopt California’s standards. Other important instances include the innovations some states have produced in regulating new stationary sources under the Prevention of Significant Deterioration and Nonattainment programs and the role the Ozone Transport Commission played in forcing EPA to address cross-border ozone pollution.

35 36 37 38

42 U.S.C. §§ 7651c–d (2012). 42 U.S.C. § 7403 (2012). See Siikama¨ki et al., supra note 27. See U.S. EPA, Renewable Fuel Annual Standards, in RENEWABLE FUEL STANDARD PROGRAM, https:// www.epa.gov/renewable-fuel-standard-program/renewable-fuel-annual-standards.

Introduction

9

1.2 why the clean air act? We have already explained some of our reasoning in studying the CAA to determine whether and how it has been both durable yet flexible. We elaborate here. Environmental Performance. First, the CAA has produced tremendous environmental results over its nearly five-decade history in reducing ambient air pollution. All six of the pollutants covered by the NAAQS have declined significantly. Carbon monoxide is down 86 percent on average nationwide. Lead has declined 99 percent. Nitrous oxide has declined 59 percent and sulfur dioxide 84 percent. Ozone, which remains one of the toughest pollutants to control, is down 32 percent. And in the fifteen years since EPA began regulating fine particles (PM2.5), concentrations have fallen 37 percent.39 All these declines have occurred while the US economy has grown dramatically (by sixteen-fold since 1971)40 and population has increased by 150 percent.41 These pollution figures are national averages from all sources for the most ubiquitous pollutants. The chapters on individual programs within the CAA that follow provide more granular data, but the overall conclusion is a powerful one: the Act has produced enormous environmental benefits. The United States will need to achieve similar declines in GHG emissions over a roughly similar period of time, by midcentury, making the CAA an important statute to study. Regulatory Diversity. Another important feature of the CAA is its embrace of a variety of regulatory tools. The centerpiece of the Act is, of course, the NAAQS, which set ambient, health-based standards for ubiquitous pollutants.42 But the NAAQS program is administered through a system of cooperative federalism that harnesses the states to implement and enforce the standards, an innovative arrangement that merits study.43 Moreover, many sources are subject not to health-based standards but to technology-based ones. So are many pollutants, 39

40

41

42 43

These figures come from the US EPA’s Air-Trends reporting. See U.S. EPA, NATIONAL AIR QUALITY: STATUS AND TRENDS OF KEY AIR POLLUTANTS, https://www.epa.gov/air-trends. According to the National Bureau of Economic Analysis, gross domestic product was, when adjusted for inflation, $1.137.8 billion in the first quarter of 1971, when the CAA took effect, and was $18,164.8 billion at the end of 2015 (EPA data about pollutants are through 2015). See U.S. Department of Commerce, Bureau of Economic Analysis, GDP AND PERSONAL INCOME, NATIONAL DATA, http://www.bea.gov/iTable/iTable.cfm?ReqID=9&step=1#reqid=9&step=3&isuri=1&903=5. See U.S. Census Bureau, HISTORICAL NATIONAL POPULATION ESTIMATES: JULY 1, 1900–JULY 1, 1999 (showing 1971 US population as 207.66 million people), http://www.census.gov/popest/data/national/ totals/pre-1980/tables/popclockest.txt; and U.S. Census Bureau, ANNUAL ESTIMATES OF THE RESIDENT POPULATION, APRIL 1, 2010–JULY 1, 2015 (showing 2015 US population as 321.42 million people), http:// factfinder.census.gov/faces/tableservices/jsf/pages/productview.xhtml?src=bkmk. 42 U.S.C. § 7409 (2012). 42 U.S.C. § 7410 (2012).

10

Ann Carlson and Dallas Burtraw

including hazardous air pollutants.44 The CAA’s system of regulating mobile sources is perhaps the most unusual, setting national standards while designating only a single state, California, to set standards more stringent than any federal standards and allowing other states to opt into those standards or follow the federal ones.45 The Act is also not confined to relatively prescriptive regulatory tools (sometimes referred to as command and control). It also includes a large number of flexible market-based programs, from the regulatory approach that led to the removal of lead from gasoline, to the control of interstate air pollution embraced in Title IV’s Acid Rain Trading Program, to the control of cross-state air pollution first through a regional state effort via the Ozone Transport Commission and later expanded to encompass the eastern half of the country.46 And the CAA embraces strong citizen participation through a broad authorization of citizen lawsuits against EPA and the inclusion of important procedural and venue provisions.47 Indeed the CAA was the first federal statute to include a citizen suit provision.48 This regulatory diversity provides a rich basis to study the ways in which durability and adaptability are promoted and/or undermined through the Act’s many tools and for understanding the cases where flexible approaches have worked well. Our chapters reflect this diversity. Notable Shortcomings. The CAA is far from perfect. Several of its provisions and programs have something to teach us from their failures, not their successes. These include well-known failures such as the exemption of many existing sources from requirements to reduce emissions49 and less well-known failures such as the fact that the boutique fuels program has produced virtually no emissions reductions.50 Because the CAA is so far reaching across so many pollutants and sources, it seems virtually inevitable that some of its provisions are ill advised or unsuccessful; some have incurred costs with little in the way of associated health or environmental benefits. Examining these failures for what they can tell us about how not to design durable yet adaptable and flexible policies is at least as important as examining the CAA’s many successes. 44

45 46

47 48

49

50

See U.S. EPA, Setting Emissions Standards Based on Technology Performance, in CLEAN AIR ACT OVERVIEW, https://www.epa.gov/clean-air-act-overview/setting-emissions-standards-based-technologyperformance. 42 U.S.C. § 7543 (2012). See U.S. EPA, Building Flexibility with Accountability into Clean Air Programs, in CLEAN AIR ACT OVERVIEW, https://www.epa.gov/clean-air-act-overview/building-flexibility-accountability-clean-airprograms. 42 U.S.C. § 7604 (2012). See Charles N. Nauen, Citizen Environmental Lawsuits after Gwaltney: The Thrill of Victory or the Agony of Defeat?, 15 WILLIAM MITCHELL L. REV. 327, 328 (1989). See Jonathan Remy Nash & Richard L. Revesz, Grandfathering and Environmental Regulation: The Law and Economics of New Source Review, 101 NW. U. L. REV. 1677, 1709 (2007). See MAXIMILIAN AUFFHAMMER & RYAN KELLOGG, Clearing the Air? The Effects of Gasoline Content Regulation on Air Quality, 101 AM. ECON. REV. 2687, 2688 (2011).

Introduction

11

Regulation of Carbon-Intensive Sources. Many of the largest sources of GHG emissions are also among the largest sources of air pollution. As we have already described, the transportation and electricity sectors contribute more than half of US carbon dioxide emissions.51 Refineries, chemical plants and other major air pollution sources also contribute a significant amount.52 It makes sense to examine how the CAA has or has not been durable, adaptable and flexible when applied to the very sources from which GHGs will need to be regulated. Ability to Withstand Political Change. The longevity of the CAA also demonstrates one of its rather remarkable accomplishments: it has endured through significant ideological and political change as the country has elected both Republican and Democratic presidents throughout the statute’s long history. The CAA was first enacted with broad bipartisan support and signed into law by a Republican president (Nixon), saw major amendments in 1977 under a Democratic president (Carter) and then underwent another round of major amendments in 1990 under Republican President George H. W. Bush. Throughout the forty-nine years since its passage, very different administrators with very different ideological predilections have led the implementing agency. We have recently witnessed another dramatic ideological shift, from eight years of aggressive environmental activism to a period of administrative retrenchment. Yet many of the features of the CAA – from federalism arrangements to statutory deadlines to citizen suits – will likely mean that the statute will continue to move forward. Comprehensive Coverage. Finally, the CAA is interesting to study in part because it is so comprehensive. This comprehensiveness manifests in several ways. The CAA contains broad definitions that extend its coverage as information emerges about the harmful effects of pollutants. It contains multiple and sometimes overlapping methods of regulating sources of pollution so that few sources escape its reach. The CAA uses more than one level of government to regulate, which can promote innovation and provide something of a regulatory backstop if one level of government is underutilizing its powers. It gives a large role to private parties, who can participate via citizen suit, notice and comment and petition. The Act provides an expert agency, EPA, with broad powers through delegation, but also for some programs it requires EPA to act through regulatory deadlines. In some instances, after experience showed that a grant of delegated power to EPA (or in some instances to the states) had proved too broad, Congress has stepped in and regulated more directly. Some parts of the CAA, then, are very prescriptive. 51

52

See U.S. EPA, INVENTORY OF U.S. GREENHOUSE GAS EMISSIONS AND SINKS 1990–2015 (Apr. 15, 2017), https://www.epa.gov/ghgemissions/inventory-us-greenhouse-gas-emissions-and-sinks-1990-2015. Id.

12

Ann Carlson and Dallas Burtraw

Chapters 2 through 6 provide remarkably detailed and in-depth analyses of individual programs within the CAA. They are invaluable in examining how these individual programs promote or undermine durability, adaptability and flexibility, and we describe them next. In our concluding chapter, we step back and look at these programs collectively, keeping sight of the comprehensiveness of the CAA and drawing lessons from the chapters as we look to future pathways for the development of energy and climate policy.

1.3 living examples in the clean air act Chapter 2, by law professor William Boyd, focuses on the NAAQS to show the ways in which the centerpiece of the CAA has produced significant pollution reductions of major pollutants across multiple decades. This chapter also sets the stage for those that follow given the centrality of the NAAQS in the way the CAA operates. Chapter 3, by law professor Hannah J. Wiseman, then examines the regulation of factories and industrial plants, known as stationary sources, that emit pollutants regulated by the NAAQS as well as stationary sources that emit hazardous air pollutants. Her comprehensive account of the many regulatory programs under the CAA shows both real successes in establishing durable yet adaptable policy and some significant – and obvious – failures. Chapter 4, by political scientist Barry Rabe, turns its attention to the regulation of mobile sources (passenger cars, trucks, off-road vehicles and so forth). Professor Rabe uses the unique federalism arrangement the CAA incorporates as his central focus. He shows that this arrangement – which gives California special authority to regulate emissions from mobile sources while preempting all other state authority to regulate separately – has been among the most durable, yet adaptable, provisions of the CAA. In Chapter 5, economist Joseph E. Aldy looks at the CAA’s regulation of fuels. The story of durability and flexibility is perhaps most mixed in these programs. The CAA has produced some great successes in the regulation of fuels – in eliminating lead from gasoline – and one of its greatest failures to date – in the attempt to produce renewable fuels. The failure in the case of the renewable fuels program may be because it is simply too adaptable, producing no clear signal to fuel producers about what will be expected of them year to year. And finally, in Chapter 6, political scientist Eric M. Patashnik steps away from examining specific programs and offers an account of a particular regulatory approach – the incorporation of flexible market-based tools – that cuts across multiple parts of the CAA. Two of his most interesting conclusions are that several market-based programs have proven to be less durable than more traditional regulatory programs and that the CAA has not always provided

Introduction

13

EPA with the authority necessary to design market-based programs effectively. Other market-based programs have, however, proven to be more successful in promoting our metrics, and this chapter provides important insights about how and why. We conclude, in Chapter 7, with broad lessons about our focus – durability, adaptability and flexibility – that we derive from the five substantive chapters.

1955

Clean Air Act of 1963 Federal research on air pollution; grants to states; authorizes abatement conferences for interstate pollution Clean Air Act Amendments of 1970 establishes NAAQS EPA revokes secondary NAAQS standard for SO2 CAA Amendments of 1977 establishes CASAC; codifies PSD and Non-Attainment programs

EPA removes Hydrocarbons from list of criteria pollutants

1960

1965

Air Quality Management Act of 1967 establishes air quality management approach; directs states to develop air quality criteria and standards; authorizes creation of air quality control regions

1970 EPA issues first NAAQS for six criteria pollutants 1975

EPA required to list Lead as criteria pollutant EPA issues NAAQS for Lead

1980

EPA revises NAAQS for Ozone

1985

CAA Amendments of 1990 strengthens connections between the NAAQS and 1990 other CAA programs; new provisions added to NonAttainment programs 1995 Supreme Court in Whitman v. American Trucking holds that EPA may not consider costs in setting NAAQS

California begins developing first modern air pollution control regime

EPA issues revised NAAQS for PM10

EPA issues revised NAAQS for Ozone and PM10 and new NAAQS for PM2.5

2000

EPA issues revised NAAQS 2005 for PM10 and PM2.5 EPA issues revised NAAQS for Lead and Ozone

EPA initiates major review of NAAQS process,creating more formal framework for review

2010

EPA issues revised NAAQS for SO2 and NOx

2015

EPA issues revised NAAQS for Ozone

EPA issues revised NAAQS for PM2.5

2 The Clean Air Act’s National Ambient Air Quality Standards A Case Study of Durability and Flexibility in Program Design and Implementation William Boyd 2.1 introduction The Clean Air Act’s National Ambient Air Quality Standards (NAAQS) can make a strong claim to being the most ambitious and successful major program in US environmental law. It is the “engine” that drives much of the Clean Air Act (CAA), the nation’s flagship environmental statute, touching most of the other major provisions of the Act.1 It embodies the very notion of cooperative federalism and mobilizes substantial resources from the states in the fight against air pollution. And since 1970, it has, in the aggregate, delivered huge benefits for environmental quality and public health. Put simply, air quality in the United States today is far better than it was in 1970, largely because of the NAAQS, with cumulative public health benefits measured in the trillions of dollars, despite significant growth in population and economic activity.2 And the program is still going strong, with continued growth and elaboration even as it enters middle age. Since establishing the NAAQS as the centerpiece of the CAA in 1970, for example, Congress has stepped in on two major occasions (1977 and 1990) to strengthen the program and its connections to other parts of the Act. The Environmental Protection Agency (EPA) has likewise developed elaborate internal policies and procedures for managing the NAAQS program and continues to devote substantial resources to major rulemakings under the NAAQS. Ongoing independent scientific reviews have come to provide an integral part of the continuous review and revision of the NAAQS. And, of course, the federal courts have played a fundamental role in shaping the program – often in response to citizen suits 1

2

See, e.g., Whitman v. American Trucking, 531 U.S. 457, 468 (2001) (observing that “§109(b)(1) and the NAAQS for which it provides are the engine that drives nearly all of Title I of the [Clean Air Act]”). See, e.g., U.S. EPA, THE BENEFITS AND COSTS OF THE CLEAN AIR ACT FROM 1990 TO 2020, (2011); U.S. EPA, THE BENEFITS AND COSTS OF THE CLEAN AIR ACT, 1970 TO 1990 (1997). See also J. Bachmann, Will the Circle Be Unbroken: A History of the U.S. National Ambient Air Quality Standards, 57 J. AIR & WASTE MGM’T ASSOC. 652, 692 (2007).

16

William Boyd

and petitions making use of important procedural opportunities embedded in the statute.3 The result is a strong science-based program that is durable, adaptive and flexible and one that depends on extensive cooperation with the states. But the story of the NAAQS is not one of unalloyed success. Despite significant improvements in air quality since 1970, air pollution continues to kill large numbers of people in the United States. Although estimates vary, recent studies have put the number of premature deaths resulting from air pollution in the United States at over 100,000 per year – roughly three times the number of people who die from handguns or automobile accidents.4 Most of these deaths result from so-called criteria pollutants – that is, pollutants that are widespread and produced by numerous and diverse sources and are the targets of the NAAQS program.5 In particular, despite substantial progress in reducing pollution and meeting the NAAQS for several criteria pollutants across most of the country, a number of areas around the United States continue under serious and even severe nonattainment for ozone and fine particulates (PM2.5), the two major sources of premature death, in part because of the ongoing difficulties of dealing with mobile source emissions as well as the challenges created by interstate transport of these pollutants and their precursors.6 This problem of interstate “downwind” pollution, in fact, has arguably been made worse by certain features of the NAAQS program, and it stands as a stark reminder that the program, despite its successes, needs to be revised and adapted yet again to deal effectively with the nation’s air pollution problems.7 3

4

5

6

7

See Elizabeth Fisher et al., Rethinking Judicial Review of Expert Agencies, 93 TEX L. REV. 1681, 1689 (2015) (noting that “[j]udicial review of the EPA’s ambient air quality standards offers one of the longest histories of judicial review of agency science, dating to the 1970s”). See, e.g., Neal Fann et al., Estimating the National Public Health Burden Associated with Exposure to PM 2.5 and Ozone, 32 RISK ANALYSIS 81, 92 (2013) (finding that between 130,000 and 340,000 premature deaths are attributable to PM2.5 and ozone using ambient measurements (2005) and nonanthropogenic background PM2.5 and ozone concentrations simulated by atmospheric chemistry models and a health impact function); Christopher J. L. Murray et al., The State of U.S. Health, 1990–2010: Burden of Diseases, Injuries and Risk Factors, 310 JAMA 591, web appendix table 8 (2013) (reporting 110,000 premature deaths in 2010 [all ages] from PM and ozone pollution); Fabio Caiazzo, Air Pollution and Early Deaths in the United States, Part I: Quantifying the Impact of Major Sectors in 2005, 79 ATMOS. ENV’T 198 (2013) (reviewing studies). The 1970 Act defined these criteria pollutants as pollutants that have “adverse effects on public health and welfare” and “result from numerous or diverse mobile or stationary sources.” Section 108, 84 Stat. 1676, 1678. These two pollutants (PM2.5 and tropospheric ozone) are responsible for the vast majority of premature deaths and illnesses attributed to air pollution in the United States. See, e.g., Richard L. Revesz, Federalism and Interstate Environmental Externalities, 144 U. PA. L. REV. 2341, 2349 (1996) (arguing “that the ambient and emissions standards . . . which form the core of the Clean Air Act, are an ineffective and poorly targeted means for dealing with the problem of interstate externalities . . . [and that] these provisions may have exacerbated the interstate spillover problem”). EPA’s recent efforts to use its FIP authority combined with the good neighbor provisions to deal with this problem through its Cross State Air Pollution Rule (also known as the Transport Rule) are obviously an important step, but such efforts are constrained by the statutory text and, more importantly, have taken decades of work. See Federal Implementation Plans: Interstate Transport of Fine Particulate Matter and Ozone and Correction of SIP Approvals, 76 Fed. Reg. 48,208 (Aug. 8, 2011)

The Clean Air Act’s National Ambient Air Quality Standards

17

In terms of instrument choice, the NAAQS program operates as a hybrid. Health-based ambient environmental standards make up the core of the program; they apply uniformly across the country and are intended to limit the concentrations of criteria pollutants to levels that are determined by the EPA administrator as necessary to protect public health with an adequate margin of safety. The Act mandates an ongoing five-year review of the science pertaining to each of the criteria pollutants and requires the administrator to revise the NAAQS as appropriate. These ambient environmental standards are then supplemented with source- and area-specific technology and performance standards that are tied to the attainment or nonattainment status of particular regions. Mobile source standards and various boutique fuels programs provide additional measures for controlling criteria pollutants.8 States are responsible for implementation and are required to submit state implementation plans (SIPs) that demonstrate how the NAAQS will be achieved in their jurisdictions. Once approved by EPA, these SIPs become federally enforceable. If a state fails to submit a SIP or EPA deems a SIP to be inadequate, EPA can impose its own Federal Implementation Plan or FIP for the state. EPA also has authority to withhold highway funds as a sanction for inadequate or incomplete SIPs. Needless to say, the exercise of establishing the NAAQS and the effort to implement the reductions necessary to achieve them can be quite challenging, requiring large scientific and technical investments at both federal and state levels.9 Taken as a whole, the NAAQS program embodies all the design principles identified in Chapter 1 as essential to durability and flexibility. The strong healthbased standard for ambient air quality provides a clear signal to the states and to sources of air pollution. The statutory requirement of an ongoing five-year review of the science provides a mechanism for continuous revision based on new information. The NAAQS program has been subjected to systematic evaluation by Congress, EPA and various independent scientific bodies over the years, which has led to a number of revisions and improvements in the program. And the program’s environmental and public health outcomes are greatly valued and perceived as such. Indeed, few, if any, programs in US environmental law have delivered more in terms of public health benefits. This chapter provides an overview of the NAAQS, with particular attention to the provisions and features of the program that contribute to flexibility and durability. Section 2.2 briefly discusses the problem of ambient air pollution. Section 2.3 describes the complex NAAQS regulatory regime, including the basic

8 9

[hereinafter Cross-State Air Pollution Rule (CSAPR)]; EPA v. EME Homer City Generation, L.P., 134 S. Ct. 1584 (2014) (reversing the DC Circuit and upholding EPA’s Cross-State Air Pollution Rule). Chapters 4 (mobile sources) and 5 (fuels) address the regulation of mobile sources. See, e.g., Bachmann, Will the Circle Be Unbroken, supra note 2 at 652–53 (noting the “extraordinary level of technical and scientific information needed to establish effect-based ambient targets, measure key pollutants, inventory sources and emissions, develop and estimate costs for alternative control scenarios, and forecast and assess results”).

18

William Boyd

statutory framework, EPA’s core responsibilities, the important role of independent scientific review, citizen suits and public participation and the monitoring and modeling infrastructure that supports the program. Section 2.4 discusses the goals of the program and the features that allow for signaling to the states and affected sources. Section 2.5 describes the structure of cooperative federalism and the critical role of the states in implementation. Section 2.6 provides a more detailed look at flexible implementation, with particular attention to statutory provisions, agency procedures and practices and the SIP process. Sections 2.7 and 2.8 look briefly at environmental performance and economic impacts, respectively. Finally, Section 2.9 discusses broader issues of political economy, durability and flexibility. Several key lessons emerge from the case study. First, history matters. The NAAQS program was able to mature and benefit from two decades of robust functional interactions between Congress, EPA and the federal courts. Many of the key provisions of the program that contribute to its durability and flexibility were added, revised and strengthened in subsequent rounds of amendments by Congress. Without this, the program would not be nearly as effective as it is today. Second, the structure and design of the program have contributed to its overall durability in part by building in provisions for flexibility and adaptation at multiple levels – all in the context of a long-term commitment to continuous review and revision of existing standards. By design, the NAAQS program is always in motion, but the goals and overall direction of the program are clear, and there are procedural mechanisms in place that continue to push the program forward. Third, process has been fundamental in bolstering the credibility of the program and strengthening its results – from EPA’s internal NAAQS review process, to the formal and informal roles of independent scientific review, to the multiple opportunities for public participation. Taken together these process features have facilitated important signaling to regulated entities and the states. They have served to ventilate, vet and strengthen proposed revisions to the NAAQS. And they have proven to be more than sufficient to satisfy judicial review in most cases. In sum, the success of the NAAQS program has derived in large part from a combination of history, structure and process. There are important lessons for energy and climate policy here, but perhaps one of the most important lessons is that the task of building a durable, adaptive and flexible policy cannot be reduced to a set of simple design choices. Complicated programs such as the NAAQS are more than the sum of their parts. Deriving lessons from them for future policies thus requires understanding them as a whole and how they have evolved over time.

2.2 the problem of ambient air pollution Decades of research across multiple disciplines have revealed significant associations between exposure to ambient air pollution and negative human health effects,

The Clean Air Act’s National Ambient Air Quality Standards

19

including respiratory ailments, cardiovascular disease and premature death.10 Although ambient air pollution is typically a complex mixture, consisting most often of particulates, ground-level ozone, oxides of nitrogen and sulfur dioxide (among others), most health effects studies have focused on individual pollutants. While the majority of this research has been conducted in Europe and North America, the associations between health effects and ambient air pollution are consistent around the world. Put simply, ambient air pollution is a major cause of disease and premature death all over the world. One recent study, for example, estimated that exposure to fine particulates (PM2.5) resulted in 4.1 million premature deaths globally in 2016 and found that 95 percent of the world’s population lived in areas that exceed World Health Organization (WHO) guidelines for fine particulates (PM2.5).11 As part of its responsibility to establish national ambient air quality standards, EPA conducts extensive reviews of the scientific literature on the health effects of the various criteria pollutants regulated under the NAAQS program. These integrated science assessments have revealed a host of specific health effects associated with exposure to individual criteria pollutants and provide the scientific basis for establishing health-protective standards. In its most recent integrated science assessment for particulates (2009), for example, EPA concluded that long-term exposure to PM2.5 is associated with premature death, heart attacks, irregular heart beat, and respiratory problems such as aggravated asthma and decreased lung function.12 Likewise, in its most recent assessment for ozone (2013), EPA concluded that short- and long-term exposure to ozone is associated with respiratory effects, cardiovascular effects and premature death.13 New evidence of the health effects of ambient air pollution is also accumulating on a near-continuous basis. This includes advances in our understanding of known 10

11

12

13

See, e.g., J. Lepeule et al., Chronic Exposure to Fine Particles and Mortality: An Extended Follow-Up of the Harvard Six Cities Study from 1974 to 2009, 120 ENV. HEALTH PERSP. 965 (2012); M. Turner et al., Long-Term Ozone Exposure and Mortality in a Large Prospective Study, 193 AM J. RESP. CRIT. CARE MED 1134 (2016); M. Jarrett et al., Long-Term Ozone Exposure and Mortality, 360 N. ENGL. J. MED. 1085 (2009); M. L. Bell et al., Ozone and Short-Term Mortality in 95 US Urban Communities, 1987–2000, 292 JAMA 2372 (2004). See HEALTH EFFECTS INSTITUTE, STATE OF GLOBAL AIR 2018: A SPECIAL REPORT ON GLOBAL EXPOSURE TO AIR POLLUTION AND ITS DISEASE BURDEN 1-3 (2018). Another recent study from the World Health Organization (WHO) estimated that exposure to ambient PM2.5 pollution resulted in more than 3 million premature deaths globally in 2012. See WHO, AMBIENT AIR POLLUTION: A GLOBAL ASSESSMENT OF EXPOSURE AND BURDEN OF DISEASE 40 (2016). This is an estimate that does not include the even larger number of deaths from indoor air pollution (estimated at more than 4 million per year). According to the WHO report, “[a]ir pollution represents the biggest environmental risk to health” in the world, with one out of every nine deaths globally the result of “air pollution related conditions.” Id. at 15. U.S. EPA, FINAL REPORT: INTEGRATED SCIENCE ASSESSMENT FOR PARTICULATE MATTER, EPA/600/R-08/ 139F (2009). U.S. EPA, FINAL REPORT: INTEGRATED SCIENCE ASSESSMENT OF OZONE AND RELATED PHOTOCHEMICAL OXIDANTS, EPA/600/R-10/076F (2013).

20

William Boyd

human health effects of air pollution such as respiratory and cardiovascular effects, as well as research on possible new effects. One recent study, for example, suggests that exposure to particulate matter in older women may contribute to the acceleration of brain aging and Alzheimer’s disease.14 Evidence is also mounting regarding other adverse neurological effects, including neurodevelopmental disorders in children, as a result of chronic exposure to ambient air pollution.15 And there is a nascent body of research on the confluence of health risks resulting from exposure to multiple pollutants, raising important questions regarding the additional health gains that might come from a multipollutant approach to air quality management.16 While the scientific understanding of the human health effects of ambient air pollution is considerable (and growing all the time), there is still much more to know about the links between air pollution and public health, including both existing and wellunderstood health endpoints as well as more subtle ones. As we learn more about these various health impacts, moreover, it raises the bar for regulation. Regulating ambient air pollution, either on a single- or multipollutant basis, thus poses a host of challenges. Health-based approaches must confront a constantly expanding body of scientific information on health effects, some of them novel. Any effective approach to pollution control must contend with the fact that ambient air pollution is widespread, affecting very large populations, and is produced by a wide range of sources and activities. Most of the major pollutants are also subject to complicated regional and even international transport dynamics that create additional challenges of upwind contributions to diminished air quality in downwind regions. Put simply, controlling these common pollutants and protecting ambient air quality require regulation of millions of mobile and stationary sources across vast areas combined with careful attention to regional transport – all in the context of constantly evolving scientific literature on health impacts.

14

15

16

See M. Cacciottolo, Particulate Air Pollutants, APOE Alleles and Their Contributions to Cognitive Impairment in Older Women and to Amyloidogensis in Experimental Models, 7 TRANSL. PSYCH. 1 (2017). See, e.g., X. Xu et al., A Review of Epidemiological Research on Adverse Neurological Effects of Exposure to Ambient Air Pollution, 4 FRONT. PUB. HLTH. 1, 1 (2016) (concluding based on a review of existing epidemiological research that there is “mounting evidence implicating adverse effects of air pollution on neurobehavioral function in both adults and children”); L. Calderon-Garciduenas et al., Air Pollution and Detrimental Effects on Children’s Brains, 8 FRONT. HUM. NUEROSCI. 1, 1 (2014) (concluding that “there is enough evidence supporting the perspective that the effects of air pollution on brains of children and teens ought to be key public health targets”). See, e.g., N. Fann et al., Characterizing the Confluence of Air Pollution Risks in the United States, 9 AIR QUAL. ATMOS. HLTH. 293, 296–99 (2016) (noting the potential for criteria pollutants and air toxics in air pollution mixtures to behave synergistically and identifying counties in the United States with elevated levels of criteria pollutants and air toxics); K. Wesson et al., A Multi-Pollutant, Risk-Based Approach to Air Quality Management: Case Study for Detroit, 1 ATMOS. POLLUTION RES. 296, 303 (2010) (finding that a multipollutant approach focused on PM, ozone and hazardous air pollutants in the Detroit urban core resulted in greater net health benefits and was more cost-effective than the standard single pollutant approach).

The Clean Air Act’s National Ambient Air Quality Standards

21

Given these characteristics, any successful regulatory approach to the problem of ambient air pollution must be both durable and flexible. Because ambient air pollution is a long-term, persistent problem that affects vast areas of the country and implicates multiple economic sectors, it is not the kind of problem that will one day be “solved.” Rather, it requires a constant, ongoing effort. A successful program thus needs to send clear signals regarding reduction goals, commitments of resources and affected sources of emissions in order to channel pollution control investments in a manner that will allow for long-term maintenance and improvement of air quality. But such a program also needs to have sufficient flexibility to adjust to new science, to tailor emissions-reduction requirements to local and regional circumstances and to adapt to a changing political-economic context. The NAAQS program has, for the most part, performed well on these fronts, combining durability and flexibility in a manner that has led to major improvements in ambient air quality. But there is still significant work to be done if the program is going to achieve its original goal of attaining the NAAQS for all criteria pollutants for all areas of the country.

2.3 the naaqs regulatory regime The NAAQS program has evolved considerably since it was established in 1970. This section reviews the history and basic elements of the program. It emphasizes specific features of the NAAQS program that contribute to its durability and flexibility, as well as some of the persistent challenges facing efforts to attain some of the NAAQS. The overall conclusion is that the NAAQS program – though hardly perfect – has successfully combined durability and flexibility as part of an effective science-based approach to controlling ambient air pollution. As discussed in more detail later, this combination of durability and flexibility is a product of several specific design features, creative use of administrative process and the fact that the program was able to develop and mature based on two decades of active engagement and adjustment by Congress, EPA and the courts. 2.3.1 The Statutory Framework The basic statutory framework for the NAAQS was established in the 1970 Clean Air Amendments.17 As noted, Congress modified the program in 1977 and again in 1990 on the basis of strong bipartisan majorities.18 In both cases, Congress was more 17 18

Clean Air Amendments of 1970, P.L. 91–604, 84 Stat. 1676 (1970). The Conference Report containing the 1977 amendments was adopted by voice vote in both chambers. The Conference Report containing the 1990 amendments passed the Senate by a vote of 89 to 10 and the House by a vote of 401 to 25. See Policy Tracker: Clean Air Act and Air Pollution, CONGRESSIONAL QUARTERLY , https://library.cqpress.com/cqalmanac/document.php?id=cqal901112490.

22

William Boyd

prescriptive than it had been in 1970, creating additional programs and requirements for both EPA and the states in their respective efforts to achieve the NAAQS and cabining some of the flexibility of the original program. During its first twenty years (1970–90), the NAAQS program was subject to robust interactions between Congress, EPA and the federal courts, with the states playing a fundamental role in implementation. Since 1990, however, other than some efforts with respect to fuels (as discussed in Chapter 5), Congress has been largely absent from review, evaluation and revision of the program. This has left EPA and the courts to adapt the program within the constraints of the existing statutory framework to deal with ongoing challenges such as interstate transport of criteria pollutants and their precursors. As discussed in more detail later, not all of these efforts have been successful, and it is unclear whether the program can respond effectively to persistent challenges such as interstate pollution in the absence of new legislation. 2.3.1.1 Establishing the NAAQS Program Although the idea of ambient air quality standards as a tool for controlling air pollution has been around since the 1930s, the first regulatory approaches that made use of such standards took shape in California during the 1950s.19 In 1955, Los Angeles developed a set of regulations based on ambient air quality standards in an effort to combat the region’s growing smog problem.20 Four years later, the state of California expanded the approach to the whole state and began to develop the first modern air pollution control regime.21 California, however, was an outlier. Air pollution control efforts in other states during the 1950s and 1960s were, for the most part, woefully inadequate.22 During this time, the federal government played a convening role and took some modest steps to support nascent state pollution control efforts. The 1963 Clean Air Act, for example, established a grants program for state air pollution control agencies and provided for abatement conferences to address interstate air pollution 19

20

21 22

See Bachmann, Will the Circle Be Unbroken, supra note 2 at 661; Harold W. Kennedy, The Legal Aspects of Air Pollution Control with Particular Reference to the County of Los Angeles, 27 S. CAL. L. REV. 374 (1954). SCOTT HAMILTON DEWEY, DON’T BREATHE THE AIR: AIR POLLUTION AND U.S. ENVIRONMENTAL POLITICS, 1945–1970 (2000) chap. 3. Bachmann, Will the Circle Be Unbroken, supra note 2 at 661. According to a 1962 report from the US Public Health Service, some 60 percent of the nation’s population (approximately 107 million people) lived in areas with air pollution problems, 43 million in areas with “major” problems. Although the number of people living in such problem areas had increased by some 23 million people since 1950, only seventeen states had air pollution programs with expenditures of $5,000 or more per year. Depending on how one interpreted enforcement, only four to six states actually enforced pollution regulations. See Jean J. Schuenenman, Air Pollution Problems and Control Programs in the United States, paper no. 62-84, U.S. Dept of Health Education and Welfare Public Health Service, Cincinnati, OH, 1 April 1962).

The Clean Air Act’s National Ambient Air Quality Standards

23

problems.23 The 1963 Act also directed the Secretary of Health, Education and Welfare to compile “criteria” summarizing current scientific knowledge on pollutants present in the air at harmful concentrations.24 Congress took further action in 1965, calling for the first federal air pollution standards for mobile sources, and again in 1967, with the Air Quality Act of 1967.25 Among other things, the 1967 Act provided a more explicit endorsement of the air quality management approach underway in California, calling for the designation of Air Quality Control Regions across the country and directing the states to develop tailored, regionally specific air quality standards based on criteria documents being prepared by the Department of Health, Education and Welfare.26 By all accounts, the 1967 Air Quality Act achieved very little.27 Without any ability to force the states to act, little progress was made in establishing (much less enforcing) air quality standards. Moreover, as growing public interest in the environment spilled over into electoral politics, the 1967 Act was soon overtaken by events. With Senator Muskie and President Nixon competing to prove their environmental bona fides, support grew for strong federal legislation on air pollution.28 The result was the Clean Air Amendments of 1970, passed by a unanimous Senate and only a single “No” vote in the House and signed into law by President Nixon on December 31, 1970.29 The 1970 legislation provided for a much stronger federal role in air pollution control and established the basic framework of what we now know as the Clean Air Act – the centerpiece of which was the National Ambient Air Quality Standards (NAAQS).30 In contrast to the 1967 Act’s preference for standards that would vary by region, the new NAAQS program would be based, as the name suggested, on a set of nationally uniform standards established by the EPA administrator and subject to ongoing federal supervision and enforcement. Concentrations of criteria 23 24

25

26

27

28

29

30

See section 4–5, 77 Stat. 392, 393–99. See section 3(c)(2), 77 Stat. 392, 395. This provided the basis for what later came to be identified as the criteria documents (and the criteria pollutants) under the NAAQS program. See Motor Vehicle Air Pollution Control Act, P.L. 89-272, 79 Stat. 992 (1965). See Air Quality Act of 1967, P.L. 90-148, 81 Stat. 485 (1967). See section 107, 81 Stat. 485, 490–91 (air quality control regions and air quality criteria); section 108, 81 Stat. 485, 491–97 (air quality standards and abatement of air pollution). These original criteria documents prepared by HEW provided the basis for the first NAAQS established in 1971. By 1970, fewer than three dozen air quality regions had been designated, as compared with an anticipated number in excess of 100, and not a single state had developed a full pollution control program. Bachmann, Will the Circle Be Unbroken, supra note 2 (recounting history). During this time (the late 1960s), there was a vigorous debate over the relative merits of ambient air quality standards versus uniform technology standards. The 1970 amendments adopted both approaches. The Senate bill passed by a 73–0 roll-call vote. A weaker House version of the bill passed by a 374–1 roll-call vote. In conference, the Senate version was adopted. See Policy Tracker: Clean Air Act and Air Pollution, CONGRESSIONAL QUARTERLY, https://library.cqpress.com/cqalmanac/document.php?id=c qal70-1293712. Clean Air Amendments of 1970, P.L. 91-604, 84 Stat. 1676 (1970).

24

William Boyd

pollutants would be established at levels necessary to protect the public health and welfare. The statute called for two types of standards: primary standards necessary to protect the public health with an adequate margin of safety and secondary standards to protect public welfare from known or anticipated adverse air pollution effects.31 EPA had a number of new duties under the program, including listing of criteria pollutants and preparation of criteria documents under section 108, establishing NAAQS for criteria pollutants under section 109, periodic review and revision of the NAAQS and review and approval of SIPs.32 The whole program was based on a model of cooperative federalism, and the states were given significant flexibility under section 110 to determine how the NAAQS would be attained in their Air Quality Control Regions.33 Among the most important and innovative provisions included in the 1970 amendments were those pertaining to citizens’ suits and public participation.34 These have provided an important additional check on agency behavior and created ample opportunities for judicial review.35 Congress also established specific statutory deadlines for attainment of the NAAQS. All states were expected to be in attainment with the NAAQS within five years, a deadline that would prove to be wildly optimistic and would have to be extended in future legislation.36 2.3.1.2 1977 Amendments Congress amended the CAA in 1977, providing a number of substantial revisions to the NAAQS program. Specifically, Congress extended the attainment deadlines.37 It changed the previous open-ended requirement of “periodic review” of the NAAQS to a mandatory five-year review process.38 The 1977 amendments also required EPA to establish a new independent Clean Air Science Advisory Committee (CASAC) that would play a formal role in the NAAQS review process going forward – an institutional innovation that would prove to be very important in reinforcing the credibility and legitimacy of the NAAQS program in the years ahead.39 31 32 33 34 35

36

37

38 39

Section 109, 84 Stat. 1676, 1679-80. Sections 108, 109, 110, 84 Stat. 1676, 1678–83. Section 110, 84 Stat. 1676, 1680–83. Section 304 (citizen suits), 84 Stat. 1676, 1706–7; Section 307 (judicial review), 84 Stat. 1676, 1707–8. These were the first such provisions in US federal environmental law, but they drew on past experience in the states. See Sections 110(a)(2)(A) (calling for SIPs that provide for attainment within three years) and 110(e)(1) (allowing for a two-year extension of attainment deadline in specific cases), 84 Stat. 1676, 1680 and 1682. Deadlines for attainment were extended by the 1977 amendments to 1982 in most cases and 1987 for areas in severe nonattainment for certain pollutants. See Section 172(a), 91 Stat. 685, 746–47. Section 109(d)(1), 91 Stat. 685, 691. Section 109(d)(2), 91 Stat. 685, 691.

The Clean Air Act’s National Ambient Air Quality Standards

25

The 1977 amendments also enhanced the SIP process, providing for a bifurcated state planning process depending on whether particular Air Quality Control Regions were in attainment or not. States were now required to classify their Air Quality Control Regions as “attainment,” “nonattainment” or “unclassifiable” for each of the NAAQS, and various controls and requirements applied depending on that status. Perhaps most significantly, the 1977 amendments codified and expanded EPA’s Prevention of Significant Deterioration (PSD) program and created a new nonattainment program.40 Among other things, the PSD and nonattainment programs provided for preconstruction review and permitting for new and modified sources that imposed technology controls on these sources depending on their attainment status. Together these became known as the New Source Review (NSR) program. Going forward, states with nonattainment areas would be required to adopt measures (reflected in their SIPs) for both existing and new and modified sources (see Chapter 3).41 Similarly, under the PSD program, state plans were required to include a preconstruction review and permitting program that imposed best available control technology (BACT) on new and modified sources in “attainment” and “unclassifiable” areas. The PSD program also established “increments” of allowable air quality deterioration over a baseline concentration based on classifications of Air Quality Control Regions.42 Congress also took aim in the 1977 amendments at the problem of interstate air pollution – a reaction in part to EPA’s almost complete lack of attention to the problem under the original 1970 provisions.43 First, Congress strengthened the socalled good neighbor provision in section 110, requiring that SIPs contain provisions 40

41

42

43

See Title I, Part C: Prevention of Significant Deterioration of Air Quality, 91 Stat. 685, 731–45; Title I, Part D: Plan Requirements for Nonattainment Areas, 91 Stat. 685, 746–51. Existing sources were subject to “reasonably available control technologies” (RACTs). New and modified sources were subject to “lowest achievable emissions reductions” (LAERs) plus a requirement to “offset” any additional increments of pollution. See Title I, Part D: Plan Requirements for Non-Attainment Areas, 91 Stat. 685, 746–51. See Title I Part C: Prevention of Significant Deterioration of Air Quality, 91 Stat. 685, 731–42. The PSD program has an interesting history that illustrates the role of the federal courts in driving certain developments under the Act. The program was initially created out of whole cloth by DC District Court Judge John Pratt in 1972 – a decision widely viewed as having no defensible basis in the operative provisions of the statute but one that was upheld on appeal. See Sierra Club v. Ruckelshaus, 34 F. Supp. 253 (D.C. Dist. 1972), aff’d per curiam 4 ERC 1815 (D.C. Cir. 1972), aff’d by an equally divided court, sub nom. Fri v. Sierra Club, 412 U.S. 541 (1973); see also A. STANLEY MEIBURG, PROTECT AND ENHANCE: “JUDICIAL DEMOCRACY” AND THE PREVENTION OF SIGNIFICANT DETERIORATION OF AIR QUALITY (1991) (tracing history); R. SHEP MELNICK, REGULATION AND THE COURTS: THE CASE OF THE CLEAN AIR ACT (1983) 71–112 (discussing evolution of the PSD program); Richard Stewart, Judicial Review of EPA Decisions, 62 IOWA L. REV. 713, 741–50 (1977) (criticizing decision on various grounds). Pursuant to Judge Pratt’s injunction, EPA issued final regulations in 1974 and 1975 that were upheld in 1976 by the DC Circuit. See Sierra Club v. EPA, 540 F. 2d 1114 (D.C. Cir., 1976). Congress then stepped in to codify and expand the program in the 1977 amendments. See Vickie Patton, The New Air Quality Standards, Regional Haze, and Interstate Air Pollution Transport, 28 ELR 10155 (1998); Revesz, Federalism and Interstate Environmental Externalities, supra note 7.

26

William Boyd

prohibiting any stationary source in the state from preventing attainment or maintenance of the NAAQS in downwind states and from interfering with PSD and visibility requirements.44 Congress also expressly prohibited states’ reliance on dispersion techniques to meet the NAAQS, disallowing emissions-reduction credit for intermittent control measures that relied on meteorological conditions and tall stacks instead of good engineering practices.45 It also added a new section 126, which allowed downwind states to petition EPA to force upwind states to modify their SIPs and abate pollution from sources that were “significantly contribut[ing]” to nonattainment with the NAAQS in downwind states.46 These provisions have proved to be important statutory hooks for some of EPA’s recent efforts to deal with cross-state transfers of criteria pollutants and their precursors.47 2.3.1.3 1990 Amendments Congress amended the CAA again in 1990.48 Among other things, the 1990 amendments overhauled section 112 dealing with hazardous air pollutants, created a cap-and-trade program for SO2 and NOx to deal with acid rain, enhanced provisions for interstate cooperation on regional air pollution problems (notably ozone in the Northeast and haze in the western United States), added extensive new mobile source provisions, adopted implementing legislation to meet US commitments under the Montreal Protocol and created the Title V permitting program. All together, the 1990 amendments ran to almost ten times the number of pages as the original 1970 amendments and almost three times that of the 1977 amendments – a reflection of the growing complexity of the problems to be addressed, the use of new policy instruments such as cap and trade and a congressional desire to constrain agency discretion through more prescriptive and precise statutory language.49 The most important provisions related to the NAAQS included revised nonattainment provisions for ozone and carbon monoxide, new initiatives focused specifically

44 45 46 47

48 49

See Section 110(a)(2)(E), 91 Stat. 685, 693. See Section 123; 91 Stat. at 721–22 See Section 126, 91 Stat. 685, 724–25. See Cross-State Air Pollution Rule (CSAPR), supra note 7. As noted, the cross-state rule was upheld by the Supreme Court in EPA v. EME Homer City Generation, L.P., 134 S. Ct. 1584 (2014). P.L. 101-549, 104 Stat. 2399 (1990). To be sure, Congress’s general lack of faith in the executive branch and its corresponding preference for increasingly prescriptive and complex provisions were readily apparent (and remarked on) in the CAA amendments of 1970 and 1977. See, e.g., William D. Ruckelshaus, Environmental Protection: A Brief History of the Environmental Movement in America and the Implications Abroad, 15 ENVT’L L. 455, 460 (1985) (“If you have ever tried to read the Clean Air Act, and I would not wish that on anybody who did not have to administer it, and if you are puzzled by what some of the language in the Act means, try thinking about it as an expression of Congress’s lack of confidence in the executive branch. A lot of provisions that are otherwise baffling become clear.”).

The Clean Air Act’s National Ambient Air Quality Standards

27

on controlling regional air pollution, particularly ozone and haze, and extensive new provisions for mobile sources. Interestingly, although the widely celebrated SO2 trading program (Title IV) was not directly related to the NAAQS program, it was in some ways a response to the problems created by EPA’s earlier decision to allow states to use dispersion techniques (tall stacks) to demonstrate compliance with the NAAQS. It also ended up generating significant NAAQS “co-benefits” by reducing PM2.5 in a number of “downwind states” (largely through the reductions in SO2 as a PM precursor). And it was ultimately rendered obsolete by the NAAQS program as new research on the health and mortality effects of PM2.5 drove more stringent reductions than those required under the cap.50 In looking back at the evolution of the NAAQS statutory framework, several developments stand out, all of which reflect an effort by Congress to reduce some of the flexibility contained in the original 1970 version of the program. First, Congress sought to strengthen and enhance the NAAQS program by providing more prescriptive sourcespecific requirements through the PSD and nonattainment provisions. This cabined some of the flexibility that states enjoyed previously under the SIP process. Second, Congress required more elaborate and formal scientific review, constraining EPA’s discretion and flexibility over the NAAQS process. This bolstered the overall credibility and legitimacy of the program, thereby enhancing its political durability. Third, Congress took a more aggressive approach to regional/interstate problems, in part because of EPA’s inability or unwillingness to use its full authority to mitigate interstate air pollution. Fourth, Congress strengthened the connections between the NAAQS and other CAA programs, designing some of these programs (fuels and mobile sources, for example) with an eye toward assisting with NAAQS attainment. The overall result is a hybrid program that combines ambient air quality standards with specific emissions control requirements built on a foundation of state implementation (see Figure 2.1). 2.3.2 EPA Responsibilities EPA’s major responsibilities under the NAAQS program can be divided into two main categories: (1) establishment, review and revision of the NAAQS and (2) review and approval of SIPs. In addition, EPA has developed extensive regulations and informal guidance under the PSD and nonattainment programs.51 And more recently, it has used its authority under the NAAQS program in new and creative ways, particularly in its efforts to deal with interstate pollution problems.52 Not 50

51 52

Richard Schmalensee & Robert N. Stavins, The SO2 Allowance Trading System: The Ironic History of a Grand Policy Experiment, 27 J. ECON. PERSP. 103 (2013). See Chapter 3. See, e.g., Finding of Significant Contribution and Rulemaking for Certain States in the Ozone Transport Assessment Group Region for Purposes of Reducing Regional Transport of Ozone, 63 Fed. Reg. 57, 356 (Oct. 27, 1998) (the NOx SIP call), upheld in Michigan v. EPA, 213 F.3d 663 (D.C.

William Boyd

28

NAAQS for Criteria Pollutants Established by EPA Administrator

Federal Implementation Plans (FIPs)

State Implementation Plans (SIPS) Air Quality Control Regions (AQCRs)

Nonattainment Program AQCRs designated Nonattainment *RACT for Existing Sources *LAER for New/Modified Sources (New Source Review) *Offsets Requirements

PSD Program AQCRs designated attainment or unclassifiable *BACT for New/Modified Sources (New Source Review) *Increments

Additional designations and requirements for Ozone, PM, CO

figure 2.1 NAAQS framework

surprisingly, much of EPA’s regulatory effort under the NAAQS program has been the subject of litigation, resulting in court-imposed deadlines for many rulemakings. 2.3.2.1 Establishment, Review and Revision of the NAAQS Since the moment it was created, EPA has devoted a considerable amount of time and resources to discharging its basic responsibilities under the NAAQS program. Establishing the NAAQS is its most fundamental responsibility in this respect. But before it can do that, EPA must first create a list of so-called criteria pollutants and develop a criteria document for each listed pollutant. As defined in the statute, criteria pollutants are pollutants which the EPA administrator determines have an adverse effect on public health or welfare and that are present in the ambient air as a result of “numerous or diverse mobile or stationary sources.”53 Criteria documents, which are a holdover from pre-1970 air pollution

53

Cir. 2000); and Cross-State Air Pollution Rule (CSAPR), supra note 7, upheld by the Supreme Court in EPA v. EME Homer City Generation, L.P., 134 S. Ct. 1584 (2014). Section 108(a)(1), 84 Stat. 1678.

The Clean Air Act’s National Ambient Air Quality Standards

29

legislation, are intended to capture the latest scientific information on the air pollutant in question and its impacts on public health and welfare.54 After establishing the list of criteria pollutants and preparing criteria documents, EPA is required by section 109 to establish two types of NAAQS for each criteria pollutant. Primary standards are set to protect public health with an adequate margin of safety, including the health of “sensitive” populations such as asthmatics, children and the elderly.55 Secondary standards are set to protect public welfare, including protection against visibility impairment and damage to animals, crops, vegetation and buildings. For all the criteria pollutants except PM2.5 and SO2, EPA has allowed the primary standard to serve as the secondary standard, raising concerns that these secondary standards may not be adequately protecting welfare.56 EPA’s efforts under the NAAQS program began in early 1971, when the monthsold agency issued the first NAAQS for six criteria pollutants, sulfur dioxide (SO2), particulate matter (PM), carbon monoxide (CO), photochemical oxidants, hydrocarbons (HC) and nitrogen oxides (NOx), based on criteria documents that had been prepared by the Public Health Service pursuant to pre-1970 air pollution legislation. The entire rule establishing the NAAQS for all six of these pollutants covered a mere fifteen pages in the Federal Register.57 In response to litigation, EPA listed lead as a new criteria pollutant in 1976 (the only new criteria pollutant it has ever listed under the program) and prepared its first criteria document as a basis for the first lead NAAQS in 1978.58 Photochemical oxidants were replaced with ozone in 1979.59 The HC standard was revoked in 1983.60 PM was revised in 1987 to include only PM10 and further revised in 1997 to include PM2.5.61 54 55

56

57

58

59

60

61

Section 108(a)(2), 84 Stat. 1678–79. Section 109, 84 Stat. 1679–80. See also Lead Industries Assoc. v. EPA, 647 F.2d 1130, 1152–54 (D.C. Cir. 1980) (confirming that EPA must set the primary NAAQS at levels that will protect the health of sensitive populations). NATIONAL RESEARCH COUNCIL, AIR QUALITY MANAGEMENT REPORT 87 (2004) (noting that the “current practice of letting the primary standard serve as the secondary standard for most criteria pollutants does not appear to be sufficiently protective of sensitive crops and unmanaged ecosystems”). See National Primary and Secondary Ambient Air Quality Standards, 36 Fed. Reg. 8186 (April 30, 1971). The only criteria pollutant for which a separate secondary standard was issued was sulfur dioxide. This standard turned out to be based on an error, which EPA fixed on remand after a challenge from Kennecott Copper. See Kennecott Copper Corp. v. EPA, 462 F.2d 846 (D.C. Cir. 1972). See Natural Resources Defense Council v. Train, 545 F.2d 320, 328 (2d Cir. 1976) (holding that once EPA found that lead met the requirements of section 108(a)(1), it was required to list the pollutant and develop a NAAQS for it). See Revisions to the National Ambient Air Quality Standards for Photochemical Oxidants, 44 Fed. Reg. 8202 (Feb. 8, 1979) (revising photochemical oxidant standard to be expressed as ozone standard). This was the only time that a NAAQS was revised upward, in this case from 0.8 to 1.2 ppm. Id. See National Primary and Secondary Ambient Air Quality Standards, 48 Fed. Reg. 628 (Jan. 5, 1983) (revoking NAAQS for hydrocarbons on grounds that there were no demonstrated direct health effects). See Revisions to National Ambient Air Quality Standards for Particulate Matter, 52 Fed. Reg. 24,634 (Jul 1, 1987) (revising existing particulate matter standard to reflect a new PM10 standard); National Ambient Air Quality Standards for Particulate Matter, 62 Fed. Reg. 38,652 (July 18, 1997) (establishing separate standards for PM10 and PM2.5).

William Boyd

30

Pollutant [links to historical tables of NAAQS reviews]

Primary/ Secondary

Carbon Monoxide (CO)

primary

Lead (Pb)

Averaging Time

Level

Form

8 hours

9 ppm

1 hour

35 ppm

primary and secondary

Rolling 3-month average

0.15 µg/m

primary

1 hour

100 ppb

primary and secondary

1 year

53 ppb

primary and secondary

8 hours

0.070 ppm

primary

1 year

12.0 µg/m3

annual mean, averaged over 3 years

secondary

1 year

15.0 µg/m

3

annual mean, averaged over 3 years

primary and secondary

24 hours

35 µg/m

primary and secondary

24 hours

150 µg/m

primary

1 hour

75 ppb(4)

99th percentile of 1-hour daily maximum concentrations, averaged over 3 years

secondary

3 hours

0.5 ppm

Not to be exceeded more than once per year

Not to be exceeded more than once per year 3(1)

98th percentile of 1-hour daily maximum concentrations, averaged over 3 years

Nitrogen Dioxide (NO2)

Ozone (O3)

PM2.5 Particle Pollution (PM)

PM10

(1)

Sulfur Dioxide (SO2)

Not to be exceeded

Annual Mean

(3)

Annual fourth-highest daily maximum 8-hour concentration, averaged over 3 years

98th percentile, averaged over 3 years

3

3

Not to be exceeded more than once per year on average over 3 years

figure 2.2 Current NAAQS

Starting in the late 1970s, EPA also issued revised NAAQS for the criteria pollutants on eleven separate occasions, most recently in 2015 when it issued a revised NAAQS for ozone (Figure 2.2 shows a table of the current NAAQS). As might be expected, EPA has struggled to meet the five-year NAAQS review requirement on a consistent basis, leading to litigation and court-mandated review schedules. Moreover, as the science of air pollution has grown in volume and sophistication, EPA’s review processes and the resulting criteria documents have become larger and more complex (“encyclopedic” in the words of one observer), raising questions about the relative merits of health-based ambient environmental standards compared with other approaches.62 62

See WENDY WAGNER, SCIENCE IN REGULATION: A STUDY OF AGENCY DECISIONMAKING APPROACHES 30 (2013) (“The scientific analyses in these [NAAQS] reviews have grown from short, relatively simple assessments to encyclopedic assessments that even experts sometimes labeled as impenetrable.”). See

The Clean Air Act’s National Ambient Air Quality Standards

31

In an effort to improve the NAAQS process, EPA has engaged in periodic reviews over the years and has made a number of adjustments.63 Specifically, EPA’s Clean Air Science Advisory Committee conducted formal evaluations of the NAAQS process in 1981 and 1985 – both of which resulted in changes to the process.64 The National Academy of Sciences has also produced several important reports on various aspects of the NAAQS and the CAA more generally, often in response to specific requests from Congress or EPA, which have informed efforts to revise the NAAQS process.65 EPA initiated a major review of the NAAQS process in 2006, leading to substantial revisions that separated analytically distinct components and created a more formal framework for NAAQS review and revision.66 In 2009, EPA Administrator Lisa Jackson issued a memorandum on the NAAQS review process that further elaborated the basic components, including an initial planning document and literature review, an integrated science assessment (previously known as the criteria document), a risk and exposure assessment, a policy assessment and the formal rulemaking67 (see Figure 2.3). In each of these steps, EPA has been “procedurally generous,” allowing for various forms of public participation that go well beyond the requirements of the Administrative Procedures Act.68 Several commentators have argued that these new procedures adopted by EPA in the mid-2000s have resulted not only in greater transparency and accountability but also in a more productive “partnership” with the courts in their effort to review agency science underlying the NAAQS.69

63

64

65

66

67

68

69

also Michael L. Livermore & Richard L. Revesz, Rethinking Health-Based Environmental Standards, 89 N.Y.U L. REV. 1184, 1258–64 (2014) (questioning the health-based approach of the NAAQS). See also Morton Lippmann, Role of Science Advisory Groups in Establishing Standards for Ambient Air Pollutants, 6 Aerosol Sci. Tech. 93, 108–13 (1987) (summarizing 1981 and 1985 CASAC assessments of and recommendations regarding the NAAQS review process). See U.S. EPA, CLEAN AIR SCIENCE ADVISORY COMMITTEE, SETTING NATIONAL AMBIENT AIR QUALITY STANDARDS: IMPROVING THE PROCESS (1981); U.S. EPA, SCIENCE ADVISORY BOARD, REPORT OF THE CLEAN AIR SCIENCE ADVISORY COMMITTEE (CASAC) ON IMPROVING THE PROCESS FOR SETTING NATIONAL AMBIENT AIR QUALITY STANDARDS: AN UPDATE (1985). See, e.g., NATIONAL RESEARCH COUNCIL, AIR QUALITY MANAGEMENT IN THE UNITED STATES (2004); NATIONAL RESEARCH COUNCIL, RETHINKING THE OZONE PROBLEM IN URBAN AND REGIONAL AIR POLLUTION (1991). See U.S. EPA, NAAQS PROCESS WORKGROUP, REVIEW OF THE PROCESS FOR SETTING THE NATIONAL AMBIENT AIR QUALITY STANDARDS (2006). See also Sidney Shapiro et al., The Enlightenment of Administrative Law: Looking Inside the Agency for Legitimacy, 47 WAKE FOREST L. REV. 463, 493–96 (2012) (describing the separate steps of the new NAAQS review process and characterizing them as a “deliberative-constitutive” process that enhances legitimacy and accountability). Lisa Jackson, Memorandum on Process for Reviewing National Ambient Air Quality Standards, May 21, 2009. See STEVEN P. CROLEY, REGULATION AND THE PUBLIC INTERESTS: THE POSSIBILITY OF GOOD REGULATORY GOVERNMENT 258 (2008) (observing that EPA was “procedurally generous” in the 1997 ozone/PM2.5 rulemaking, soliciting comments and input from a wide variety of stakeholders and experts that went well beyond the minimum requirements of the APA). See Fisher et al., Rethinking Judicial Review of Expert Agencies, supra note 3; Shapiro et al., The Enlightenment of Administrative Law, supra note 66. See also CROLEY, REGULATION AND THE PUBLIC INTERESTS, supra note 68 at chap. 9.

William Boyd

32

Peer-reviewed scientific studies

Workshop on science-policy issues

Integrated Science Assessment: concise evaluation and synthesis of most policy-relevant studies

Integrated Review Plan: timeline and key policyrelevant issues and scientific questions

CASAC review and public comment

Risk/Exposure Assessment: concise quantitative assessment focused on key results, observations and uncertainties

CASAC review and public comment

EPA proposed decision on standards

Policy Assessment: staff analysis of policy options based on integration and interpretation of information in the ISA and REA

Agency decision making and draft proposal notice

Interagency review

Public hearings and comments on proposal

Agency decision making and draft final notice

Interagency review

EPA final decision on standards

figure 2.3 NAAQS review process

In particular, they point to the development of internal agency “yardsticks” and a coherent “epistemic framework” for managing the complex process of NAAQS review and revision and weighing scientific evidence as key elements of EPA’s new process that have facilitated a more hospitable climate for public participation and judicial review.70 As discussed in more detail later, the development of robust internal procedures at EPA has been an important contributor to the overall legitimacy of the program. Administrative process, in this respect, has been a key component of the success and durability of the NAAQS. 2.3.2.2 Review, Approval and Enforcement of SIPs EPA’s other major responsibility under the NAAQS program is to review, approve (or disapprove) and enforce the SIPs. Under the statute, once a state submits a SIP or SIP revision to EPA, the agency can approve or disapprove the SIP in whole or in part.71 EPA must approve a SIP if it finds that it meets all applicable requirements, and the courts have expressly held that EPA cannot impose particular emissions controls or requirements other than those called for in the statute on the states 70

71

Fisher et al., Rethinking Judicial Review of Expert Agencies, supra note 3. See also Wagner, SCIENCE IN REGULATION, supra note 62 (“ Although it has not been easy, EPA appears to have finally developed a transparent process that produces analyses that are accessible to expert onlookers and that manages successfully to bridge science and policy in ways that appear worthy of replication.”). Before reviewing, however, EPA must determine whether the submission is complete. See section 7410(k)(1). EPA’s criteria for whether a submission is complete are set out at 40 C.F.R. pt. 51, App. V; 40 C.F.R. pt. 51, App. V.

The Clean Air Act’s National Ambient Air Quality Standards

33

through its SIP authority.72 If EPA finds that a SIP does not meet the requirements of the CAA, it may disapprove the SIP. EPA can also partially or conditionally approve a SIP or SIP revision.73 Conditional approvals require states to agree to adopt enforceable measures by a definite date no more than one year from the conditional approval.74 If the state later fails to adopt those measures, the SIP is treated as disapproved.75 Finally, EPA can also ask states to revise SIPs that have been previously approved if it later finds that these SIPs are inadequate. This process, known as a SIP call, has been used by EPA in a variety of contexts, most notably in its efforts since the late 1990s to force “upwind” states to consider and deal with their contributions to attainment problems in “downwind” states.76 If a state fails to submit a satisfactory SIP, EPA must develop its own compliance plan within two years. This plan is known as a federal implementation plan (FIP).77 In the past, EPA has generally used FIPs in situations where states have failed to correct some previously identified deficiency. Recently, however, EPA has taken a more proactive approach in using its FIP authority. In its 2011 Cross-State Air Pollution Rule (also known as the Transport Rule), for example, EPA simultaneously announced a regulatory emissions budget that quantified states’ obligations for curbing interstate NOx and SO2 air pollution and promulgated FIPs to implement the budget and associated controls in noncomplying states.78 Significantly, EPA did not first give states an opportunity to meet the newly quantified obligations in their SIPs, based on the theory that the states’ existing SIPs were inadequate to satisfy the good neighbor provision and that immediate federal intervention was therefore authorized. This approach was 72

73 74 75 76

77 78

40 C.F.R. pt. 51. See Train v. NRDC, 421 U.S. 57, 79 (1975) (“Under § 110(a)(2), the Agency is required to approve a state plan which provides for the timely attainment and subsequent maintenance of ambient air standards and which also satisfies the section’s other general requirements. The Act gives the Agency no authority to question the wisdom of a State’s choices of emission limitations if they are part of a plan which satisfies the standards of §110(a)(2), and the agency may devise and promulgate a plan of its own only if a State fails to submit an implementation plan which satisfies those standards” (emphasis in original); Virginia v. EPA, 108 F.3d 1397, 1407–9 (D.C. Cir. 1997) (holding that EPA had exceeded its authority under section 110 by requiring certain states to adopt California’s vehicle emission program as the principal means of reducing ozone precursors). 42 U.S.C. § 7410(k)(3). A partial approval is not complete until EPA approves the entire plan. Id. § 7410(k)(4). Id. Id. § 7410(c)(1). EPA used its SIP call authority in 1998 when it issued the NOx SIP call to twenty-two states and the District of Columbia, mandating that they revise their SIPs to mitigate interstate transport of ozone and its precursors (NOx). See Finding of Significant Contribution and Rulemaking for Certain States in the Ozone Transport Assessment Group Region for Purposes of Reducing Regional Transport of Ozone, 63 Fed. Reg. 57,356 (1998). The major provisions of this rule were upheld by the DC Circuit in Michigan v. EPA, 213 F.3d 663 (D.C. Cir. 2000). 42 U.S.C. § 7410(c)(1). 76 Fed. Reg. 48,208 (Aug. 8, 2011). As noted earlier, the good neighbor provision requires SIPs to control emissions contributing significantly to nonattainment or that interfere with NAAQS maintenance in another state. See 42 U.S.C. § 7410(a)(2)(D)(i)(I). A closely related provision requires states to consider the impacts of their air pollution on foreign countries. See id. §§ 7410(a)(2)(D)(ii), 4215.

34

William Boyd

recently upheld by the Supreme Court in EPA v. EME Homer City Generation, L.P.79 Four brief observations are worth making regarding EPA’s major responsibilities under the NAAQS program. First, the NAAQS review process has become much more complex and science intensive, requiring mobilization of significant internal and external resources. Second, EPA’s efforts to establish more formal and transparent procedures for the NAAQS review process, including its reliance on independent scientific review, has been a critical component of the overall durability of the program. Third, the SIP review and approval process has not received the same level of attention as the NAAQS review process, leading to criticisms that it is overly bureaucratic and ineffective. Fourth, EPA has made more creative use of its SIP/FIP authorities and the CAA’s good neighbor provisions in recent years to deal with persistent problems of interstate transport of criteria pollutants and their precursors. 2.3.3 Independent Scientific Review Independent scientific review has played a fundamental role in the NAAQS program, bolstering EPA’s own scientific efforts and helping to maintain the credibility of the program. This has contributed significantly to the political durability of the NAAQS. As noted earlier, Congress was an important early driver in this respect, instructing EPA in the 1977 amendments to create an independent Clean Air Science Advisory Committee (CASAC) that would play a formal role in the NAAQS review and revision process.80 As amended, section 109 required EPA to appoint an independent seven-member CASAC that includes at least one member of the National Academy of Sciences and one representative from state air pollution control agencies. The CASAC reviews the air quality criteria and NAAQS at fiveyear intervals and makes recommendations to the EPA administrator regarding new criteria pollutants and revisions of existing air quality criteria and NAAQS as appropriate.81 Section 307 also requires that EPA explain in its rulemakings the reasons for any differences between the proposed or final NAAQS and CASAC recommendations.82 79 80

81

82

EME Homer City, 134 S.Ct. at 1595–96 (2014). The CASAC has been subject to the Federal Advisory Committees Act (FACA) since it was established. For an illuminating discussion of CASAC’s role and relationship with EPA management during its early years, see SHELIA JASANOFF, THE FIFTH BRANCH: SCIENCE ADVISERS AS POLICYMAKERS 101–22 (1990). The Environmental Research, Development, and Demonstration Authorization Act of 1978 established EPA’s Science Advisory Board (SAB) and also calls for review of the NAAQS. EPA has historically relied on the CASAC review process to satisfy these ERDDAA provisions. In conducting its reviews, the CASAC is supplemented with additional subject-matter experts forming a CASAC Review Panel. In its 2008 revision of the ozone NAAQS, for example, EPA chose a standard (0.075 ppm) above the range recommended by the CASAC (0.060–0.070 ppm). See National Ambient Air Quality Standards

The Clean Air Act’s National Ambient Air Quality Standards

35

Since its creation, the CASAC has played an active role in the NAAQS review process. In addition to reviewing and providing feedback on the various formal documents prepared by EPA as part of the process, the CASAC also convenes various meetings and workshops with the broader scientific community.83 At the end of the process, the CASAC provides a formal assessment (known as a closure letter) to the EPA administrator containing its views on the state of the science and suggesting a range for a revised standard if it concludes that the existing standard needs to be changed. The CASAC also provides advice to EPA on a variety of other issues, including air quality modeling and monitoring, research needs and reform of the NAAQS review process itself.84 In addition to these formal efforts to ensure that independent scientific review is well integrated into the NAAQS process, there have been multiple outside scientific reviews by the National Research Council (NRC) and other bodies looking at the NAAQS program as a whole as well as at specific challenges associated with particular pollutants.85 In some cases, these various reviews and reports were called for by Congress, demonstrating again a congressional commitment to ensuring ongoing independent assessment of the NAAQS program and its challenges. The 2004 NRC report, Air Quality Management in the United States, for example, was prepared in response to a congressional request for an independent evaluation of the effectiveness of the NAAQS and the CAA more generally.86 Likewise, the 1992 NRC report, Rethinking the Ozone Problem in Urban and Regional Air Pollution, was developed in part as a response to a congressional request for a National Academy of Sciences study on ozone precursors and their role in ozone formation and control.87 In these cases and others, the resulting reports have provided important additional input to EPA as it has worked to strengthen the scientific foundations of the NAAQS and to improve the overall process.

83

84

85

86

87

for Ozone, 73 Fed. Reg. 16,436, 16,482 (March 27, 2008). EPA’s decision was upheld by the DC Circuit in Mississippi v. EPA, 723 F.3d 246–9 (D.C. Cir. 2013). The key documents here are now known as the Integrated Science Assessment, previously known as the Criteria Document, a Risk/Exposure Assessment and a Policy Assessment, previously known as the Staff Paper. The CASAC typically reviews and comments on multiple drafts of these documents before they are finalized. On CASAC reports on the NAAQS review process, see CASAC, SETTING NATIONAL AMBIENT AIR QUALITY STANDARDS: IMPROVING THE PROCESS (1981); CASAC, IMPROVING THE PROCESS FOR SETTING NATIONAL AMBIENT AIR QUALITY STANDARDS: AN UPDATE (1985). See, e.g., NATIONAL RESEARCH COUNCIL, GLOBAL SOURCES OF LOCAL POLLUTION: AN ASSESSMENT OF LONG-RANGE TRANSPORT OF KEY AIR POLLUTANTS TO AND FROM THE UNITED STATES (2009); NATIONAL RESEARCH COUNCIL, AIR QUALITY MANAGEMENT IN THE UNITED STATES (2004); NATIONAL RESEARCH COUNCIL, RETHINKING THE OZONE PROBLEM IN URBAN AND REGIONAL AIR POLLUTION (1992). See NATIONAL RESEARCH COUNCIL, AIR QUALITY MANAGEMENT IN THE UNITED STATES (2004). The report contains a number of important recommendations for improving the NAAQS process and adapting the program to deal with difficult challenges such as regional transport issues. See NATIONAL RESEARCH COUNCIL, RETHINKING THE OZONE PROBLEM IN URBAN AND REGIONAL AIR POLLUTION (1992).

36

William Boyd

The NAAQS program thus draws on a broad and robust network of independent scientific review. The integrated science assessments are based on a thorough review (internal and external) of literally thousands of scientific studies.88 These documents are designed to reflect the current state of the science and now often stretch well beyond 1,000 pages (not including appendices). During each individual NAAQS review, moreover, EPA’s science assessment, its risk and exposure assessments and its policy assessment are all subject to multiple rounds of review and input from the CASAC as well as more informal review and input from the broader scientific community and from the public at large. At a more general level, the NRC and others provide yet another layer of independent review of various aspects of the NAAQS program (and other aspects of the CAA), as well as particular challenges confronting the program (e.g., ozone chemistry, long-range transport of pollutants, etc.). To a considerable extent, the evolution of the NAAQS program and the increasingly formal role that independent scientific review has come to play in the NAAQS review process reflect a broader trend in regulatory science toward increased reliance on external peer review as a means of insulating agency decision making from claims of bias and excess discretion. This “renegotiation of expertise” has often translated into a focus on process rather than the substance of specific decisions as a means to bolster the credibility of EPA’s decision making.89 Without question, difficult and inescapable policy judgments remain at the heart of NAAQS standard setting, and a preoccupation with process can create its own problems (including a loss of flexibility). On the whole, though, it is clear that the efforts by Congress and EPA to create a robust process of independent scientific review as a fundamental part of the NAAQS process has contributed significantly to the overall success of the NAAQS in withstanding challenges (legal and political) and in maintaining public trust and credibility. 2.3.4 Public Participation Like other environmental statutes, the CAA contains multiple avenues for public participation. Most prominently, the “citizen suit” provisions under section 304 and the petition process under section 307 have allowed various citizens groups, advocacy organizations, state and local governments and industry groups (among others) to challenge EPA action (or inaction) on many fronts and to force the agency to 88

89

WAGNER, SCIENCE IN REGULATION, supra note 62 at 30 (“A single NAAQS review can involve the analysis of thousands of studies.”). See, e.g., Shelia Jasanoff, Science, Politics, and the Renegotiation of Expertise at EPA, 7 OSIRIS 194, 197 (1992) (“[I]n the arena of environmental decision making public representation of science has shifted away from an emphasis on testable knowledge claims to a preoccupation with the processes of knowledge production. Under continual assault from political adversaries, EPA’s environmental science has more and more justified itself in terms of its legal, institutional, and procedural underpinnings rather than the truth-value of the facts it alleges.”).

The Clean Air Act’s National Ambient Air Quality Standards

37

discharge its responsibilities under the CAA. Section 304 allows “any person” to commence a civil action against any other person (including the United States) alleged to be in violation of an emission standard or limitation under the Act or an order issued by EPA or a state with respect to such a standard or limitation.90 It also allows such persons to commence an action against the EPA administrator for failure to discharge any nondiscretionary duty.91 Section 307 contains procedural and venue provisions regarding petitions for review of specific actions taken by EPA, including various procedural steps that EPA takes in discharging its duties.92 Taken together, these provisions have allowed for robust participation (manifest in a large amount of litigation) by citizens groups, industry representatives and others, resulting in major judicial decisions, court-imposed deadlines and a more general check on regulated entities and EPA that have profoundly shaped the NAAQS program. Indeed, since 1970, scores of lawsuits have been filed on various aspects of the NAAQS program, resulting in dozens of appellate decisions. Among other things, citizen suits and section 307 petitions have resulted in new listings of criteria pollutants (lead), court-imposed deadlines for the NAAQS review process, challenges to the NAAQS as promulgated, challenges to PSD and NSR regulations, challenges to individual SIPs and the SIP process, challenges to attainment and nonattainment designations and challenges to EPA’s efforts to use various authorities to deal with interstate transport issues. In sum, hardly any aspect of the NAAQS program has escaped judicial review, and it is fair to say that the NAAQS program would not look anything like it does today in the absence of these provisions for public participation. These important procedural features of the statute have contributed to both durability and flexibility. Robust citizen participation and judicial review have worked to reinforce the overall quality (and the legitimacy) of EPA’s efforts under the NAAQS program, although there are surely cases where EPA has focused too much on surviving judicial review rather than on substantive outcomes. The constant threat of litigation from interest groups on all sides has likely protected the program from various forms of capture and has arguably allowed it to respond to evolving social priorities. Court-imposed deadlines have also given EPA muchneeded political cover, allowing it to move forward with expensive and controversial rulemakings. In effect, these provisions have kept the pressure on EPA, providing an 90 91

92

Section 304(a)(1), 84 Stat. 1676, 1706. Section 304(a)(2), 84 Stat. 1676, 1706. This provision also waives the traditional requirements of amount in controversy. A third set of activities subject to the citizen suits provision was added in the 1977 amendments pertaining to persons who construct any new major emitting facility without a permit under either the PSD or nonattainment program. See new Section 304(a)(1)(3), 91 Stat. 685, 771. Section 307(b), 84 Stat. 1676, 1708. This provision would be elaborated further in the 1977 amendments with a long list of specific EPA actions to which it applied. See new Section 307(d), 91 Stat. 685, 772–76.

38

William Boyd

additional check on agency behavior and allowing the program to move forward even in the face of politically challenging circumstances. But, of course, lawsuits are cumbersome and time-consuming and do not always result in progress toward the overall objectives of the NAAQS program. Other, less formal (and less litigious) opportunities for public participation have also enhanced the durability of the program. In particular, over the last decade or so, EPA has created multiple opportunities for public review and input as part of the NAAQS review process, often in the form of workshops and solicitations of feedback on draft assessments. None of these are required as part of any rulemaking; they are in addition to formal notice and comment provisions. But they serve the important function of allowing EPA to vet revised standards before any formal proposal is made. This, in turn, allows the Agency to receive valuable feedback on possible revisions and, at the same time, to signal to the regulated community and other stakeholders the Agency’s thinking about future NAAQS revisions.93 By the time EPA releases its formal proposal for a revised NAAQS, no one can really claim surprise. Given prior vetting of the proposed revision, moreover, the rulemaking is almost certainly stronger and less vulnerable to attack (from whatever side) than it would have been. As a result, these informal means of public participation contribute further to the overall political legitimacy and durability of the program. 2.3.5 Monitoring and Modeling Infrastructure In contrast to some traditional end-of-pipe pollution control standards, ambient environmental quality standards require an elaborate technical and analytical infrastructure. As noted earlier, EPA’s efforts to establish the NAAQS (and to review and revise the NAAQS over time) are extremely science-intensive, requiring review of thousands of studies from multiple disciplines.94 The entire enterprise, moreover, depends on a range of tools for monitoring and modeling air quality, assessing risk and exposure and developing scenarios to evaluate future controls. The nation’s air quality monitoring network provides the empirical foundation for the NAAQS.95 As such, it has evolved considerably over the years as EPA has worked to make it more responsive to the needs of the program. Similarly, air quality 93

94 95

See CROLEY, REGULATION AND THE PUBLIC INTERESTS. supra note 68; Fisher et al., Rethinking Judicial Review of Expert Agencies, supra note 3; WAGNER, SCIENCE IN REGULATION, supra note 62. WAGNER, SCIENCE IN REGULATION, supra note 62. The 1977 Amendments to the Clean Air Act directed EPA to establish an air quality monitoring system in the United States. See section 309, 91 Stat. 781–82 (directing the EPA administrator to promulgate regulations establishing an air quality monitoring network throughout the United States that would, among other things, use uniform criteria and methodology, provide for air quality monitoring in major urban areas in a manner that supplemented state monitoring programs, provide for daily analysis and reporting of air quality and compile air monitoring data to inform the administrator’s actions). See also NATIONAL RESEARCH COUNCIL, AIR QUALITY MANAGEMENT IN THE UNITED STATES 219 (2004) (“Since the 1980s, the United States has had an extensive air quality monitoring network that routinely measures the concentrations of selected air pollutants in some locations.”); id (noting that the network

The Clean Air Act’s National Ambient Air Quality Standards

39

modeling has been central to the entire NAAQS enterprise since its inception, and the courts have generally been deferential to EPA in reviewing how it uses models.96 Modeling has been particularly important regarding EPA’s efforts on interstate regional pollution issues and not always to good effect. In the late 1970s and early 1980s, for example, EPA repeatedly used the limits of regional air quality models to avoid taking strong action on interstate air pollution – a course of action that the courts largely upheld.97 The SIP process has also long relied on emissions inventories and air quality models to determine control strategies and demonstrate compliance. Emissions inventories, for example, have been critical in developing estimates of current air pollution and in identifying sources. Air quality models, in turn, have provided the basis for developing scenarios for future controls in order to demonstrate attainment with the various NAAQS. All these tools have limits and shortcomings, some of which can have important consequences for the effectiveness of the NAAQS program.98 Efforts to control ozone offer a cautionary tale in this respect. Until the early 1990s, efforts to comply with the ozone NAAQS focused primarily on controlling volatile organic compounds (VOCs) rather than NOx, both of which are ozone precursors. This control strategy, however, was based on erroneous emissions inventories that had been systematically underestimating VOC emissions. At relatively low VOC to NOx ratios, controlling VOCs is the best way to reduce ozone. At higher ratios, however, the preferred mitigation strategy shifts to NOx control. When field data from the 1980s and 1990s revealed much higher emissions of VOCs (from mobile and

96

97 98

was “largely designed to monitor urban pollution levels and compliance with the National Ambient Air Quality Standards (NAAQS)”). The CAA requires states to establish air monitoring stations for the criteria pollutants. There are around 4,000 of these state and local air monitoring stations (SLAMS) distributed across the country based largely on the needs of state and local air pollution agencies to meet their SIP obligations. A subset of the SLAMS network (about 1,800 stations) is designated as national air monitoring stations (NAMS), which are typically located in urban and multisource areas. The 1990 amendments also required EPA, in partnership with state and local agencies, to carry out more extensive monitoring of ozone and its precursors in areas of persistent nonattainment. EPA responded by establishing a network of photochemical assessment monitoring stations (PAMS) in twenty-four urban areas. See id at 221–26 (discussing SLAMS, NAMS and PAMS). See Thomas O. McGarity and Wendy E. Wagner, Legal Aspects of the Regulatory Use of Environmental Modeling, 33 ELR 10751 (2003) (reviewing thirty years of judicial challenges to EPA’s use of models to support various regulations); NATIONAL RESEARCH COUNCIL, MODELS IN REGULATORY DECISION MAKING 76–79 (2007) (discussing legal challenges to EPA’s use of models to support environmental regulations). See Patton, supra note 43, at 10168 (discussing cases). See NATIONAL RESEARCH COUNCIL, AIR QUALITY MANAGEMENT IN THE UNITED STATES 113 (2004) (“Literature estimates for individual components of an air quality model – emissions, chemistry, transport, vertical exchange, deposition – typically indicate uncertainties of 15–30%, but when the supporting data sets are weak, the uncertainties can be significantly higher . . . Relying solely on the output of an air quality model to resolve emission-control issues or to demonstrate attainment of an air quality standard or objective is problematic.”); id. at 99 (“The consequences of errors in emission inventories can be profound.”).

40

William Boyd

biogenic sources) than was apparent from previous emissions inventories, it became clear that ozone mitigation needed to focus more on NOx controls than on VOC controls.99 It would take the better part of a decade for EPA to shift its overall approach, resulting in the NOx SIP call in 1998, the Clean Air Interstate Rule (CAIR) and, finally, the Cross-State Air Pollution Rule (CSAPR), which was recently upheld by the Supreme Court in the EME Homer Generation case.100 This episode provides another reason why flexibility is so important to the NAAQS program. As tools for monitoring and modeling air quality improve, new problems and new aspects of old problems come into view. These new ways of seeing condition not only how such problems are understood but also how EPA and others think about the possibilities for response.101 At the same time, when our ways of seeing particular problems are uneven, incomplete or faulty, large misallocations of resources can sometimes result.102 Ongoing review, assessment and, where necessary, revision of the underlying monitoring and modeling infrastructure that supports the NAAQS are thus a critical part of the success of the program. Here again, the flexibility to make adjustments as tools improve and as the underlying science evolves is critical to the long-term success and durability of the program.

2.4 goals, deadlines and long-term signaling The goal of the NAAQS program is deceptively simple: to reduce concentrations of criteria pollutants in the ambient air to levels that will protect public health and welfare. EPA sets the standards and regulates mobile sources, while the states are charged with implementing controls on stationary sources necessary to meet the federal standards. The Supreme Court held in 2001, as the DC Circuit previously held in 1978, that EPA may not consider costs in setting the NAAQS.103 The statute 99

100

101

102

103

See NATIONAL RESEARCH COUNCIL, AIR QUALITY MANAGEMENT IN THE UNITED STATES 107–8 (2004); NATIONAL RESEARCH COUNCIL, RETHINKING THE OZONE PROBLEM IN URBAN AND REGIONAL AIR POLLUTION (1991). See EPA v. EME Homer City Generation, L.P., 134 S. Ct. 1584 (2014) (discussing the various EPA rulemakings on interstate air pollution and upholding the Cross-State Air Pollution Rule). See William Boyd, Ways of Seeing in Environmental Law: How Deforestation Became an Object of Climate Governance, 37 ECOLOGY L.Q. 843, 898–915 (2010); William Boyd Genealogies of Risk: Searching for Safety, 1930s–1970s, 39 ECOLOGY L.Q. 895, 944–47 (2012). These issues can also emerge as important sources of conflict. Although often hidden from public view, the choices made about how and what to monitor, as well as the assumptions and protocols used in air quality modeling, can elicit intense scrutiny and opposition from the regulated community and other stakeholders. See, e.g., Art Fraas, John D. Graham & Jeff Holmstead, EPA’s New Source Review Program: Time for Reform?, 47 ENVT’L L. REP. 10026 (2017) (criticizing EPA’s current modeling assumptions and guidance as a component of NSR permitting and arguing for adoption of a probabilistic modeling approach). See Whitman v. American Trucking, 531 U.S. 457, 465 (2001) (“Section 109(b)(1) instructs the EPA to set primary ambient air quality standards ‘the attainment and maintenance of which . . . are requisite to protect the public health’ with ‘an adequate margin of safety.’ Were it not for the hundreds of pages of briefing respondents have submitted on the issue, one would have thought it fairly clear that this

The Clean Air Act’s National Ambient Air Quality Standards

41

thus requires that the EPA administrator make a policy judgment based on the scientific case contained in the criteria document and supporting analyses in determining the level at which the NAAQS for a particular criteria pollutant will protect public health “with an adequate margin of safety.”104 This is an ambitious, precautionary standard. Critics have argued that it puts EPA in a difficult (if not impossible) situation with respect to nonthreshold pollutants, requiring the Agency to set the NAAQS at zero if, in fact, it is going to protect public health with an adequate margin of safety. But there is no evidence in the legislative history that Congress intended for the “margin of safety” language to mean “zero risk,” and EPA and the federal courts have consistently interpreted safety as “acceptable risk” in order to allow for more flexibility in determining the level at which to set the NAAQS.105 The NAAQS program thus provides a clear signal to the states and affected sources that the standards will be set at levels necessary to protect public health and welfare without consideration of technological or economic feasibility. To be sure, considerations of cost and feasibility do come into play in the implementation of the NAAQS (see below), and they are clearly visible in the Regulatory Impact Analyses (RIA) and Office of Information and Regulatory Affairs (OIRA) reviews of NAAQS rulemakings. At times, these cost considerations (and the associated politics) have intruded into the NAAQS process (e.g., 2011 OIRA letter to Lisa Jackson directing EPA to pull the revised ozone NAAQS106). But as a formal legal matter, EPA is barred from considering costs in setting the NAAQS. The ambitious health-protective goals of the NAAQS have been tempered in some respects by adjusting attainment deadlines and by various provisions that allow for flexible implementation. As noted earlier, the initial 1970 version of the NAAQS program proved to be quite unrealistic with respect to the time it would take for states to comply with the NAAQS, leading to frustration on the part of some state agencies

104 105

106

text does not permit EPA to consider costs in setting the standards. The language, as one scholar has noted, ‘is absolute’”, (citations omitted.)) See also Lead Industries v. EPA, supra note 55. But see Livermore & Revesz, Rethinking Health-Based Environmental Standards, supra note 62 (arguing for a reinterpretation of American Trucking to allow for cost–benefit analysis to serve as a regulatory floor for setting the NAAQS, especially for nonthreshold criteria pollutants). For a brief response to Livermore and Revesz, see Gary Guzy, Rethinking Rethinking Health-Based Environmental Standards and Cost-Benefit Analysis: A Solution in Search of a Problem?, 46 ENVT’L L. REP. NEWS & ANAL. 10,681 (2016). Section 109. See Bachmann, Will the Circle Be Unbroken, supra note 2 at 667 (noting that there is no evidence that “margin of safety” meant zero risk for nonthreshold pollutants). But see NATIONAL RESEARCH COUNCIL, AIR QUALITY MANAGEMENT IN THE UNTED STATES 77–78, 87 (2004) (noting challenges to setting NAAQS based on “adequate margin of safety” language for nonthreshold pollutants). For a broader discussion of how safety was redefined as acceptable risk across multiple domains of US health, safety and environmental law in the 1970s and early 1980s, see William Boyd, Genealogies of Risk: supra note 101, 964–83. See Letter from Cass R. Sunstein, Administrator of OIRA, to Lisa Jackson, Administrator of EPA, dated September 2, 2011.

42

William Boyd

and gaming of the SIP process.107 In response, Congress adjusted the attainment deadlines in the 1977 amendments. Moreover, in order to deal with the persistent and widespread nonattainment with the NAAQS for ozone, Congress established in the 1990 amendments a sequence of attainment dates based on an area’s particular nonattainment classification, thereby providing a more flexible timetable for state and local agencies to deal with the problem.108 Notwithstanding the relaxation of attainment deadlines, the ongoing five-year review of the NAAQS combined with the requirement that EPA revise the standards as necessary have together established a set of general expectations about the overall direction of the program. In this context, EPA’s NAAQS review process allows the Agency to start signaling to the states and affected sources well in advance of any final rule establishing a new standard. States and affected sources then have additional time to comply with the NAAQS after the new standard is finalized. This kind of long-term signaling combined with flexibility on timing has likely been an important source of the program’s overall political durability, but it also begs the question about the overall efficacy of the program. How long will it take, in other words, to bring persistent areas of ozone nonattainment into attainment, and what does this say about the success of the program? SIPs can also perform an important signaling function with respect to affected sources. By laying out an implementation framework and schedule that includes specific controls, the SIP process provides a forum for affected sources to develop their own compliance plans and to negotiate with state regulators. The courts have also held that SIPs can include emissions control requirements that are not technologically feasible under current conditions – that they can be “technology-forcing,” which also sends a powerful signal to affected sources.109

2.5 cooperative federalism: the sip process The NAAQS SIP process is one of the oldest and best examples of cooperative federalism in US environmental law. States are the primary implementers of the NAAQS and, historically, have enjoyed considerable flexibility in allocating emissions reductions across various sectors and sources.110 In effect, the SIPs provide the 107 108 109

110

NATIONAL RESEARCH COUNCIL, AIR QUALITY MANAGEMENT IN THE UNITED STATES 129 (2004). Id. at 132. See, e.g., Union Electric Co. v. EPA, 427 U.S. 246, 265 (1976) (concluding that claims of economic and technological infeasibility cannot be considered by EPA in deciding whether to approve a SIP); see also id., at 269 (“Technology forcing is a concept somewhat new to our national experience and it necessarily entails certain risks. But Congress considered those risks in passing the 1970 Amendments and decided that the dangers posed by uncontrolled air pollution made them worth taking.”). As is the case in various other environmental statutes, the states can impose more stringent standards and controls than those required under the NAAQS (Train v. NRDC, 421 U.S. 60 (1975); Union Electric, supra note 108).

The Clean Air Act’s National Ambient Air Quality Standards

43

basic implementation framework for the NAAQS, linking the substantive standards established by EPA with state regulations and federal oversight and enforcement. Under section 110 of the CAA, states are required to submit SIPs within three years of the establishment of a new NAAQS for one of the criteria pollutants. Tribes submit tribal implementation plans (TIPs). Each SIP (or TIP) must demonstrate (largely on the basis of modeling) how the various Air Quality Control Regions within the state (or reservation) will attain the NAAQS. The statute sets forth the general requirements for SIPs, including (1) an emissions inventory, (2) attainment demonstrations based on air quality models and other analyses, and (3) a description of emission control strategies and enforcement measures that will allow achievement of the required reductions.111 SIPs are also required to include provisions that will prohibit emissions that will interfere with other “downwind” states’ attainment or maintenance of the NAAQS.112 While states are charged with developing SIPs, the CAA establishes planning procedures regarding who should be involved in the process.113 SIP planning must include state, regional and local government officials.114 Preparation of the SIP is done by an organization certified by the state and must include state air pollution regulators, state transportation planning officials, metropolitan transportation planning officials and local elected officials.115 Once approved by EPA, SIPs are federally enforceable. As noted earlier, if EPA finds that a SIP is inadequate or that a state is delinquent in implementing its SIP, it can develop a FIP. The agency can also impose sanctions (loss of highway funds) on states that fail to carry out their responsibilities under the SIP process. The SIP process was designed to be one of the most important sources of flexibility under the NAAQS program. Individual states and tribes would be allowed to determine in large part which sources they would control (and how) in order to attain and maintain the NAAQS within their respective Air Quality Control Regions. This flexibility, which derived from the CAA’s commitment to cooperative federalism, was intended to contribute to the overall political durability of the program. Efforts by EPA to force the states to adopt specific measures in their SIPs have repeatedly been rejected by the courts.116 Starting with the 1977 amendments, however, Congress constrained some of the flexibility that states had historically enjoyed in the SIP process. The PSD and nonattainment programs, for example, required certain technology controls for 111 112

113 114 115 116

42 U.S.C § 7410(a)(2). This is known as the good neighbor provision. See 42 U.S.C. § 7410(a)(2)(D)(i). See also EPA v. EME Homer City Generation, L.P., 134 S. Ct. 1584, 1587 (2014). 42 U.S.C. § 7504(a). Id. Id. See Train v. NRDC, 421 U.S. 60 (1975); Virginia v. EPA, 108 F.3d 1397 (D.C. Cir. 1997). See also Michigan v. EPA, 213 F.3d 663, 686–87 (D.C. Cir. 2000) (discussing Train-Virginia federalism bar to EPA efforts to condition SIP approval on adoption of particular control measures).

44

William Boyd

specific sources depending on the attainment status of the various Air Quality Control Regions in the state. Likewise, the 1990 amendments provided for specific requirements and schedules depending on the severity of nonattainment for ozone, carbon monoxide and particulate matter.117 State flexibility in the SIP process has also been subject to additional requirements and controls to deal with the persistent and difficult issue of interstate air pollution. To be sure, some of this stemmed from EPA’s inability or unwillingness to force meaningful consideration of interstate issues for many years. Although section 110(a)(2)(E) of the 1970 Act required SIPs to contain “adequate measures for intergovernmental cooperation, including measures necessary to ensure that emissions of air pollutants [inside the state] will not interfere with the attainment or maintenance” of the NAAQS in another state, EPA’s implementing regulations minimized this requirement by calling for information exchange only. A pair of court challenges by the Natural Resources Defense Council (NRDC) arguing that these regulations were inadequate were rejected.118 This created an opening for states to adopt policies allowing tall stacks and other dispersion techniques that would move air pollution out of their states in order to achieve compliance, creating more downwind pollution in the process.119 By the time Congress stepped in to deal with this in 1977, more than a dozen states and hundreds of sources had taken advantage of the “tall stacks” loophole.120 And despite this effort by Congress to force more consideration of interstate pollution, EPA continued throughout the 1980s (and much of the 1990s) to avoid using the full extent of its SIP authority to force more attention by states to downwind pollution impacts. Congress thus stepped in again in the 1990 amendments with several provisions intended to further strengthen the SIP process and promote regional air planning and coordination, particularly with respect to ozone. The Ozone Transport Commission and the voluntary Ozone Transport Assessment Group both established important multistate and multistakeholder processes to fashion regional solutions to the ozone problem in the Northeast.121 Among other things, these efforts directly influenced EPA’s 1998 NOx SIP call and laid the groundwork for the subsequent CAIR and CSAPR rulemakings. Here again, 117

118

119

120 121

See 42 U.S.C. Ch. 85, subch. I, pt. D, subpts. 2–4. These provisions distinguish areas by their level of nonattainment and impose stricter measures with tailored deadlines for areas depending on the severity of nonattainment. Five categories apply for ozone: “marginal,” “moderate,” “serious,” “severe,” and “extreme.” The CO and PM provisions include two categories: “moderate” and “serious.” See NRDC v EPA, 483 F2d 690 (8th Cir. 1973); NRDC v. EPA, 494 F.2d 519 (2d Cir. 1974); see also Patton supra note 43. Revesz, Federalism and Interstate Environmental Externalities, supra note 7. See also RICHARD L. REVESZ & JACK LIENKE, STRUGGLING FOR AIR: POWER PLANTS AND THE WAR ON COAL 82–99 (2016) (discussing problem of tall stacks and downwind pollution). See Patton, supra note 43 at 10162. See Ann Carlson, Iterative Federalism and Climate Change, 103 N.W. UNIV. L. REV. 1097 (2009) (discussing OTC and OTAG).

The Clean Air Act’s National Ambient Air Quality Standards

45

though, the SIP process (and the statutory language of the good neighbor provision) has limited the effort to create robust multistate trading programs as a means for dealing with interstate pollution. Needless to say, developing and revising SIPs takes considerable time and resources. The process has been criticized on various grounds, including charges that it is overly bureaucratic, focuses too heavily on attainment demonstrations, is limited to single pollutants and does not adequately consider interstate transport issues.122 The SIP process has also long been a source of litigation between EPA and the states – something that will surely continue as states struggle with attainment of increasingly stringent NAAQS for pollutants such as ozone.123 Despite these shortcomings and challenges, however, the SIP process, at least in its better moments, continues to provide an important space for states to tailor their air pollution control efforts based on their own priorities and to experiment with new ideas and approaches. As such, it is a good example of how flexible implementation can contribute to political durability.

2.6 flexible implementation As discussed earlier, the NAAQS program was designed to combine ambitious health-based standards that would apply uniformly across the country with flexible state-led implementation. In the original 1970 version of the program, Congress intended that the SIP process would give states maximum flexibility in deciding how to allocate the emissions reductions necessary to achieve the NAAQS. While Congress came back in 1977 and again in 1990 with new provisions that constrained some of that flexibility, the basic model of cooperative federalism at the heart of the NAAQS program continues to provide the most important example of flexible implementation under the program. In effect, the SIP process gives states significant flexibility in determining how implementation will proceed and provides a forum for affected sources to seek flexibility with respect to their own compliance obligations. Complementing the flexibility inherent in the SIP process, several other statutory provisions provide for flexibility with respect to the timing of implementation. These include, most obviously, the decisions by Congress to modify and relax compliance deadlines and to create specific compliance schedules that are tied to the current level of nonattainment for ozone, CO and PM. But they also include statutory provisions allowing for postponements of deadlines and variances from the normal 122 123

See NATIONAL RESEARCH COUNCIL, AIR QUALITY MANAGEMENT IN THE UNITED STATES, 128–32 (2004). See, e.g., Mississippi v. EPA, 744 F.3d 1334 (D.C. Cir., 2013) (addressing challenges to the 2008 ozone NAAQS); Findings of Failure to Submit State Implementation Plan Submittals for the 2008 Ozone National Ambient Air Quality Standards (NAAQS), 82 Fed. Reg. 58,118 at 58,120 (Dec. 11, 2017) (finding that New Jersey, Illinois and California had failed to meet various requirements in their SIPs for the 2008 ozone NAAQS). Various petitions have also been filed challenging the 2015 NAAQS for ozone.

46

William Boyd

SIP process as well as EPA’s authority to waive compliance deadlines for certain stationary sources if sufficient control measures are unavailable and “the continued operation of such sources is essential . . . to the public health or welfare.”124 EPA also enjoys some limited flexibility in discharging its responsibilities to establish and revise the NAAQS. The policy judgment at the heart of section 109, for example, provides for some discretion on the part of the EPA administrator in determining the precise concentration that will protect public health “with an adequate margin of safety.” More recently, EPA efforts to use its authority under the SIP and good neighbor provisions to deal more effectively with interstate pollution can be seen as examples of flexible implementation. The 1997 joint rulemaking for ozone and PM2.5, the NOx SIP call, and the CAIR and CSAPR all demonstrate an effort on EPA’s part to use its authority under the statute in more creative and flexible ways than it has in the past. Finally, the White House has played an important, though limited, role in flexible implementation of the NAAQS program. The 1997 Presidential Implementation Memorandum on Ozone and PM NAAQS, for example, stressed the need for flexibility in implementation, the importance of regional approaches and the attractiveness of market-based approaches.125 The memorandum provided guidelines for EPA to follow in implementing the new standards and strongly endorsed regional approaches “to respond to the fact that pollution travels hundreds of miles and crosses many State lines.”126 This provided important political support to EPA as it sought to use its existing authorities in more creative and flexible ways to fashion regional market-based approaches to controlling ozone and PM pollution, first with the NOx SIP call and then with the CAIR and CSAPR rulemakings. The 2011 letter from OIRA Director Cass Sunstein to EPA Administrator Lisa Jackson provides a different example of White House involvement in the NAAQS process – one directed at slowing down EPA in the context of the controversial revision of the 2008 ozone NAAQS.127 In returning the rule to EPA and stating emphatically that the President “does not support finalizing the rule at this time,” the White House flexed its political muscle to force EPA to adjust the timing of the ozone revision.128 Regardless of one’s views on the merits of this action, it does represent a rather dramatic example of the use of Presidential authority to inject additional flexibility into the NAAQS process. That said, it is not at all clear whether this contributed to the long-term political durability of the program, and it raises important questions about whether this sets any sort of precedent for the future. 124 125 126 127

128

See section 110(f)(1), 84 Stat. 1683. See 62 Fed. Reg. 38,421 (1997). Id. at 38,421. See Letter from Cass R. Sunstein, Administrator of OIRA, to Lisa Jackson, Administrator of EPA, dated September 2, 2011. Id.

The Clean Air Act’s National Ambient Air Quality Standards

47

2.7 environmental performance By any measure, the environmental performance of the NAAQS program has been impressive. From 1970 to 2015, aggregate national emissions of the six criteria pollutants declined by an average of 71 percent while the US population grew by 57 percent, gross domestic product (GDP) by 246 percent, vehicle miles traveled by 184 percent and energy consumption by 44 percent.129 These emissions reductions have led to dramatic improvements in air quality. Between 1980 and 2015, national ambient concentrations declined by 99 percent for lead, 84 percent for carbon monoxide, 84 percent for sulfur dioxide, 60 percent for nitrogen dioxide and 32 percent for ozone. Fine-particle concentrations declined by 37 percent and coarse-particle concentrations by 36 percent between 2000, when trends data began for fine particles, and 2015130 (see Figure 2.4). With substantial improvements in air quality have come very significant public health benefits. As a result of reductions in PM2.5 and ozone, for example, a 2011 EPA study on the benefits and costs of the CAA estimated that close to 240,000 premature deaths will have been avoided by 2020.131 Unhealthy air days – a broad measure that combines PM and ozone concentrations into an air quality index – have declined continuously across the country since the index was first used in the early 2000s.132 These improvements in air quality have resulted in the avoidance of hundreds of thousands of hospitalizations for asthma and other respiratory and cardiovascular ailments, as well as avoidance of millions of lost work and school days.133 Likewise, as a result of the massive reduction in ambient concentrations of lead due to the phaseout of leaded gasoline under the CAA’s fuels provisions and the NAAQS program, blood lead levels in the US population, notably children, have declined dramatically, leading to substantial reductions in cognitive impairment that comes from lead exposure.134 Taken together, one estimate puts the public health benefits from improvements in air quality to 2020 as a result of the CAA at more than $2 trillion.135 129 130 131

132 133 134

135

These figures come from the EPA Air Trends site, https://www.epa.gov/air-trends. Id. See U.S. EPA, THE BENEFITS AND COSTS OF THE CLEAN AIR ACT FROM 1990 TO 2020 (2011). The vast majority of this comes from reductions in PM2.5. EPA Air Trends, https://www.epa.gov/air-trends. U.S. EPA, THE BENEFITS AND COSTS OF THE CLEAN AIR ACT FROM 1990 TO 2020 (2011). See Herbert L. Needham, The Removal of Lead from Gasoline: Historical and Personal Reflections, 84 ENVIRONMENTAL RESEARCH 20 (2000) (discussing removal of lead from gasoline in the United States and substantial reductions in children’s blood lead levels that followed); Phillipe Grandjean & Philip J. Landrigan, Developmental Neurotoxicity of Industrial Chemicals, 368 LANCET 2167, 2169–70 (2006) (“A 90% reduction in blood-lead concentrations followed the termination of lead additives in gasoline.”); Philippe Grandjean & Philip J. Landrigan, Neurobehavioral Effects of Developmental Toxicity, 13 LANCET NEUROL. 330, 335 (2014) (citing studies showing that the prevention of neurodevelopmental toxicity resulting from phaseout of lead in gasoline in the United States has generated economic benefits of $200 billion in each annual birth cohort since 1980, leading to “an aggregate benefit in the past 30 years of over $3 trillion”). U.S. EPA, THE BENEFITS AND COSTS OF THE CLEAN AIR ACT FROM 1990 TO 2020 (2011).

William Boyd

48

Carbon Monoxide

Ozone 0.12

10 9

0.1

8

0.08

6

PPM

PPM

7 5

0.06

4 0.04

3 2

0.02

1 0 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016

1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016

0

31% decrease

85% decrease

100 90 80 70 60 50 40 30 20 10 0

Particulate Matter (PM2.5) 16 14

ug/m3

12 10 8 6 4 2 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016

0 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 2010 2013 2016

ug/m3

Particulate Matter (PM10)

42% decrease

39% decrease

NO2 & SO2

99% decrease

180 160 140 120 100 80 60 40 20 0 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016

ppb 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 2010 2013 2016

ug/m3

Lead 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0

NO2 61% decrease

SO2 87% decrease

figure 2.4 Progress in reducing criteria pollutants

Despite such remarkable success, however, there are still a number of areas around the country that continue to struggle with stubborn and persistent nonattainment issues, particularly for ozone.136 Los Angeles, for example, may never be able to achieve attainment with the ozone NAAQS, raising important questions about the success of the program in delivering clean air to all Americans. 136

Information on NAAQS nonattainment status across the United States by criteria pollutant is compiled by EPA in the so-called Green Book, available at https://www.epa.gov/green-book. For a map of areas in the United States that are in nonattainment for one or more of the NAAQS, see https://www3.epa.gov/airquality/greenbook/mapnpoll.html.

The Clean Air Act’s National Ambient Air Quality Standards

49

More generally, it is important to recognize that aggregate trends showing substantial progress in reducing ambient concentrations of criteria pollutants do not reveal the unevenness across and even within Air Quality Control Regions and the associated distributional impacts of poor air quality on specific populations. The techniques of measuring and monitoring air pollution, as well as the form of the standard itself (average concentrations over particular time periods), can also hide peak air pollution episodes over the course of a day, a month or even a season. And the existing air quality monitoring network often does not reveal local hot spots of high ambient air pollution.137 Put simply, a nontrivial portion of the population of the United States continues to breathe unhealthy air, and far too many people continue to die or get sick as a result. With new, more stringent standards for ozone and PM2.5, moreover, the burdens on nonattainment areas will increase. This is particularly true for ozone, given that the most recent revision to the standard in 2015 is pushing up against background levels in some areas. All of which makes the distributional issues associated with the NAAQS program’s impressive overall record of environmental performance increasingly important. Without question, the NAAQS program can rightly claim to be among the most (if not the most) successful major programs in US environmental law. But its work is still not finished, and the remaining improvements in air quality needed to meet the original 1970 goal that all areas of the country would be in attainment with all the NAAQS will likely be even harder to achieve than the progress already made. In the meantime, millions of Americans will continue to live in areas of the country with air quality that still does not protect public health with an adequate margin of safety.

2.8 economic impacts Although the benefits of the NAAQS program far outweigh the costs, the program is expensive with widespread impacts across multiple economic sectors.138 To take one recent example, the 2015 revision of the ozone NAAQS, which reduced the allowable concentration of ozone in the ambient air by 0.05 ppm (from 0.075 to 0.070 ppm) was projected to result in compliance costs ranging from $12 billion to $20 billion.139 Every time a NAAQS is revised, each state must submit a new SIP 137

138

139

See JOHN WARGO, GREEN INTELLIGENCE: CREATING ENVIRONMENTS THAT PROTECT HUMAN HEALTH 207–40 (2009) (discussing problems of existing air quality monitoring and averaging). See also Ann Carlson, The Clean Air Act’s Blind Spot: Microclimates and Hotspot Pollution, 65 UCLA L. REV. 1036 (2018). Various cost-benefit analyses of the program confirm very large net benefits. U.S. EPA, THE BENEFITS AND COSTS OF THE CLEAN AIR ACT FROM 1990 TO 2020 (2011) (finding that central benefits estimate exceeds costs by a factor of more than 30:1; high benefits estimate exceeds costs by 90:1; low benefits estimate exceeds costs by 3:1). See SUMMARY OF THE UPDATED REGULATORY IMPACT ANALYSIS (RIA) FOR THE RECONSIDERATION OF THE 2008 OZONE NATIONAL AMBIENT AIR QUALITY STANDARD (NAAQS) at S1–4 (indicating annual costs ranging from $12 billion to $20 billion as a result of change in ozone standard from 0.075 to 0.070 ppm).

50

William Boyd

that includes provisions relating to the many and varied sources of the pollutant in question. More stringent standards can also push air quality control regions into nonattainment (or a more serious category of nonattainment), which, in turn, triggers additional controls and requirements for sources of emissions. All of this can be costly. The new 2015 ozone standard, for example, is expected to push some Air Quality Control Regions from attainment into nonattainment while raising the severity of nonattainment for a number areas already out of attainment.140 Given that the new ozone standards are starting to push up against background levels in some areas (e.g., the Colorado Front Range), the costs imposed on these areas in terms of foregone economic activity and jobs will likely increase. Congress, of course, has long been aware that a program built around healthbased ambient environmental standards has significant economic impacts. The 1967 Air Quality Act, for example, called for a study on economic impacts,141 and the 1977 and 1990 amendments both mandated studies looking at the costs and benefits of the CAA as a whole.142 As noted earlier, the general conclusion that results from these studies (and others) is that the benefits of the CAA, and the NAAQS program in particular, have far outweighed the costs. In contrast to other major regulatory efforts under the CAA, which can impose large costs on a specific sector or even specific facilities within a sector (see, e.g., the Mercury Air Toxics Standards or the Clean Power Plan), the costs associated with the NAAQS program (and with revisions to individual NAAQS) are often more diffuse, are filtered through many different SIPs and kick in over an extended time frame. This likely has an important impact on the political economy of the program that may contribute to its durability. Opposition is less focused and less intense but potentially more widespread.

2.9 political economy, durability and flexibility The NAAQS program has proven to be highly durable since it was established in 1970. From a public choice perspective, this may seem odd given a diffuse class of regulatory beneficiaries (the public at large), the fact that the most important beneficiaries (children and future generations) are not exactly top of mind for elected officials and the large number of affected sources. Although the NAAQS program would almost certainly not enjoy the widespread bipartisan support today 140

141

142

EPA completed the designations for attainment and nonattainment for the 2015 ozone NAAQS in the spring and summer of 2018. See EPA, Air Quality Designations for Ozone, available at https://www .epa.gov/ozone-designations. The Air Quality Act of 1967 called for “a detailed estimate of the cost of carrying out the provisions of this Act; a comprehensive study of the cost of program implementation by affected units of government; and a comprehensive study of the economic impact of air quality standards on the Nation’s industries, communities, and other contributing sources of pollution.” Section 2, 81 Stat. 505. See, e.g., U.S. EPA, THE BENEFITS AND COSTS OF THE CLEAN AIR ACT FROM 1990 TO 2020 (2011); U.S. EPA, THE BENEFITS AND COSTS OF THE CLEAN AIR ACT, 1970 TO 1990 (1997).

The Clean Air Act’s National Ambient Air Quality Standards

51

that it benefited from in 1970 and again in 1977 and 1990, the program has worked well enough and long enough to give it a certain amount of staying power that makes it hard to undo even if it has outlived the bipartisan support that produced it. Put another way, the continued success of the program, imperfect as it is, may be one of the most important components of its overall political durability. The real and tangible nature of the benefits of the NAAQS program – improved air quality and avoided premature deaths – make it hard to attack; few politicians want to run against clean air. And because the costs are often spread across many different entities, accrue over extended time frames and are buried in many different SIPs, they are in some ways less likely to provide a basis for focused opposition. The program has also benefited from the fact that it was able to mature over a twodecade period (1970–90) marked by active engagement from Congress, EPA and the federal courts. This allowed for important statutory adjustments to the program as it evolved and confronted new problems, creating a set of expectations with respect to its day-to-day workings. In the process, the NAAQS reached deep into the organizational and administrative capacities of the states, mobilizing substantial state resources in the fight against air pollution. And all the while, citizen suits and public participation kept the pressure on and forced agency action – even and especially in cases that were politically fraught. As this chapter has demonstrated, specific design features of the program have also contributed to its durability. Credible goal setting combined with a clear signal that the NAAQS will be reviewed every five years and revised as appropriate has created a clear set of expectations about the NAAQS among the states and affected sources. Everyone knows that the wheels are turning – that the program is subject to a continuous ratchet. Moreover, EPA’s efforts to develop a robust and transparent set of internal procedures for the NAAQS review process, combined with independent scientific review and layers of public comment and participation, have allowed for long-term signaling and vetting of proposed changes well before any final rules are issued.143 Finally, the provisions allowing for citizen suits and public participation have created an additional check on the program, leading to court-imposed deadlines and schedules for EPA action and providing important political cover for the agency to move forward on controversial rulemakings. A large part of the success of the program also derives from the fact that it was designed to be flexible. This is perhaps most apparent in the requirement that the 143

See Fisher et al., Rethinking Judicial Review of Expert Agencies, supra note 3 at 1689–90 (“Because the air quality of the entire nation is riding on the [NAAQS] (as well as the compliance requirements for the millions of sources of pollution), a diverse set of interest groups closely follows EPA’s NAAQS process and participates vigorously in it.”). See also CROLEY, REGULATION AND THE PUBLIC INTERESTS, supra note 68 at chap. 9 (describing the robust administrative process followed by EPA, marked by extensive external peer review and layers of public comment, combined with a commitment to flexibility on the timing of implementation, as critical in allowing the controversial 1997 ozone and PM2.5 rules to survive in the face of strong opposition from industry, state and local governments and Congress).

52

William Boyd

NAAQS be reviewed every five years and revised as appropriate to take account of new scientific information. But flexibility is also built into the program at multiple levels. The policy judgments at the heart of section 109’s standard-setting exercise, the SIP process, adjustments to deadlines and compliance schedules, creative use of existing authorities as a basis for EPA rulemakings and targeted White House interventions are all important examples of flexibility. To be sure, there are limits to this flexibility, and the program continues to struggle with the original goal of attaining all the NAAQS across the entire country. Substantial numbers of people continue to live in areas that are out of attainment for at least one of the NAAQS. EPA’s long-standing difficulties in creating a robust trading program under the NAAQS program to deal with regional air pollution problems have stemmed in large part from the limits of the existing statutory language and Congress’s inability or unwillingness to further revise the program. Similarly, the traditional single-pollutant approach of the NAAQS and the singlestate approach of the SIP process have further blocked the development of regional multipollutant approaches that might be more effective and efficient.144 Going forward, it seems unlikely that the NAAQS program will be dismantled any time soon, much less that it will somehow complete its work in protecting public health. It is like a machine that continues on under its own power, and it will surely continue to deliver major public health benefits for years to come. But whether it will be able to fully live up to its purpose and potential – to fully achieve the ambitious goals of the program as established almost a half century ago – will likely depend on new statutory adjustments and modifications. To state the obvious, this seems highly unlikely in the current political environment, and of course, there is always a risk that Congress could do long-term damage to the program if it were ever to reengage. Thus, while the NAAQS program has struggled over the last twenty-five years in the absence of robust engagement and support from Congress, it seems that EPA and the courts (and the public at large) will have to continue to find ways, at least for the foreseeable future, to work with the program we have and to continue adapting it to deal with new and persistent problems.

2.10 conclusions: lessons for energy and climate policy No policy is perfect, and few work as intended. Major government programs such as the NAAQS are always works in progress – complicated political undertakings crafted under a particular set of circumstances and legal constraints, informed by particular understandings of problems and based on a particular coalition of supporters. If they are to survive beyond the conditions of their making, such programs must be able to evolve and adapt to new circumstances, new understandings and 144

See NATIONAL RESEARCH COUNCIL, AIR QUALITY MANAGEMENT IN THE UNITED STATES 271 (2004) (discussing the importance of regional multipollutant approaches to deal with ozone, PM and regional haze).

The Clean Air Act’s National Ambient Air Quality Standards

53

new political landscapes. Doing that while holding onto their core principles and continuing to deliver on their original objectives is a sure sign of durability.145 Viewed in this light, the NAAQS program has performed admirably over the last half century. In the process, it has generated a set of experiences that hold important lessons for future efforts to craft policies that successfully combine durability and flexibility in an effort to deal with long-lived problems such as global climate change. As this chapter and others have demonstrated, there are specific design choices tied to particular mechanisms, instruments and authorities that can help to strike the right balance between flexibility and durability and allow a program to survive over extended time periods.146 Core structural features of the NAAQS program that would seem to be relevant considerations in an effort to craft energy and climate policy include the five-year NAAQS review, the role of independent scientific evaluation as part of the NAAQS process, the model of cooperative federalism and the flexibility of the SIP process and citizen suits. In addition, EPA’s efforts to use its authority under the SIP process and the good neighbor provisions to deal with regional transport issues are an example of the ways in which broad, even if longdormant, statutory provisions can be mobilized to fashion responses to new and persistent problems. Administrative process has also been a critical part of the success of the NAAQS program. EPA’s efforts to develop and refine its own internal process for NAAQS review and revision – an exercise that played out over many years – has resulted in a robust science-based approach to the NAAQS that allows for multiple layers of scientific review (formal and informal), extensive public participation and careful vetting of proposed revisions well before any formal proposals are made. This has not only strengthened the proposed revisions once they are made but also has provided an important signal to states and the regulated community regarding the content of any proposed revisions. And it has given comfort to the federal courts in their review of EPA’s efforts, providing strong evidence that the Agency has more than satisfied the requirements of reasoned decision making. These commitments to and elaborations of process are not outcomes or features of the program that can be reduced to a simple set of design choices. Rather, they look more like organic, emergent properties of the program that took considerable time to develop based on years of trial and error. To that end, it is important to recognize that durability and flexibility are about more than a set of design choices. They cannot be reduced to a recipe that will guide future policy designs. There is no single portfolio of instruments and authorities that 145

146

See Chapter 6. Obviously, some policies are politically durable, even while they fail to deliver on their initial objectives. See Ann Carlson & Robert W. Fri, Designing a Durable Energy Policy, 142 DAEDALUS 119 (2013). See also Richard J. Lazarus, Super Wicked Problems and Climate Change: Restraining the Present to Liberate the Future, 94 CORNELL L. REV. 1153 (2009).

54

William Boyd

can be plugged in and optimized for the next set of challenges. Put another way, policy instruments and mechanisms are not widgets, even though it is sometimes helpful to think of them as such. They don’t always work as intended when we transplant them from one context to another, something we need to recognize as we set forth on the truly daunting task of trying to design energy and climate policy for the next half century. It is a mistake, therefore, to see a complex program such as the NAAQS as simply a collection of rules, instruments, actors and authorities from which we can draw lessons regarding the elements of flexibility and durability. The program is more than the sum of its parts, and any effort to summarize the key features of the NAAQS program that have allowed it to be flexible and durable needs to be complemented with an effort to understand the program as a whole and over time – how it has evolved and taken on new features, how it has developed new and thicker connections (internal and external) across various domains and constituencies and how it has responded to political and legal challenges. On this broader register, one of the most important reasons why the NAAQS program has been able to survive (and even thrive) over the last half century is that it was able to mature over its first two decades with active involvement by all three branches of government, particularly Congress. Many of the most important features of the NAAQS program were added and subsequently revised and strengthened by Congress in the 1977 and 1990 amendments, including the five-year mandatory review, the Clean Air Science Advisory Committee, the PSD and nonattainment programs and new and stronger provisions to deal with regional transport issues, among others. Had Congress not stepped back in to revise the program, we can say with some confidence that it would not be nearly as effective as it is today. In the absence of future statutory updates, moreover, the NAAQS program will likely continue to struggle with persistent problems such as regional, interstate air pollution. It may be that federal climate and energy policy, assuming that it is even possible to get comprehensive legislation in the future, will not have the luxury of ongoing constructive engagement by Congress. If true, this may lead to different design choices at the outset. That is, if we assume that Congress will likely not be available to come back and make important adjustments as it did with the NAAQS program, the initial choices in designing the program may need to be different. This might argue, for example, for more administrative discretion and flexibility in adjusting the program. But there are obvious limits to how far this approach can go. History matters in at least one other respect as well. With the original NAAQS program, Congress was not writing on a completely blank slate, but it did not face major constraints in terms of preexisting regulatory regimes (at the state or federal level), and it was drafting the new legislation in the midst of an expansive bipartisan lawmaking moment. With the exception of California and a few other nascent state efforts, there was no extensive record of air pollution regulation on which to draw.

The Clean Air Act’s National Ambient Air Quality Standards

55

Prior federal efforts had been largely limited to trying to assist and nudge the states to act, and almost everyone recognized that strong federal action was needed. With greenhouse gases, the context will surely be quite different. Bipartisan majorities seem to be a thing of the increasingly distant past. If and when federal legislation starts to take shape, moreover, it will have to confront a much more complicated landscape of preexisting efforts and regulatory models given ongoing efforts by states to move forward with all manner of energy and climate policies. With much of the focus on the electricity sector, any such effort will also have to contend with the complex regulatory framework for electricity – one that involves a different structure of federalism, a different set of state regulatory agencies and a more diverse set of regulatory models across the country. Long-standing preferences for market-based approaches will also likely exert an important influence on instrument choice and program design. None of this is intended to suggest that the experience of the NAAQS program does not hold important lessons for future efforts to craft energy and climate policy. As this chapter has demonstrated, there are many valuable lessons in the NAAQS experience – for climate policy and beyond. But perhaps one of the most important lessons is that we need to be careful about drawing too many firm lessons about individual design choices and mechanisms; that we need to recognize that it is the interactions between history, structure and process that ultimately shape these programs and provide the conditions for their success.

1965

1970

Clean Air Act Amendments of 1970 direct EPA to set NAAQS and identify and write standards for HAPs, establish New Source Performance Standards, and direct states to write SIPs

CAA Amendments of 1977 1975 establish New Source Review program and direct EPA to write technology-based standards for HAPs 1980

1985

1990

CAA Amendments of 1990 establish list of HAPs, set more stringent limits for SO2 and NOx emissions and establish SO2 allowance trading program

1995

2000

Clean Air Interstate Rule largely replaces NOx Budget Trading Program

NOx Budget Trading Program begins 2005

2010

EPA issues CPP Final Rule under Section 111(d) of the CAA EPA issues CPP Repeal Proposed Rule

2015

Cross-State Pollution Rule replaces CAIR due to North Carolina v. EPA, 531 F.3d 896 (D.C Cir. 2008)

3 Stationary Sources, Movable Rules Intransigence and Innovation under the Clean Air Act Hannah J. Wiseman

3.1 introduction The regulation of stationary sources of air pollutants such as electric power plants and factories under the Clean Air Act (CAA) presents a complex, seemingly paradoxical story of intransigence and innovation. On the one hand, under relatively rigid “brightline” thresholds that define new and existing plants, the CAA grandfathers in old sources built before certain new statutes and rules were issued and places stringent air pollution limits on new sources. This allows many of the largest polluters to continue polluting1 and discourages innovative modifications or replacements that could substantially reduce pollution. Additionally, the Act’s similarly rigid definitions of major and minor sources have allowed certain sources to continue operating without facing the stricter technologybased controls that apply only to the major sources. On the other hand, the CAA – despite its tendency to stymie some innovations – has accomplished dramatic pollutant reductions, in part through technology-based requirements that mandate continuous consideration of improved pollutant reduction technologies available for each new source built. The United States once suffered from air pollution that rivaled modern-day problems seen in developing countries such as China and India. Smog permeated the air, choking and even killing individuals who dared to walk outdoors. In infamous incidents such as the 1948 Denora, Pennsylvania, smog, at least twenty people died, and many more were sickened when a blanket of air pollution from industrial plants enveloped the city for days.2 Thousands of people in the United States still experience health problems and early deaths as a result of air pollution,3 but the CAA has made great progress in part due to stationary source reductions.4 1 2 3

4

NATIONAL RESEARCH COUNCIL, AIR QUALITY MANAGEMENT IN THE UNITED STATES 183 (2004). DEVRA DAVIS, WHEN SMOKE RAN LIKE WATER (2003). See, e.g., Fabio Caiazzo et al., Air Pollution and Early Deaths in the United States, Part I: Quantifying the Impact of Major Sectors in 2005, 79 ATMOSPHERIC ENV’T 198 (2013). Numerous factors have contributed to the reduction in US air pollutants over time, including, among others, increased income and associated demand for clean air as well as technological innovation. But there is evidence that the CAA has played a causal role. For a discussion of various studies attempting

58

Hannah J. Wiseman

Progress on the stationary source front appears to result from several factors. Old, high-polluting plants can only operate for so long. As the US economy has grown, with accompanying demands for more electricity and other products, new, more efficient stationary sources with lower emissions have emerged.5 Additionally, the CAA’s regulatory design – which combines mechanisms for adapting and improving regulatory requirements over time and provisions for flexible implementation of the Act – likely plays a part in the pollutant reductions achieved. The Acid Rain Program (ARP), discussed in more detail in Chapter 6, also has been a major cause of pollutant reductions. For example, in 2005 the program accounted for 35 percent of overall sulfur dioxide (SO2) reductions in the electric power sector and the reduction of approximately 50 percent of anticipated nitrogen oxide (NOx) emissions.6 Other trading programs also have contributed to the CAA’s successes.7 With respect to regulatory design, the CAA appears to have had success due to several specific facets of its policy approach. These ensure that the Act adapts to new circumstances and science over time and allows for flexible implementation of federal statutory requirements. First, the CAA requires something that few other environmental statutes do: it places a firm, explicit threshold on the total acceptable amount of air pollution in the ambient air (the National Ambient Air Quality Standard [NAAQS])8 and then limits individual sources of air pollutants based on this threshold. This is a relatively fixed standard that sends a clear signal to states and regulated sources of the goal that must be achieved. Yet, through a key adaptation mechanism, the CAA requires periodic review of the standard9 and allows the Environmental Protection Agency (EPA) to make the standard more stringent if this will achieve health-based goals. This helps to ensure continued progress toward durable CAA air quality goals. Stationary sources in areas that have failed to achieve the firm threshold – even old, grandfathered sources – are subject to more stringent pollutant limits. And in another example of an important adaptation tool, when new sources are built in these nonattainment areas, they must offset the additional pollution they will contribute and even demonstrate that they will achieve net progress toward reducing air pollutants. Regulating based on the total environmental harm caused by particular sources – regardless of the number of sources that emerge over time – appears to be a key to success.10 In the many environmental fields in which these “total harm” limits do not

5

6

7 8 9 10

to identify the cause of air pollutant reductions, see NATIONAL RESEARCH COUNCIL, AIR QUALITY MANAGEMENT, supra note 1, at 37. For a graphic showing emissions declining as the economy has grown, see U.S. EPA, COMPARISON OF GROWTH AREAS AND EMISSIONS, https://www.epa.gov/sites/production/files/2017-02/1970-2015_baby_ graphic.png. For a general discussion of the higher emissions and lower efficiency of older plants, see NATIONAL RESEARCH COUNCIL, AIR QUALITY MANAGEMENT, supra note 1, at 184. U.S. EPA, ACID RAIN PROGRAM 2005 PROGRESS REPORT 2 (2006), https://www.epa.gov/sites/production/ files/2015-08/documents/2005report.pdf. Id. 42 U.S.C. § 7409(a)(1). 42 U.S.C. § 7409(d). See Hannah J. Wiseman, Remedying Regulatory Diseconomies of Scale, 94 B.U. L. REV. 235 (2014).

Stationary Sources, Movable Rules

59

exist, the emergence of numerous new, individual sources of pollutants quickly dwarfs progress toward environmental quality. Even though there are individual pollutant limits on these sources, their sheer increase in number causes the limits to be insufficient.11 A second important attribute of the CAA stationary source program appears to be its combination of durable federal controls that establish a minimum “floor” of stringency with flexible mandates, which require that states place pollution limits on stationary sources within their borders. The CAA requires states to issue plans for implementing the Act and achieving its air quality limits noted earlier, and it specifies the basic required framework of these plans while leaving states substantial latitude in determining how to achieve the NAAQS. The plans must demonstrate not only that the state has limits on air pollution designed to achieve federal air quality standards but also that it has adequate resources to enforce those limits. It allows states to deploy the most effective strategies for their unique political, economic, technological and geographic conditions while maintaining a durable federal goal.12 States, with the help of federal guidance, also issue individual permits to polluters, and this is a key adaptation mechanism as well as a tool that tends to leverage technological innovation. The CAA requires this case-by-case approach, providing through the New Source Review (NSR) program that major stationary sources may not be constructed or modified without first receiving a permit. This is an important adaptation mechanism because it forces individualized consideration of the best available or demonstrated technologies. Each and every time a source is built or modified, the state issuing the permit approves a particular technology that that source must implement in order to reach a technology-based emission limit. This case-by-case review encourages a ratcheting up of the stringency of pollution controls over time. This review would not be fully effective without another unusual and effective informational instrument. EPA maintains and periodically updates a federal clearinghouse that contains technology options from which states may choose when they issue permits.13 This clearinghouse catalogues the many different types of experimental and proven pollution control technologies installed at stationary sources around the country (and even globally) and is an unusual example14 of the federal government providing key information that states need for effective policy 11

12

13

14

See id. for a discussion of this problem. But see David E. Adelman, Environmental Federalism when Numbers Matter More Than Size, 32 UCLA J. ENVTL. L. & POL. 238 (2014) (arguing that the NAAQS excessively constrain the flexibility of the states). But see Adelman, Environmental Federalism, supra note 11 (arguing that states have too little flexibility to operate above the federal floor due to the NAAQS and NSR requirements and thus that they cannot focus on the sources that now contribute to the most pollution, which are not subject to NSR). U.S. EPA, RACT/BACT/LAER CLEARINGHOUSE (RBLC), https://cfpub.epa.gov/rblc/index.cfm? action=Home.Home. For a discussion of how many sub–federal policy experiments are not in fact experiments or involve relatively haphazard implementation of different approaches due to informational deficiencies, see Hannah Wiseman, Regulatory Islands, 89 N.Y.U. L. REV. 1661 (2014).

60

Hannah J. Wiseman

experimentation.15 Regulatory officials and industry actors bargaining over the technologies to be included in a permit have information on the many approaches that have been tried, which provides an important baseline. And, as discussed in Section 3.2, under a top-down approach that EPA decided to mandate, states choosing the technology that will be required within each permit issued to a stationary source must look to other technologies, most of which are in the database, and select the most stringent one unless they provide adequate justification for a laxer technological requirement.16 Thus, as pollution control technologies improve over time, these technologies are incorporated into new permitting requirements. A third important approach that the CAA takes – one that is also related to the clearinghouse – is the Act’s incorporation of mechanisms for innovation within its statutory design. The federal technology-based emissions standards that states must achieve in issuing each permit are worded in ways that encourage innovation and diffusion of policy improvements among states and sources, and the clearinghouse amplifies this effect. The sources subject to the most stringent pollution limits under the CAA must achieve standards that are written based on the “best available”17 or “maximum achievable”18 control technology, among other standards. And it is important to note that these are technology-based emissions standards, not technology standards. This encourages innovation because it requires sources to meet a particular emissions limit – one defined by the pollutant reductions achieved by similar sources using particular technologies – using whatever technology will meet that limit. Sources might be incentivized to agree to a stringent, innovative pollution control technology simply because the permit is the key to their commencing construction and operation, and quickly obtaining a permit is a high priority. And, with the clearinghouse, if a source in one state implements a new, more effective innovative control technology and the state requires this technology in a permit that it issues, this technology is entered into the clearinghouse. While other states are not required to select this specific technology when issuing their permits, they must follow a top-down approach in which they must justify choices that are less stringent than technologies implemented in other states.19 Finally, other adaptive portions of the CAA that help to ensure that EPA updates standards over time and performs other functions required by the Act are the citizen suit provisions within the Act. These allow nonprofit watchdog groups and other entities to require EPA to hew to the language of the CAA and perform nondiscretionary duties under the Act. When one entity responsible for implementing the 15

16 17 18 19

For discussion of information-based limits on state and local policy experimentation and the need for federal information collection and dissemination, see id. See infra notes 138–146 and accompanying text. 42 U.S.C. § 7475(a)(4). 42 U.S.C. § 7412(g)(2)(A). See infra note 139and accompanying text.

Stationary Sources, Movable Rules

61

CAA fails, citizens can fill this gap, thus helping to contribute to the overall durability of CAA goals. Firm yet adaptable standards and flexible implementation mechanisms under the CAA have not always contributed positively to durable CAA requirements for maintenance of and improvements in air quality. First, when these mechanisms are combined with inadequate resources for implementation or when they place firm limits on pollutants for which the risks and best mechanisms for controlling the pollutants are complex, they can backfire. For example, Congress’s initial directive to EPA to list hazardous air pollutants (HAPs) – a directive designed to be adaptive so that EPA could address new risks as they became apparent and could use its own expertise and judgment to identify those risks – resulted largely in inaction.20 This was due to the difficulty of accurately characterizing the safe levels of each type of pollutant21 as well as the overwhelming task placed on an underfunded agency. As a result, Congress had to later step in and reduce this discretion by specifically listing the HAPs within the CAA.22 Second, firm, brightline standards designed to ensure progress toward air pollution improvements can backfire in that they present rigid rules with specific numbers or other limits that regulated sources can avoid.23 For example, the numerical limits that separate major from minor sources have allowed certain sources to operate for years just under the threshold,24 and the cutoff that separates old from new sources also allows grandfathered higher-polluting sources to continue operating for long periods of time. Additionally, the flexibility that states have in writing state implementation plans (SIPs) and individual emission controls within permits might sometimes allow states to operate below the federal floor by issuing inadequately stringent SIP approaches or permits and then litigating for years or avoiding EPA’s politically controversial imposition of a federal plan. In other respects, though, there is too little state flexibility. Some legal scholars argue that the central focus on achieving the NAAQS, combined with stringent federal technology-based requirements for new major sources, has overly constrained the states’ ability to focus on other, more important sources that now contribute to more pollution problems.25 20 21 22 23

24

25

See infra note 168 and accompanying text. NATIONAL RESEARCH COUNCIL, AIR QUALITY MANAGEMENT, supra note 1, at 55. Id. This chapter uses the term brightline to refer to standards that include a single, clear rule, such as a numerical limit, rather than a vaguer narrative standard that allows for a variety of compliance options. If a source falls below the brightline rule, it avoids the rule altogether, whereas if it falls at or above the threshold, it is subject to the rule. For a discussion of the difference between brightline and vaguer rules in the property law context, in which the author refers to brightline rules as “hardedged,” see Carol Rose, Crystals and Mud in Property Law, 40 STANFORD L. REV. 577 (1988). See infra notes 230–233 and accompanying text for an example of sources operating just below the threshold that defines a major stationary source. Adelman, Environmental Federalism, supra note 11.

62

Hannah J. Wiseman

This chapter explores the adaptive, nonadaptive, flexible and inflexible aspects of the stationary source portion of the CAA and explores which regulatory tools appear to have contributed to the durability of the Act’s goals for achieving clean air and which appear to have detracted from these goals. Section 3.2 introduces the major programs within the CAA that address stationary sources, regulations that implement these programs and the role of the federal government and states in implementing CAA regulations. That section further discusses the standards’ durability and feasibility in terms of achieving meaningful air pollutant reductions while also accounting for political economy considerations. Section 3.3 then explores air quality improvements that result from CAA stationary source regulations, and Section 3.4 normatively assesses the stationary source portion of the CAA from the perspective of adaptability, flexibility and durability. That section uses case studies from several stationary source types to provide real-world examples within these normative metrics.

3.2 the caa’s major stationary source programs 3.2.1 The Structure of the Act The CAA categorizes sources of air pollutants in a variety of ways, but two of the major distinctions within the Act include mobile sources and stationary sources of air pollutants. Mobile sources such as cars and trucks are regulated at the federal level. Stationary sources, which remain fixed in one location, are more limited in number than mobile sources and are regulated under two different approaches. For ubiquitous criteria pollutants – discussed next – the CAA uses a cooperative federalism scheme that involves both the federal government and states in the regulatory process. States can, and do, take different approaches to regulating emission reductions from stationary sources in a manner that meets or is more stringent than a federal minimum floor for required emission reductions. However, states are meaningfully constrained in their flexibility by specific federal technology-based standards placed on new major sources of pollutants. In contrast to criteria pollutants, the federal government regulates stationary sources of hazardous air pollutants (HAPs) and does not rely on a cooperative federalism approach. The CAA sets emission controls for stationary sources based on the type of pollutant that they emit as well as their size, location, age and type (e.g., various types of industrial plants versus power plants). First, with respect to pollutant type, the CAA sets different emissions limits for criteria (relatively common) pollutants, HAPs, “other” pollutants that do not fall within the criteria or hazardous categories and pollutants that cause acid rain. Additionally, there are distinct limits on sources’ air emissions that contribute to interstate air pollution problems. A stationary source that emits pollutants from all of these pollutant categories is therefore subject to

Stationary Sources, Movable Rules

63

emission limits from at least five different portions of the CAA and from regulatory programs under these parts of the Act. Second, with respect to size, major sources – defined by the amount of pollution they emit or have the potential to emit – are subject to more stringent controls than are smaller sources, which are called minor or area sources depending on the type of pollutant they emit. Third, regarding location, sources located in regions that fail to attain the NAAQS (regions with relatively polluted air – called nonattainment areas) generally face stricter emissions limits than stationary sources within cleaner areas. In nonattainment areas, smaller sources of air pollution are thus classified as “major” because they contribute to further problems in areas that are already polluted. Fourth, sources built before the effective date of new air quality standards – existing sources – are generally grandfathered under old standards, meaning that they may continue to comply with old rules under which they were operating and need not improve their emissions performance to meet newly promulgated standards. There are exceptions, however, in areas with poor air quality, where states must regulate both new and existing sources. Despite this requirement, states have had difficulty regulating grandfathered sources.26 Although there are economic and practical justifications for this size- and time-based threshold approach, this has been a substantial factor that has limited progress in US air pollutant reductions.27 Finally, different types of stationary sources, such as power plants and a variety of industrial units, must achieve different degrees of pollutant reductions depending on the proven emission reductions that have been achieved for these types of plants. When setting national emission limits for stationary sources under the CAA for a particular pollutant, EPA investigates the feasibility of emission reductions achievable for that pollutant by sources around the country within a particular source category – such as coal-fired power plants – and then sets a national emission limit for all plants within this category. States then regulate these sources to ensure that they meet the national emission limit; states issue individualized permits and make case-by-case technology-based emission limit determinations, and they therefore have some flexibility in determining how sources should meet this national limit. These technology-based limits help to allow the CAA to adapt over time as available pollution controls evolve and help to ensure that pollutant reductions under the Act are maintained – thus contributing to the durability of the Act’s goals. 3.2.2 An Introduction to Specific Statutory Provisions Addressing Stationary Sources Within the basic structure of the CAA there are detailed statutory provisions that create a maze of requirements. While many of these requirements send clear signals 26 27

NATIONAL RESEARCH COUNCIL, AIR QUALITY MANAGEMENT, supra note 1, at 183. Id. at 184.

64

Hannah J. Wiseman

to sources that must comply with them, the complexity of the regulations promulgated to implement the requirements can be quite confusing. With respect to the types of pollutants regulated, through major amendments in 1970 – the amendments that created the CAA as we know it today – Congress first directed EPA to list pollutants that in the judgment of the EPA administrator “have an adverse effect on public health and welfare” and are emitted from “numerous or diverse mobile or stationary sources.”28 EPA was then to establish “air quality criteria” for these criteria pollutants, meaning that it was to determine safe levels of these pollutants in the air based on scientific studies of their effects and then set NAAQS for these pollutants.29 As explained in Chapter 2, NAAQS are the maximum levels of criteria pollutants in the air that must not be exceeded in order to protect human health, crops, buildings, and other resources.30 In the 1970 amendments, Congress additionally required that states write SIPs to regulate air emissions from stationary sources in order to achieve the NAAQS.31 As of 2017, there are six criteria pollutants for which EPA has set a NAAQS: sulfur dioxide, nitrogen oxides, carbon monoxide, particulate matter, ground-level ozone, and lead.32 The flexible, broad definition of criteria pollutant and, more generally, the term air pollutant under the CAA allow for periodic updating of the Act through regulation of new pollutants as their risks are better understood. To try to ensure progress toward the NAAQS and prevent further deterioration of air quality as new sources were built, the 1970 amendments also established national emission standards that applied directly to newly built or modified individual major and minor stationary sources and that states were to incorporate into their SIPs. These standards are called standards of performance for new stationary sources,33 or New Source Performance Standards (NSPS). As of 1970, any time an entity proposed to build a new major stationary facility or modify an existing one, the entity had to ensure that it would limit its emissions of criteria pollutants to the amount established by the NSPS if EPA had established a NSPS for that category of source. EPA sets the NSPS for specific sources – typically a numerical, rate-based limit on emissions34 based on the emissions the Agency estimates the source can control as a result of demonstrated emission reductions at similar sources. For sources for which EPA has delegated NSPS implementation to states,35 states issue the NSPS 28 29 30 31 32 33 34

35

Pub. L. No. 91-604, § 108 (1970). Id. at § 108(a)(2). Id. at § 109. Id. at § 110. U.S. EPA, CRITERIA AIR POLLUTANTS, https://www.epa.gov/criteria-air-pollutants. Pub. L. No. 91-604 at § 111. See, e.g., 40 C.F.R. § 60.42 (providing a NSPS for particulate matter from fossil fuel–fired plants in the form of maximum nanograms per joule of plant heat input). Some standards instead require a “general control device” or work practice. See 40 C.F.R. § 60.18. 42 U.S.C. § 7411(c). See also Philip Weinberg et al., Environmental Law and Regulation in New York, 9 N.Y. PRAC., ENVTL. L. AND REG. IN N.Y. § 5:10 (2d ed.) (Oct. 2016 update) (describing how EPA has delegated NSPS permitting for numerous sources to the state of New York).

Stationary Sources, Movable Rules

65

permit with which a new or modified stationary source must comply.36 States have flexibility in requiring the technologies and methods that sources must deploy to achieve the NSPS. These basic federal emission limits on criteria pollutants from new and modified stationary sources, combined with state pollution controls under SIPs, did not do enough to achieve the NAAQS. In another set of major amendments to the CAA in 1977, Congress therefore established strict limits on annual emissions from newly built major stationary sources of pollutants or existing stationary sources that made major modifications.37 These limits are part of a CAA program called New Source Review (NSR). NSR limits exist in addition to NSPS, and unlike NSPS, they are tied directly to the NAAQS in order to help ensure that new sources being built in an attainment area do not cause attainment to “slip” out of compliance and that sources built in nonattainment areas do not contribute to further nonattainment problems. Unlike NSPS, NSR also applies only to major stationary sources. Through NSR, any company proposing to build a new major stationary source of criteria pollutants or to make a major modification to the source must, before commencing construction, obtain a permit that requires the source to meet a federal emission limit. The emission limit varies depending on whether the new source will be constructed in a region that has met or exceeded the NAAQS or is in nonattainment. For attainment regions that have met or exceeded the NAAQS, the new or modified source is subject to a CAA program called Prevention of Significant Deterioration (PSD), or preconstruction review.38 Sources in nonattainment areas are subject to the nonattainment NSR program.39 As with the NSPS program, states must incorporate these emission limits into their SIPs, and states have flexibility to determine the specific practices or technologies that sources must deploy to meet the nonattainment NSR or PSD emission limit. As a result of Congress’s incremental additions of federal emission limits for individual source types, state SIPs now contain numerous programs and permitting requirements, all with the goal of achieving the NAAQS. For example, SIPs must include a program to regulate criteria pollutants from minor new sources.40 SIPS also must implement nonattainment NSR or PSD programs as well as NSPS programs. In some cases, EPA has allowed states to operate their own NSR or PSD programs – provided that these programs meet the federal floor. In other cases, states simply directly implement federal NSR and PSD requirements as part of their SIPs.41 To implement nonattainment NSR and PSD requirements, states issue individualized permits to each new or modified source. These permits contain 36 37 38 39 40 41

See id (describing how state operating permits for sources contain the NSPS). See Pub. L. No. 95-95, §§ 165, 171 (1977). Id. at § 165; 42 U.S.C. §§ 7470–79 (2016). Pub. L. No. 95-95, § 171; 42 U.S.C. §§ 7470–79. U.S. EPA, MINOR NSR BASIC INFORMATION, https://www.epa.gov/nsr/minor-nsr-basic-information. See, e.g., Minnesota Pollution Control Agency, Minnesota State Implementation Plan (SIP) (describing the difference between state nonattainment NSR and PSD programs that meet minimum federal

66

Hannah J. Wiseman

a nonattainment NSR and/or PSD emission limit determined on a case-by-case basis by the state, thus allowing flexibility in state permitting. However, as discussed further below, EPA guidance directs the states to adopt a “top-down” approach to permitting42 in which the presumption is that within each permit the state will adopt the most stringent technology (and associated emission limits) unless it has a specific, EPA-defined justification for selecting a less stringent technology. This still allows room for state discretion but helps to ensure that durable CAA goals of improving air quality are met. In contrast, NSR is so complex that it dominates states’ and EPA’s regulatory focus – as well as regulated sources’ compliance strategies43 – and in some cases has pulled important focus and resources away from other, more problematic pollution sources not covered by NSR.44 For HAPs, Congress in 1970 did not set acceptable ambient levels or direct states to write SIPs for them. Congress instead directed EPA to make a list of these pollutants and set national health-based emission limits for these pollutants – National Emissions Standards for Hazardous Air Pollutants (NESHAPS).45 Congress allowed EPA to delegate to states the responsibility of “implementing and enforcing” the standards.46 Congress later (in 1977) directed EPA to write technology-based standards for HAPs but to study the health risks not addressed by these standards and recommend additional measures necessary to address these health risks.47 Despite the lack of an ambient standard, a major benefit of HAPs is that Congress required EPA to regulate both new and existing sources of these pollutants, thus preventing the grandfathering of older plants.48 In 1970 Congress also created an additional category of pollutants that were not criteria pollutants,49 meaning that EPA had not established air quality criteria for the acceptable levels of pollutants in the air based on these pollutants’ effects,50 and the pollutants also were not HAPs. Congress anticipated that some pollutants would still have problematic effects despite the lack of air quality criteria having been developed for these pollutants and despite the pollutants not having been listed as hazardous. Congress therefore directed EPA to require states to write plans similar to SIPs and to create emission standards for sources of these “other” air pollutants.51 These plans were to apply to existing stationary sources of these other pollutants. EPA has used this provision, section 111(d) of the CAA, to issue regulations for some

42 43 44 45 46 47 48 49 50 51

requirements and “delegated” programs in which the state directly implements federal nonattainment NSR and PSD requirements). See infra note 139 and accompanying text. NATIONAL RESEARCH COUNCIL, AIR QUALITY MANAGEMENT, supra note 1, at 182. Adelman, Environmental Federalism, supra note 11. Pub. L. No. 91-604, at § 112. Id. at § 112(d)(1). See infra note 118 and accompanying text. NATIONAL RESEARCH COUNCIL, AIR QUALITY MANAGEMENT, supra note 1, at 187. Pub. L. No. 91-604, at § 111(d)(1). Pub. L. No. 91-604, at § 108. Id. at § 111(d)(1).

Stationary Sources, Movable Rules

67

pollutants emitted from municipal waste combustors52 and for carbon dioxide emissions from existing power plants.53 It is an important provision that is both flexible and adaptive; it allows for flexible implementation of the CAA by EPA to address new air quality problems that emerge or become better known over time. Additionally, although SO2 and NOx were already was regulated as criteria air pollutants, Congress determined in 1990 that these types of emissions from stationary sources were contributing to acid rain, which damaged forests, crops, buildings and aquatic life and caused other problems.54 Congress therefore amended the CAA to further limit SO2 and NOx emissions from power plants – setting emission limits above and beyond those already contained in SIPs, NSPS, PSD, and nonattainment NSR programs. Specifically, Congress capped the amount of SO2 that each plant could emit.55 In order to make compliance easier, it allowed sources to trade with each other. After establishing caps, allowances were created representing the total amount of SO2 emissions allowed, and allowances were distributed to plants.56 Plants that could easily meet or exceed their emissions cap sold their extra allowances to plants that had more difficulty meeting their individual SO2 cap.57 Finally, a good neighbor provision within the CAA addresses the problem of states approving stationary sources that emit pollutants that contribute to other states’ air pollution problems. EPA, with the help of numerous states that were part of a regional commission addressing the interstate transport of ozone, developed and implemented a NOx Budget Trading Program, in which sources that overachieved pollutant reductions could sell credits to those that had trouble complying. This program was recently expanded and essentially replaced by the Transport Rule (also called the Cross-State Air Pollution Rule), which allows some trading of pollutant reductions among sources.58 This section discusses these CAA programs for stationary sources in further detail and describes how they operate, including the types of stationary sources to which they apply (major or minor, new or existing, etc.) and the specific standards imposed on these sources. It does not discuss SIPs or the acid rain or interstate air emissions programs because these topics are covered in Chapters 2 and 6, respectively. After describing the programs, this section briefly explores the specific role of states in implementing these programs. Finally, it analyzes the adaptability, flexibility and durability of these programs in terms of their ability to achieve and maintain the goals of the CAA. 52 53

54 55 56 57 58

40 C.F.R. § 60.30b–§ 60.39b. 80 Fed. Red. 64,662 (Oct. 23, 2015) (stayed by Order in Pending Case, West Virginia v. EPA, No. 15A773, 2016 WL 502947 (Feb. 9, 2016)). Pub. L. No. 101-549 § 401 (1990). Id. at § 404. Id. at § 403. For detailed discussion of these trading programs, see Chapter 6. See EPA v. EME Homer City Generation, L.P., 134 S. Ct. 1584 (2014), for a discussion of the rule.

68

Hannah J. Wiseman

In general, NSPS, NSR, and the HAPs provisions are all adaptive in that they require sources to implement some of the leading pollution control technologies at the time the sources are built. NSPS and NSR are also flexible in some respects because many states have taken over these programs and issue the permits themselves, selecting the technology that sources must implement but operating under the constraints of federal rules regarding these technologies. And NSPS is both adaptive and flexible in that it has a catch-all category that applies to pollutants that do not fall under the HAPs or criteria pollutant categories. This allows EPA to address newly understood or emerging air pollutant problems. However, both NSPS and NSR contain rigid brightline rules that are in some respects nonadaptive and inflexible because they allow old sources to continue operating largely unregulated. States and EPA generally lack flexible options for regulating these problematic sources. The NSR’s brightline major-minor source threshold presents a similar problem. Additionally, some scholars argue that the complex provisions of NSR mean that states have to devote most of their time and resources to addressing federal NSR standards and cannot adequately regulate other sources that are not covered by NSR but now cause the majority of air pollution.59 3.2.3 Specific Stationary Source Statutes and Regulations: Types of Sources Regulated and Standards Imposed A single stationary source is typically subject to several different portions of the CAA because the source emits several different types of pollutants. For each type of pollutant regulated, this section describes the size, age and type of stationary source to which the CAA applies. 3.2.3.1 Criteria Pollutants As introduced in Section 3.2.2, the CAA limits criteria pollutants emitted from stationary sources in three primary ways. First, states must write SIPs designed to limit emissions from these sources and thus achieve the NAAQS, as discussed in more detail in Chapter 2. Second, regardless of whether stationary sources are in a region that meets or exceeds the NAAQS or are in a nonattainment area (in which the amount of at least one type of criteria pollutant in the air exceeds the NAAQS), new and modified major and minor stationary sources must limit criteria pollutant emissions under the NSPS program. Third, via NSR requirements, new and modified major stationary sources must also meet more stringent annual limits on criteria pollutant emissions, which vary depending on whether the sources are in a region that meets or exceeds the NAAQS or in a nonattainment area. Existing sources (those already built when new regulations became effective) are also subject to 59

Adelman, Environmental Federalism, supra note 11.

Stationary Sources, Movable Rules

69

national emission limits if they are located in nonattainment areas, but states, which implement these limits, have encountered political snags when they attempt to regulate these sources due to the high costs of retrofits.60 All three of these programs help to ensure the durability of achievements made under the CAA by setting mandatory thresholds on acceptable ambient air pollution and limiting pollutants from new sources that could dwarf progress previously achieved. Additionally, the CAA requires constant adaptation of air pollution control approaches as new information emerges about available pollution control technologies. It does this because the NSPS and attainment and nonattainment NSR programs all are implemented through permits issued to individual plants, thus ensuring case-by-case review and updating as information becomes available. But, by exempting existing sources of air pollution except in nonattainment areas, the CAA creates a substantial impediment to this durability; as old sources age and continue to operate, they emit large quantities of problematic pollutants. Additionally, the firm thresholds that separate major from minor sources in the NSR programs block certain needed adaptations over time as small sources increase in number and contribute to pollution problems. new source performance standards (nsps). NSPS – the most basic national emission limits that apply to stationary sources – apply only to new and modified sources, as their name implies. A stationary source that is subject to NSPS includes “any building, structure, facility, or installation which emits or may emit any air pollutant,”61 so EPA also may issue NSPS for sources that emit noncriteria pollutants. NSPS applies to both major and minor sources,62 and it does not distinguish between sources in nonattainment or attainment areas. All “new” sources subject to NSPS must meet the requirement regardless of where they are located. With respect to the age of the source, the program applies to sources that were constructed or modified after the standard established for that source was published or proposed.63 A source is modified if there is “any physical or operational change to an existing facility which results in an increase in the emission rate to the atmosphere of any 60 61 62

63

NATIONAL RESEARCH COUNCIL, AIR QUALITY MANAGEMENT, supra note 1, at 183. 40 C.F.R. § 60.2. See 42 U.S.C. § 7411(b)(1)(A)–(B) (requiring the EPA administrator to list categories of sources that “contribute significantly to . . . air pollution which may reasonably be anticipated to endanger public health or welfare” and to “publish proposed regulations, establishing Federal standards of performance for new sources within such category,” without including the term major). See also U.S. EPA, BACKGROUND ON ESTABLISHING NEW SOURCE PERFORMANCE STANDARDS UNDER THE CLEAN AIR ACT 1, https://www.epa.gov/sites/production/files/2013-09/documents/111background.pdf (“Section 111 gives EPA significant discretion to identify the facilities within a source category that should be regulated. To define the affected facilities, EPA can use size thresholds for regulation and create subcategories based on source type, class or size. Emission limits also may be established either for equipment within a facility or for an entire facility.”). 40 C.F.R. § 60.1; 42 U.S.C. § 7411(a)(2).

70

Hannah J. Wiseman

pollutant to which a standard applies.”64 However, routine modification, repair and replacement (RMRR) of existing stationary sources do not count as modifications for NSPS purposes under a regulatory exemption added in 1971.65 Other exceptions also apply. For example, increasing the hours of operation of a source does not count as a modification.66 If a source makes replacements to the facility, and the fixed capital cost of the replacements “exceeds 50 percent of the fixed capital cost that would be required to construct a comparable entirely new facility,”67 this does not meet the definition of a replacement, and the source will be subject to NSPS. If a stationary source falls under the definition of a newly constructed or modified source, it must achieve an emission limit (a maximum emission rate or other type of limit68) that “reflects the degree of emission limitation achievable through the application of the best system of emission reduction [BSER] which . . . the [EPA] Administrator determines has been adequately demonstrated.”69 When EPA identifies the BSER for a particular category of source, such as a coal-fired power plant, it must consider “the cost of achieving such reduction and any non–air quality health and environmental impact and energy requirement.”70 Although NSPS only applies to new sources, there are other limits on specific types of existing sources, including those in nonattainment areas and existing power plants that emit greenhouse gases, as explained further later. Appendix 3.1 provides a decision tree for determining whether a source is subject to NSPS. new source review (nsr). The NSR portion of the CAA sets national technologybased emissions standards for any sources that are newly constructed after the effective date of the standards or that are modified, as defined by regulation, after the effective date. There are two different types of NSR; a new or modified source will be subject to both types of NSR regulation if it will emit a criteria pollutant for which the region in which the source is located is in attainment and another criteria pollutant for which the region is in nonattainment. New and modified stationary sources built in regions of the United States that have achieved or exceeded the NAAQS for a particular pollutant are subject to an NSR program called Prevention of Significant Deterioration (PSD), as introduced earlier. For example, if the new or modified Source A will increase emissions of particulate matter (PM) and the source is proposed to be constructed or modified in a region that is in attainment for PM, the source will be subject to a PSD standard for PM. New stationary sources built in 64 65

66 67 68 69 70

40 C.F.R. § 60.14(a). 40 C.F.R. C.F.R. § 60.14 (e)(1); Thomas O. McGarity, When Strong Enforcement Works Better Than Weak Regulation: The EPA/DOJ New Force Review Enforcement Initiative, 72 Md. L. Rev. 1204, 1211 (2013). 40 C.F.R. C.F.R. § 60.14(e)(3). 40 C.F.R. § 60.15(b). See supra note 34. 42 U.S.C. § 7411(a)(1). Id.

Stationary Sources, Movable Rules

71

nonattainment areas are subject to nonattainment NSR. If Source A will increase emissions of carbon monoxide and the region in which the source is proposed to be located is in nonattainment for carbon monoxide, the source will be subject to nonattainment NSR standards for carbon monoxide. The emissions limits that new stationary sources must achieve under PSD or nonattainment NSR are defined by the total amount of emissions reductions that these plants can achieve by installing different types of technologies or engaging in other practices such as using advanced fuels and thus the total annual amount of pollutants these plants are allowed to emit assuming that they will install these technologies or take similar emission reduction approaches. New plants under the PSD program must achieve the emissions limit determined by the Best Available Control Technology (BACT) that similar plants around the country have installed,71 which is defined as follows: [BACT] means an emission limitation based on the maximum degree of reduction of each pollutant . . . which the permitting authority, on a case-by-case basis, taking into account energy, environmental, and economic impacts and other costs, determines is achievable for such facility through application of production processes and available methods, systems, and techniques, including fuel cleaning, clean fuels, or treatment or innovative fuel combustion techniques for control of each such pollutant.72

New plants under nonattainment NSR must achieve the emissions limit established by the Lowest Achievable Emission Rate (LAER)73 that plants in similar source categories have achieved, which is defined as follows: [LAER] means for any source, that rate of emissions which reflects (A) the most stringent emission limitation which is contained in the implementation plan of any State for such class or category of source, unless the owner or operator of the proposed source demonstrates that such limitations are not achievable, or (B) the most stringent emission limitation which is achieved in practice by such class or category of source, whichever is more stringent.74

In order to be subject to PSD or nonattainment NSR, a stationary source must emit a particular type of air pollutant (a criteria pollutant), must be a “major” source (meaning that it must have the potential to emit a minimum amount of air pollution) and must be “new” (meaning that it must be newly constructed or modified, as defined by regulation, after a particular date established by statute or regulation). Thus any source that emits one of the six regulated criteria air pollutants – nitrogen oxide, sulfur oxides, ground-level ozone, particulate matter, or lead – may be subject 71 72 73 74

42 U.S.C. § 7475(a)(4). 42 U.S.C. § 7479(3). 42 U.S.C. § 7503(a)(1). 42 U.S.C. § 7501(3).

72

Hannah J. Wiseman

to PSD or nonattainment NSR. But the source also must be major and new to be subject to one or both of these programs. To establish whether the source is major and thus subject to PSD or nonattainment NSR, one must determine the amount of pollution that a newly constructed or modified source has the potential to emit in a year. The threshold for major source differs depending on the region in which the source is located. In regions that are in nonattainment for at least one criteria pollutant, any source that has the potential to emit 100 tons per year of a regulated NSR pollutant is major and is subject to nonattainment NSR, provided that it is considered to be a new or modified source. However, for regions that are in “serious,” “severe” or “extreme” nonattainment for a limited set of pollutants, the threshold for defining a stationary source as major is lower. For example, any sources that have the potential to emit 50 tons per year of carbon monoxide and are a located in a “serious” nonattainment area for ground-level ozone are considered major for nonattainment NSR purposes.75 For stationary sources in attainment areas, the threshold for major source is the potential to emit 100 tons per year of any air pollutant if the source is on a statutory list of some of the major types of polluters (such as power plants, iron and steel mills or phosphate rock processing plants, among others)76 or 250 tons per year of any air pollutant if the stationary source is not on this list.77 One wrinkle in determining whether a stationary source is major arises depending on whether the source will be newly built or is an existing source that is being modified. If an entity proposes to build a new plant that has the potential to emit a criteria pollutant above the relevant emissions threshold, it is clearly a major source. Additionally, a plant that was built before August 7, 1977, and thus is a major source but not a new source, becomes a major new source if the plant will be modified and will cause a significant net emissions increase. Finally, a plant that was already built before August 7, 1977 (and thus is an existing source) and that does not emit at the major threshold will only be considered a major source if a modification is made at the plant (such as replacing a boiler), and the modification itself is major.78 For a modification itself to be major, the modification must have the potential to cause the plant to increase emissions by the amount that causes any plant to be a major plant. In an attainment area, the modification must enable the plant to have the potential to emit 100 or 250 more tons per year of a criteria pollutant to push the source into the major category.79 In a nonattainment area, the modification of the existing nonmajor plant must be a modification that causes the plant to have the potential 75 76 77 78

79

40 C.F.R. § 51.165(a)(x)(D). 42 U.S.C. § 7479(1). Id. See, e.g., 40 C.F.R. § 51.165(a)(2)(i) (explaining that preconstruction review “shall apply to any new major stationary source or major modification that is major for the pollutant for which the area is designated nonattainment”). CAA § 169(1).

Stationary Sources, Movable Rules

73

to emit 100 tons or more per year of a regulated NSR pollutant or a lower threshold in some serious and severe nonattainment areas.80 Thus, in one example of a plant in a nonserious or severe nonattainment area, if an existing power plant in this nonattainment area currently emits 80 tons per year of sulfur dioxide (below the major threshold) and proposes a modification, and the modification will cause the plant to emit 21 more tons per year of SO2, thus causing total potential SO2 emissions to be 101 tons per year, this plant is not major and is thus not subject to nonattainment NSR. However, if the modification will cause the plant to emit 100 or more tons per year of SO2, this is a major modification and will cause the plant to be considered a major source. Even if a stationary source emits a criteria pollutant or pollutants and is determined to be a major source, it will not be subject to PSD or nonattainment NSR unless it also is new. To be new, it must be newly constructed or modified after August 7, 1977.81 Thus there are many stationary sources with annual pollutant emissions that far exceed the major threshold, but because they were built before 1977, and because they made very small, incremental modifications that do not result in significant net emissions increases, they are not considered new and thus avoid PSD and nonattainment NSR. As introduced in the previous paragraph, a source is only considered new if it meets one of the following criteria: the source is (1) newly built, (2) an existing major source (i.e., a source built prior to August 7, 1977, that emits the defined threshold of pollutants for a major source) and will cause a significant net emissions increase when it is modified,82 or (3) the plant is an existing source that is not a major plant but is modified and has the potential to cause a net emissions increase that itself is major. For existing stationary air pollution sources that are already major sources but are grandfathered and thus avoid PSD and nonattainment NSR, the only event that will trigger PSD or nonattainment NSR requirements is a modification to the plant that will cause the plant to potentially emit criteria pollutants in an amount considered to be a “significant” net emissions increase. Note that the significant net emissions increase amount required to make an existing grandfathered major source a new major source is different from the amount required to make an existing nonmajor source a major source. Existing nonmajor sources must experience modifications that would, independently viewed, cause the sources to be major sources in order to count as major. But existing major sources that have been grandfathered need only experience a significant net emissions increase in order to count as major. The required net emissions increase threshold to make these sources new sources depends on whether the sources are in an attainment or nonattainment area and the type of criteria pollutant emissions that will increase. In attainment areas, in order for a net emissions increase to be considered significant – thus causing an 80 81 82

CAA § 302(j). 42 U.S.C. § 7475(a). 40 C.F.R. § 52.41(a)(2)(iv)(a).

Hannah J. Wiseman

74

existing major source that is being modified to be considered a new source – the potential annual net emissions increase caused by the modification must be 100 tons per year of CO, 40 tons per year of NOx, 40 tons per year of SO2, etc.83 The significant net emissions increase thresholds for existing major sources in nonattainment areas are similar.84 These brightline thresholds help to ensure durable policy achievements of the CAA in that they ensure that pollutant reductions are maintained as new, large sources are constructed and operated. But their inflexibility, in that they do not allow entities implementing the CAA to place similarly stringent limits on grandfathered sources that can cause substantial pollution problems (with the exception of old sources in nonattainment areas), detracts from progress under the Act. Similarly, the lack of NSR limits on minor sources and states’ inability to focus on minor sources even though their SIPs must address these sources due to their investment of resources in complex NSR permitting for major sources cause some of the most dominant forms of pollution to remain largely unchecked. Appendix 3.2 provides decision trees for determining whether a source is subject to nonattainment NSR or PSD. other requirements for attainment and nonattainment areas. Beyond requiring stationary sources in attainment and nonattainment areas to meet specific technology-based standards, the CAA contains additional requirements that attempt to ensure that attainment areas will remain in attainment and that nonattainment areas will reach attainment within five years of being designated as nonattainment areas. These are expressly adaptive standards in that they attempt to ensure – ex ante – that as conditions change over time these conditions will not cause deterioration of air quality and that positive reductions in air pollutants will increase over time. For attainment areas, prior to being constructed or modified, new and modified major sources must demonstrate that the source will not cause emissions of a criteria pollutant to increase by more than a designated “maximum allowable increase” more than once annually, in addition to other requirements.85 In nonattainment areas, one of the most significant requirements in addition to technology-based standards is that any entity proposing to build a new major source in a nonattainment area or modify a major source must offset its added emissions to ensure that total emissions from new major stationary sources, existing stationary sources and nonmajor sources will be lower than they previously were. The emissions must be lowered to the point where they represent “reasonable further progress” toward the area reaching attainment for the criteria pollutant at issue.86

83 84 85 86

40 C.F.R. § 52.21(b)(23)(i). 40 C.F.R. § 51.165(a)(1)(x)(A). 42 U.S.C. § 7475(a)(3). 42 U.S.C. § 7503(a)(1)(A).

Stationary Sources, Movable Rules

75

The CAA also requires states to take additional actions relating to nonattainment areas. Specifically, states must write SIPs that ensure that existing stationary sources within the nonattainment area adopt, at minimum, reasonably Available Control Measures Technology (RACT) as soon as possible.87 RACT is defined as “the lowest emission limitation that a particular source is capable of meeting by the application of control technology that is reasonably available considering technological and economic feasibility.”88 States also must inventory all actual emissions from existing sources and identify additional pollutant emissions that will result from new and modified sources in the area.89 This helps to achieve durable CAA goals of achieving the NAAQS by requiring states to take a backward look at nonattainment problems previously created by sources and mandating that states address these problems. However, as noted earlier, states have had difficulty meaningfully regulating old, grandfathered sources due to political opposition.90 3.2.3.2 Hazardous Air Pollutants (HAPs) The CAA regulates HAPs differently from criteria pollutants because it does not set an acceptable level of these pollutants in the air and then require states to regulate the pollutants to achieve this acceptable level. Rather, as introduced earlier, Congress in 1970 directed EPA to list HAPs and issue health-based standards for those pollutants.91 These standards were supposed to represent the maximum permissible emissions of a listed hazardous air pollutant from a stationary source within a particular source category92 and were to be written to “provide an ample margin of safety to protect the public health from such hazardous air pollutant.”93 The standards were to apply to new and modified stationary sources.94 Unlike SIPs for criteria pollutants, Congress allowed but did not require states to implement these national standards, providing that states could “develop and submit” to EPA a procedure for “implementing and enforcing the standards.”95 By 1977 it was apparent that health-based standards proved difficult to write and enforce. Congress therefore amended the CAA to provide that if in the judgment of the EPA administrator the health-based standard was “not feasible to prescribe or enforce” due to technical limitations on measurement of the pollutant or the lack of a conveyance that emits or captures the HAP, the administrator could promulgate 87 88 89 90 91 92 93 94 95

42 U.S.C. § 7502(c)(1). 44 Fed. Reg. 53761 (Sept. 17, 1979). 42 U.S.C. § 7502(c)(3)–(4). See supra text accompanying note 60. Pub. L. No. 91-604 § 112(b)(1)(A)–(B) (1970). Id. at § 112(b)(1)(C). Id. at § 112(b)(1)(B). Id. at § 112(c)(1)(A). Id. at § 112(d)(1).

76

Hannah J. Wiseman

technological standards.96 Specifically, these standards could be based on the design of the stationary source, particular equipment (technology) used at the source, work practices, operational practices or a combination of these factors.97 The 1977 CAA amendments also provided that state governors could apply to EPA to add a pollutant to the list; if the application showed that the pollutant caused mortality or “an increase in serious irreversible, or incapacitating reversible, illness,” EPA was required to add the pollutant to the list of HAPs.98 This required, at least in theory, periodic updating of standards to address new risks identified and enlisted states in this adaptive effort. But the amendments allowed too much discretion and flexibility with respect to implementation. Although they were intended to allow the CAA to adapt to new information, EPA – in part due to the difficulty of properly understanding and documenting the risks and acceptable levels of pollutants, as well as limited resources – failed to meet its mandate.99 Controlling HAPs from stationary sources continued to be challenging, and EPA was slow to list HAPs and set standards for them. In the 1990 CAA amendments, Congress therefore established the list itself100 and, through an important adaptive measure, required EPA to periodically review the list and, where appropriate, revise it.101 Pursuant to the 1990 amendments, EPA must consider revisions to the list at least once every eight years.102 Congress also largely jettisoned the health-based standards approach, directing EPA to identify categories of sources of HAPs103 and establish technology-based emission standards for those sources.104 Congress established different standards for new and modified sources of HAPs, but existing sources of HAPs throughout the country had to comply with a congressionally established technology-based emission standard. This contrasts with the criteria pollutant approach, in which Congress established national emission standards for existing sources only in nonattainment areas. For new, modified and existing sources of HAPs, Congress in the 1990 amendments required the EPA administrator to establish standards that reflected the “maximum degree of reduction in emissions of the hazardous air pollutants” subject to the Act that the administrator, “taking into consideration the cost of achieving such emission reduction, and any non–air quality health and environmental impacts and energy requirements, determines is achievable” for a particular category or subcategory of pollutants.105 This technology-based standard is called 96 97 98 99 100 101 102 103 104 105

Pub. L. No. 95-95 § 110(e)(1)–(2) (1977). Id. at § 110(e)(1). Id. at § 109. See infra note 168 and accompanying text. § 112(b)(1). Id. at § 112(b)(2). Id. at § 112(c). Id. Id. at § 112(d). Id. at § 112(d)(2).

Stationary Sources, Movable Rules

77

Maximum Achievable Control Technology (MACT).106 Under the MACT program, new and modified sources had to achieve the “emission control that is achieved in practice by the best controlled similar source, as determined by the Administrator,” whereas existing sources had to meet “the average emission limitation achieved by the best performing 12 percent of the existing sources” (excluding sources that had only recently achieved a particular limitation).107 New sources under the HAPs program are any stationary sources for which construction commences after a regulation containing an emission standard is first proposed,108 and modified sources are those at which a physical or operational change increases actual HAP emissions by more than a de minimis amount or causes the source to emit more than a de minimis amount of a HAP that the source did not previously emit.109 An existing source of HAPs is “any stationary source other than a new source.”110 In contrast to its approach to criteria pollutants, in which Congress mandated new source standards only for major sources, Congress provided that these technology standards were to apply to new and modified major and minor (“area”) sources for HAPs,111 although the EPA administrator could apply a less stringent standard to area sources. Major source MACT had to be based on measures that reduced the volume of or eliminated emissions from the sources; collected, captured or treated pollutants; or were “design, equipment, work practice, or operational standards,”112 whereas area source measures could be based on “generally available control technologies or management practices.”113 All MACT standards were to take into account “the cost of achieving” the emission reduction and “any non–air quality health and environmental impacts and energy requirements.”114 A major source under the HAP program is defined as follows: Any stationary source or group of stationary sources located within a contiguous area and under common control that emits or has the potential to emit considering controls, in the aggregate, 10 tons per year or more of any hazardous air pollutant or 25 tons per year or more of any combination of hazardous air pollutants.115

When writing these technology-based standards, Congress provided that the EPA administrator could take into account any health-based standards that already had been developed for HAPs.116 The administrator then had to prepare studies 106 107 108 109 110 111 112 113 114 115 116

42 U.S.C. § 7412(g)(2)(A). Pub. L. No. 95-95, at § 112(d)(3). 42 U.S.C. § 7412(a)(4). 42 U.S.C. § 7412(a)(5). 42 U.S.C. § 7412(a)(10). Pub. L. No. 95-95, at § 112(d)(1). Id. at § 112(d)(2). Id. at § 112(d)(5). Id. at § 112(d)(2). 42 U.S.C. § 7412(a)(1). Pub. L. No. 95-95, at § 112(d)(4).

78

Hannah J. Wiseman

examining the health risks remaining after the application of technology-based standards and make recommendations to Congress on methods of addressing these additional risks; if Congress did not act, the administrator was directed to promulgate additional standards necessary to “provide an ample margin of safety to protect public health.”117 This residual risk provision further provides [i]f standards promulgated pursuant to subsection (d) of this section [technologybased standards] and applicable to a category or subcategory of sources emitting a pollutant (or pollutants) classified as a known, probable or possible human carcinogen do not reduce lifetime excess cancer risks to the individual most exposed to emissions from a source in the category or subcategory to less than one in one million, the Administrator shall promulgate standards under this subsection for such source category.118

As summarized by a law review note addressing this provision, EPA has interpreted the residual risk provision to mean that HAPS with lifetime cancer risks below 1 in 1 million do not require a residual risk assessment because these levels are presumptively safe, and thus an ample margin of safety is preserved.119 For HAPs with a greater than 1 in 1 million cancer risk, EPA must promulgate health-based standards.120 And for HAPs that are not clearly in one of these two categories, EPA has discretion but conducts an individualized rulemaking for each pollutant to determine the appropriate risk level.121 Despite some relatively strict directives to EPA under the HAPs program, EPA has struggled to implement the HAPs. For example, the Agency took years to issue an effective rule to control mercury from power plants and has similarly delayed other HAPs rules.122 These ongoing difficulties represent the challenges that a regulatory agency faces while attempting to achieve durable goals and ensure adaptation as new information about risks emerges. Durability and effective adaptation require the agency implementing a statute to have adequate resources – particularly when the agency is tasked with relatively strict, timeconsuming work – and EPA has a slim budget and staffing in light of its duties under various federal environmental acts.

117 118 119

120 121 122

Id. at § 112(f). 42 U.S.C. § 112(f)(2)(A). Alex Jackson, EPA’s Fuzzy Bright Line Approach to Residual Risk, 36 ECOL. L.Q. 439, 457 (2009) (citing to National Emission Standards for Hazardous Air Pollutants; Benzene Emissions from Maleic Anhydride Plants, Ethylbenzene/Styrene Plants, Benzene Storage Vessels, Benzene Equipment Leaks, and Coke By-Product Recovery Plants (Benzene NESHAP), 54 Fed. Reg. 38,044 (Sept. 14, 1989). Id. at 457–58. Id. See Sierra Club v. Johnson, 444 F. Supp. 2d 46, 47, 51 (D. D.C. 2006) (describing some of EPA’s failures to issue required HAPs standards); Brett S. Dugan, Sierra Club v. Jackson, The Failed Potential of Judicial Review to Spur Timely Clean Air Act Regulations, 38 VT. L. REV. 1045 (2014) (describing similar failures documented in this case and others).

Stationary Sources, Movable Rules

79

3.2.3.3 “Other” Pollutants The CAA directs EPA to regulate additional air pollutants that are not designated as criteria pollutants or HAPs. Specifically, section 111(d) of the 1970 amendments provided the following: The Administrator shall prescribe regulations which shall establish a procedure similar to that provided by section 110 [state implementation plans] under which each State shall submit to the Administrator a plan which (A) establishes emission standards for any existing source for any air pollutant (i) for which air quality criteria have not been issued or which is not included on a list under section 108(a) [criteria pollutants] or 112(b)(1)(A) [HAPs] but (ii) to which a standard of performance under subsection (b) [new source performance standard] would apply if such existing source were a new source, and (B) provides for the implementation and enforcement of such emission standards.123

Thus Congress provided that for existing sources of pollutants that were not criteria or hazardous pollutants, the EPA administrator must establish emission standards for these sources and a procedure by which states are to meet the standards if they would be considered eligible for NSPS but for their status as old sources. Sources are eligible for NSPS (or section 111(d) emission standards) if they “may contribute significantly to air pollution which causes or contributes to the endangerment of public health or welfare.”124 Thus the standard applies to both major and minor sources. An existing source under the NSPS section of the CAA is “any stationary source other than a new source”125 and, as defined by associated regulations, a source for which construction was commenced before the date on which a standard was proposed.126 This means that under section 111(d), EPA must establish emission guidelines for sources that may contribute significantly to air pollution that causes or contributes to the endangerment of public health and welfare and that were built prior to the proposal of emission guidelines for these sources. In 1990, Congress directed EPA to write emission guidelines for municipal solid waste incinerators, including standards under section 111(d) of the CAA for existing incinerators.127 These guidelines require states – through a special state plan128 – to regulate129 numerous emissions from incinerators, including particulate matter, 123 124 125 126 127 128

129

Pub. L. No. 91-604, at § 111(d)(1). Id. at § 111(b)(1)(A). 42 U.S.C. § 7411(a)(6). 40 C.F.R. § 60.2. Pub. L. No. 101-549, at § 129(a). U.S. EPA, MUNICIPAL WASTE COMBUSTION: SUMMARY OF THE REQUIREMENTS FOR SECTION 111(D)/129 STATE PLANS FOR IMPLEMENTING THE MUNICIPAL WASTE COMBUSTOR EMISSION GUIDELINES viii (1996), https://nepis.epa.gov/Exe/ZyPDF.cgi?Dockey=2000FA6U.pdf. As opposed to NSPS, emission guidelines provide “guidance for regulating landfill gas emissions which the States are required to implement through individual State plans. While State plans must generally be as stringent as the EG, there is flexibility, on a case-by-case basis, to apply less stringent limitations or compliance schedules if certain criteria are met.”). Municipal Solid Waste Landfill

80

Hannah J. Wiseman

sulfur dioxide, hydrogen chloride, NOx, carbon monoxide, lead, cadmium, mercury and dioxins/furans.130 This plan is separate from the state’s SIP.131 More recently, EPA used section 111(d) to write emission guidelines for greenhouse gas (GHG) emissions from power plants that were under construction as of January 8, 2014.132 This section 111(d) regulation accompanied a second NSPS for new sources. Specifically, under section 111(b) of the CAA, EPA established NSPS for new power plants that emit carbon dioxide (plants that emit specified quantities of this gas and that are newly built or modified after the proposal of the standard), and under section 111(d), EPA established a GHG NSPS for existing power plants – those built or modified prior to the proposal of the standard. This latter regulation is called the Clean Power Plan (CPP). The Trump administration has proposed to withdraw this rule and has proposed a weaker replacement. Although the CPP might ultimately be significantly modified or wholly replaced, section 111(d) of the CAA as a whole, with its broad wording, provides an important example of how a statute can allow for adaptation over time. Indeed, the wording allows EPA to regulate pollutants that simply do not fall under other categories under the Act but still pose pollution problems that are important to address. NSPS also contributes to the durability of progress made under the CAA in that once a NSPS is finalized, it limits states’ discretion and requires minimal regulation of new and modified sources covered by the NSPS. In addition to NSPS for existing sources of air pollutants, the CAA also contains requirements for states to write SIPs, for power plants to control SO2 and NOx emissions that contribute to acid rain and for states to avoid approving stationary sources that contribute to other states’ air quality problems, as discussed in Chapters 2 and 6. In summary, the CAA applies to numerous types, ages and sizes of stationary sources, with the most stringent regulations addressing new and modified sources that emit HAPs or criteria pollutants in nonattainment areas. Table 3.1 summarizes the major portions of the CAA that apply to these sources. 3.2.4 The States’ Role in Stationary Source Regulation States play a central role in interpreting and implementing the stationary source standards described in Section 3.2. For example, although Congress directly imposed national emission standards on stationary sources through the NSPS and NSR programs, states are required to define and implement these standards through

130

131 132

New Source Performance Standards (NSPS) and Emission Guidelines (EG): Questions and Answers 2, https://www.epa.gov/sites/production/files/2016-09/documents/msw_landfqa.pdf. See, e.g., LOUISIANA DEPT. OF ENVTL. QUALITY, CLEAN AIR ACT SECTION (CAA) 111(d) PLAN FOR OTHER SOLID WASTE INCINERATION (OSWI) UNITS (summarizing the guidelines). See, e.g., id. 80 Fed. Red. 64,662, 64,715–16, supra note 53 and accompanying text.

(continued)

Non–air quality health and environmental impacts and energy requirements

BSER Any new source that “causes, or contributes significantly to, air pollution which may reasonably be anticipated to endanger public health or welfare” and emits pollutants that are not criteria pollutants or HAPs

Existing: Construction commenced before is standard proposed

111(d)

Major: Potential to emit at least 100 tons per year (tpy) of any air pollutant

Non–air quality health and environmental impacts and energy requirements

Considerations for emission standard

BSER Any new source that “causes, or contributes significantly to, air pollution which may reasonably be anticipated to endanger public health or welfare”

Emission standard

Major: and minor New and modified New: Construction commenced at time regulation with NSPS is proposed Modified: Source at which physical or operational change increases hourly emission rate

Type of source

111(b)

Size of source

Age of source

CAA section

table 3.1 Major CAA Provisions for Stationary Sources, Excluding SIPs and the Acid Rain Program

133

42 U.S.C. § 7412(a)(10).

Sources that emit listed Major and area (minor) Existing, new, and modified HAPs Major: Stationary source or New: Construction contiguous sources under commenced after MACT common control that first proposed emit 10 tpy of one HAP or Modified: Change at source 25 tpy of combination of increases HAP emissions by HAPs de minimis amount or causes source to emit more Area: All other sources of HAPs than a de minimis amount of a HAP that it previously did not emit Existing: “[A]ny stationary source other than a new source” (133)

Type of source

112

Size of source

Age of source

CAA section

table 3.1 (continued) Considerations for emission standard

Cost of achieving MACT emission New and modified: reduction, any Emission control non–air quality achieved in health and practice by the best environmental controlled similar impacts and source energy Existing: Emission requirements control achieved by best performing 12 percent of existing sources

Emission standard

135

134

(continued)

Major: Potential to emit 100 Sources that emit any BACT determined on “Energy, New and modified regulated NSR tpy or more of any air environmental, case-by-case basis: New: Sources commencing pollutant (a pollutant pollutant for specific and economic “fuel cleaning, construction after August 7, regulated anywhere in listed sources; for impacts and clean fuels or 1977 the CAA, provided the nonlisted source, other costs” treatment or Modified: Source with physical regulation “requires potential to emit 250 tpy innovative fuel or operational change that actual control of the or more of any air combustion causes significant net emisquantity of emissions pollutant techniques” (135) sions increase 134 of that pollutant” [ ])

40 C.F.R. § 51.166(b)(49). 42 U.S.C. § 7479(3).

160 et seq. (NSR – PSD)

40 C.F.R. § 51.165(a)(1)(xxxvii).

New and modified New: Source commencing construction after August 7, 1977 Modified: Sources with physical or operational change that causes significant net emissions increase

171 et seq. (nonattainment NSR)

136

Age of source

CAA section

table 3.1 (continued)

Type of source

Emission standard

Considerations for emission standard

No additional Major: Potential to emit 100 Sources that emit any New and modified: considerations regulated NSR tpy or more of any air Lowest achievable for LAER; pollutant (NOx and pollutant (or lower emission rate LAER: RACT is specific threshold in severe, most VOCs, any criteria to each source extreme, etc. stringent limitation pollutant with a and is based on nonattainment regions) for a category of NAAQS, precursors to technological sources in any state’s NOx and VOCs [136]) and economic SIP, or feasibility actually achieved by a source within the category Existing: Reasonably available control measures (technologies) (RACT)

Size of source

Stationary Sources, Movable Rules

85

their SIPs. Specifically, although EPA writes NSPS for a variety of sources, the CAA provides that [e]ach State may develop and submit to the Administrator a procedure for implementing and enforcing standards of performance for new sources located in such State. If the Administrator finds the State procedure is adequate, he shall delegate to such State any authority he has under this chapter to implement and enforce such standards.137

States that choose to implement NSR also must write plans for attainment and nonattainment, and these plans must contain a permitting program to implement NSR for all major new and modified stationary sources that will be built in these areas.138 Specifically, no new or modified major sources may be built until the state issues a permit that, among other things, sets the specific BACT (for attainment areas) or LAER (for nonattainment areas) for the source. The BACT or LAER within the individual permit typically includes a specific technology that the source must install or design, operational changes or other changes required at the source. All BACT and LAER requirements are inputted into a national clearinghouse, thus allowing for the potential diffusion of states’ approaches across state lines. When EPA updates technology-based requirements or guidelines for states’ BACT and LAER requirements, it also looks to this database, thus drawing from state approaches. As a result, standards tend to be gradually ratcheted up over time. As part of their responsibilities for moving nonattainment regions into attainment, states must approve the emissions offsets that new and modified sources must achieve – overall reductions of pollutants in the region – in order to be built. In attainment areas, states must ensure that sources do not exceed maximum permissible annual emissions increases or emissions concentrations and do not cause violations of NAAQS and other standards. Both the technology standards and the clearinghouse encourage the adaptation of pollution standards over time as new and improved technologies emerge. The state process for issuing construction permits to new sources under the PSD program demonstrates the general approach to selecting NSR technologies for each construction permit. Permit applicants must initially propose BACT emission limits and technological means of achieving those limits.139 When a state agency reviews this proposal and issues a BACT determination in each permit, EPA guidance from 1990140 directs the state to follow a top-down process. The state must first identify all available control options for the specific source being permitted – those with 137 138

139 140

42 U.S.C. § 7411(c). 42 U.S.C § 7502 (c)(5) (required permitting program for nonattainment areas); 42 U.S.C. § 7475 (required permitting program for attainment areas). In re: Masonite Corporation Permittee, 5 E.A.D. 551 at *4 (EPA 1994). In re: Mississippi Lime Co., 15 E.A.D. 349 at *6 (E.P.A. 2011) (“The NSR Manual guides permit issuers reviewing new sources under the CAA and sets forth a ‘top-down’ process for determining BACT for a particular regulated pollutant.”); In re Rockgen Energy Center at *4, *4 n. 10, 8 E.A.D.

86

Hannah J. Wiseman

“practical potential for application to the emissions unit and the regulated pollutant under evaluation,” including technologies “employed outside of the United States” and LAER technologies.141 The state must then “eliminate technically infeasible options,” “rank remaining control technologies by control effectiveness,” “evaluate the most effective controls” (choosing the most stringent control option or explaining why the most stringent option has been rejected due to “relative cost effectiveness of the alternative technologies”142) and select BACT – the “[m]ost effective option not rejected.”143 In the second step, in which the state rejects technically infeasible options, any options that are “available” and “applicable” – even undemonstrated control technologies – may not be rejected.144 And available is broadly defined by EPA guidelines as a technology that has “reached the licensing of the commercial sales stage of development.”145 Thus, when it is merely inconvenient for a source seeking a PSD permit to install a particular technology or fuel input that is the most stringent control – for example, when using natural gas rather than oil in starting up a plant because the plant is far from the nearest gas pipeline – the EPA Environmental Appeals Board that reviews NSR determinations has determined that this does not count as technical infeasibility.146 The clearinghouse substantially affects individual state decisions because when states identify all available control options, they must consider all technologies for that particular type of plant that are in the database. Because the database contains technologies required in permits in all fifty states, it shows the panoply of available control options and thus prevents states from excluding highly effective pollution control options that happen to have been selected as BACT for a similar source in another state. This is especially true because, as noted earlier, states issuing BACT determinations must, when ranking potential technologies, include the highly “stringent” LAER technologies. States might then have difficulty proving in later steps of the ranking process that the LAER is technically infeasible or not costeffective. Although the clearinghouse does not establish a default technology that states must choose, states or EPA often use it to defend their permitting decisions. For example, in a case where a regional EPA office selected as BACT an emission control efficiency that was several percentage points lower than a potentially higherefficiency control, the EPA appeals court – upholding the PSD BACT decision –

141

142 143 144 145 146

536 (E.P.A. 1999) (noting that “[a]though the NSR Manual is not a binding rule, we [EPA] have looked to it as a statement of the Agency’s thinking on certain PSD issues” and observing that “[i]n making BACT determinations, permit issuers frequently rely on a guidance document issued by the Agency in 1990.”). U.S. EPA, NEW SOURCE REVIEW WORKSHOP MANUAL B.5 (Draft, Oct. 1990), https://www.epa.gov/sites/ production/files/2015-07/documents/1990wman.pdf. In re Mississippi Lime Co., supra note 140, at *7. U.S. EPA, WORKSHOP MANUAL, supra note 141, at B.5–B.6. In re Mississippi Lime Co., supra note 140, at *9. U.S. EPA, WORKSHOP MANUAL, supra note 141, at B.18. In re Mississippi Lime Co., supra note 140, at *10.

Stationary Sources, Movable Rules

87

noted the regional office’s evidence that the efficiency selected “rank[ed] among the top of the list of control technologies for control efficiency,” as shown by the clearinghouse.147 In another important regulatory role for stationary sources, states are the entities primarily responsible for regulating existing sources of criteria pollutants in nonattainment areas, specifically by setting RACT standards for those sources; states must include RACT provisions within their SIPs.148 Thus, while EPA writes guidelines for RACT to help states determine the best potential technologies and practices, the states ultimately determine how to define the RACT for a particular source.149 When EPA writes RACT guidelines, it often looks to state and local programs that already have been implemented for existing sources. For example, when drafting RACT guidelines for reducing the emissions of volatile organic compounds (VOCs) from oil and gas well sites, compressor stations and other parts of the oil and gas industry, EPA looked to existing standards from Colorado, Montana, Wyoming and California’s Bay Area Air Quality Management District (AQMD) as well as general permitting requirements for VOC emissions from the oil and gas industry in West Virginia, Ohio and Pennsylvania.150 These standards existed prior to the establishment of federal RACT standards because states like Colorado already had implemented stringent controls on oil and gas VOCs under a special SIP in an effort to prevent certain parts of the state from being designated as nonattainment.151 Under an Early Action Compact between Colorado and EPA, the state issued this SIP in an effort to delay a nonattainment designation for the Denver area.152 The clearinghouse also influenced EPA’s oil and gas RACT guidelines because in writing the guidelines EPA looked to the clearinghouse into which all state requirements for BACT, LAER and RACT are inputted. This type of learning, in which EPA gleans lessons from states and local governments, implements these lessons into national guidelines and then influences other states through issuing these guidelines, is similar to the productive “iterative federalism” that occurs for mobile source standards, in which California often pushes for more stringent air quality standards, some of which are then incorporated at the federal level.153 This learning and updating process is adaptive, helping to ensure that the CAA does not remain stagnant and that further progress in reducing air pollution is achieved over time. 147 148 149

150 151

152 153

Id. at *7. 42 U.S.C. § 7502(c)(1). See, e.g., U.S. EPA, CONTROL TECHNIQUES GUIDELINES FOR THE OIL AND NATURAL GAS INDUSTRY (Draft) at 1-1 (Aug. 2015). Id. at 2-2 to 2-3. See State of Colorado Comments, Docket ID EPA-HQ-OAR-2008-0699; FRL 9918-43-OAR at 1-2 ( M a r . 1 7 , 2 0 1 5 ) , h t t p s : / / w w w . c o l o r a d o . g o v / p a c i fi c / s i t e s / d e f a u l t / fi l e s / A P - P O ColoradoCommentsOzoneNAAQS.pdf. Id. at 1. See Ann Carlson, Iterative Federalism and Climate Change, 103 NORTHWESTERN U. L. REV. (2009). See also Chapter 4 of this book, in which Professor Barry G. Rabe discusses leveraged federalism.

Hannah J. Wiseman

88

3.2.5 Durability of Stationary Source Standards The stationary source requirements described earlier are, on the whole, durable in that they represent a genuine attempt by Congress to ensure that air quality would remain at levels that promote human health and protect public welfare. This durability is demonstrated in several ways. First, although standards for the control of air pollution from stationary sources are technology-based standards, these standards aim to achieve health-based goals. In the case of criteria pollutants, by directing EPA to identify pollutants with adverse effects on human health and public welfare and to set acceptable ambient levels of these pollutants, Congress attempted to ensure that air quality would in fact remain at a safe level. All the criteria pollutant statutes and regulations described earlier are designed to ensure that as the number of new individual stationary sources of air pollutants increase – thus adding more pollution to the air – pollution limits will not be exceeded. And Congress gave EPA broad latitude in defining criteria pollutants with adverse effects by describing those pollutants as those which, “in the judgment of the Administrator,” create these effects.154 Congress also attempted to ensure that HAPs would not reach a level unsafe to human health by directing EPA to write health-based standards. Although Congress ultimately abandoned this effort, replacing NESHAPS with MACT, EPA is still required to review the adequacy of the technologybased standards, identify health risks that remain, and make recommendations to Congress regarding additional controls that are needed. And similar to criteria pollutants, Congress gave EPA broad latitude to define HAPs, initially defining these pollutants in 1970 as those, which, “in the judgment of the Administrator may cause, or contribute to, an increase in mortality or an increase in serious irreversible, or incapacitating reversible, illness.”155 When EPA was too slow to list these pollutants, Congress provided a statutory list of pollutants deemed to cause these problems. Beyond incorporating health-based goals into technological standards, the CAA’s approach to stationary sources is also durable in that it attempts to preserve safe levels of air quality once they are achieved and avoid slippage. For example, in addition to meeting technology-based standards, sources in nonattainment areas must prove that they will offset the new air emissions that they will create and that total overall emissions in the region will be lower after the new source is built. Thus, even though the new source is not the cause of the existing nonattainment problem, in order to build in an area with already dirty air, the source must find ways to reduce some of the existing air quality problems caused by other sources. States that apply to EPA to redesignate a nonattainment area as being in attainment also must submit a maintenance plan showing how the state will maintain the NAAQS “for at least 10 years 154 155

Pub. L. No. 91-604, at § 108. Id., at §112(a).

Stationary Sources, Movable Rules

89

after the redesignation.”156 This is an important forward-looking requirement that helps to prevent slippage and is a strong example of durability within the CAA. Further, in attainment areas, in addition to meeting technology-based standards, new sources must show that they will not increase emissions above a specific maximum amount or concentration or cause any NAAQS violations. Another indicator of the durability of the stationary source standards is the extent to which they aim to achieve and maintain progress by reducing a relatively comprehensive set of air pollutants and, to some extent, numerous sources of these pollutants. Covering this range of pollutants helps to prevent backsliding on air quality overall. Here the CAA is more durable in some areas than others. The CAA is most comprehensive with respect to HAPs, in that it regulates both major and area (small) sources and new and existing sources of air pollutants in meaningful ways. This is because HAPs have localized effects even when emitted in relatively small quantities. For criteria pollutants, the CAA addresses minor sources more marginally. It requires states to inventory all sources that contribute to pollutants in a nonattainment area (including minor sources).157 Additionally, for nonattainment areas with particularly problematic air quality, Congress requires states to reduce emissions from a limited number of smaller sources. For example, in areas with extreme ground-level ozone problems, each new or modified electric utility or industrial or commercial boiler that emits “more than 25 tons per year of oxides of nitrogen” must burn natural gas, methanol or ethanol or a fuel with comparably low emissions, or it must “use advanced control technology.”158 Further, in some cases, EPA has imposed controls on minor sources of criteria pollutants through the NSPS program. For example, EPA has established NSPS for petroleum storage vessels,159 surface coating of metal furniture160 (with the exception of smaller facilities that use less than 3,842 liters of coating annually161) and other coating operations, residential woodstoves,162 oil and gas wells and some components of oil and gas distribution and pipeline systems163 and leaks of VOCs from equipment and tanks at various types of facilities,164 among other smaller sources. Finally, the CAA requires states to review 156 157 158 159 160 161 162 163 164

42 U.S.C. § 7505(a). 42 U.S.C. § 7502(a)(3). 42 U.S.C. § 7511a(e)(3). 40 C.F.R. §§ 60.110a–60.116(a) (subpart KA). 40 C.F.R. §§ 60.310–60.316 (subpart EE). 40 C.F.R. § 60.310(c). 40 C.F.R. §§ 60.530–60.539(b) (subpart AAA). 40 C.F.R. §§ 60.5360–60.5430 (subpart OOOO). See, e.g., 40 C.F.R. §§ 60.480–60.489 (subpart VV – equipment leaks of VOCs in synthetic organic chemical manufacturing industry for plants constructed or modified in a specified time period); 40 C.F.R. §§ 60.480(a)–60.489(a) (subpart Vva – same, covering another time period); 40 C.F.R. §§ 60.590–60.594 (subpart Ggg – equipment leaks of VOCs in petroleum refineries for refineries constructed or modified in a specified time period); 40 C.F.R. §§ 60.590(a)–60.593(a) (subpart Ggga – same, covering another time period); 40 C.F.R. §§ 60.630–60.636 (subpart Kkk – equipment

90

Hannah J. Wiseman

the proposed construction of minor sources under their programs that implement NSR for attainment and nonattainment areas.165 States often regulate and limit emissions from minor sources via their minor source NSR programs.166 But the requirements for states’ minor source NSR programs are quite flexible.167 Beyond these types of limited minor source programs, minor sources of criteria pollutants remain largely unregulated. Finally, the CAA is both durable and adaptive in the sense that it covers pollutants not yet identified by Congress as subject to regulation; it also requires EPA to consider regulating a broad range of pollutants by broadly defining the pollutants covered. Section 111(d) of the CAA, which addresses noncriteria and non-HAP pollutants, requires EPA to regulate pollutants not included under other portions of the Act. Although EPA has only rarely used section 111(d), this is a precautionary and adaptive portion of the statute in that it recognizes that Congress might not have specifically identified all pollutants that could potentially have harmful effects and requires EPA to consider and address these other pollutants. Similarly, section 111 (b)’s broad definition of pollutants to be covered by NSPS – including each source that “in [the EPA administrator’s] judgment . . . causes, or contributes significantly to, air pollution which may reasonably be anticipated to endanger public health or welfare” – requires EPA to list a broad range of source categories and regulate them even if they are not criteria pollutants. This section, along with section 111(d), allowed EPA to regulate substances such as methane and other GHGs.168 Although in many respects the CAA as applied to stationary sources is highly durable, in application, its achievement and maintenance of certain standards have

165

166

167

168

leaks of VOCs in onshore natural gas processing plants for plants constructed or modified in a specified time period). 42 U.S.C. § 7410(a)(2)(c) (requiring state SIPs to include a program for “regulation of the modification and construction of any stationary source within the areas covered by the plan as necessary to assure that national ambient air quality standards are achieved”); Luminant Generation Co. v. EPA 922 (5th Cir. 2012) (noting that “the CAA prescribes only the barest of requirements for ‘minor’ NSR”); id. at 923 (“The EPA has recognized that because ‘the Act includes no specifics regarding the structure or functioning of minor NSR programs’ and because the implementing regulations are ‘very general . . . SIP-approved minor NSR programs can vary quite widely from State to State’” (quoting 74 Fed. Reg. 51,418, 51,421 (Oct. 6, 2009)). See, e.g., U.S. EPA, MINOR NSR BASIC INFORMATION, https://www.epa.gov/nsr/minor-nsr-basic-infor mation (“minor NSR permits often contain permit conditions to limit the sources [sic] emissions to avoid PSD or nonattainment NSR”). 40 C.F.R. § 51.160 (requiring SIPs to “set forth legally enforceable procedures that enable the State or local agency to determine whether the construction or modification of a facility, building, structure or installation, or combination of these” will cause NAAQS violations or interfere with the state’s control strategy and, in terms of the size of the source covered by this requirement, only requiring the state to “identify types and sizes of facilities, buildings, structures or installations which will be subject to review under this section”). 81 Fed. Reg. 35,824 (June 3, 2016), https://www.gpo.gov/fdsys/pkg/FR-2016-06-03/pdf/2016-11971.pdf (limits on methane from newly fractured oil and gas wells). Although methane is not a criteria pollutant, it is a GHG for which EPA has made an endangerment finding in the mobile source context, as EPA notes in its methane rulemaking. Id. at 35,828.

Stationary Sources, Movable Rules

91

been somewhat more limited. As demonstrated by the HAP experience, EPA was initially so slow to list any HAPs that Congress then stepped in and formed the list itself.169 This was a move to reduce flexibility previously allowed by Congress in its direction to the Agency to list HAPs; Congress simply selected the HAPs and listed them. This is an example in which flexibility allowed for too much shirking of regulatory duties. This slow action under a somewhat flexible scheme was for good reason – an immense amount of time and large investment of resources were required to gather the data required to write the health-based NESHAPs – but it nonetheless negatively delayed regulation of highly damaging pollutants. As a result of these and other challenges, writing pure health-based standards for HAPs proved to be so difficult that Congress switched primarily to a technology-based standard.

3.3 evidence of durability: air quality achievements from stationary source regulation The regulation of stationary sources has been highly effective for many pollutants. National emissions totals from stationary sources have, for the most part, declined dramatically, as demonstrated by Figure 3.1.170 In the area of HAPs, EPA reports that “[s]tationary sources today emit about 1.5 million tons less toxic air pollution per year than in 1990.”171 Further, “[m]ercury emissions fell by about 75 percent between 1990 and 2008” due largely to HAP regulation applied to “municipal waste combustion and medical waste incineration,”172 although reductions from coal-fired power plants were also significant.173 The primary remaining emitters of mercury are power plants.174 For the period 1990–2020, EPA estimates that CAA controls on all pollutants emitted from both stationary and mobile sources would yield benefits that exceeded costs “by a factor of more than 30 to 1” as a central estimate (calculated annually or using present value), and the high estimate shows benefits exceeding costs by approximately 90 to 1.175 The costs of achieving these benefits are estimated to be 169

170 171

172 173 174 175

See Patrick D. Traylor, Presumptive MACT As a Regulatory Tool to Streamline the Development of National Emission Standards for Hazardous Air Pollutants, 4 ENVTL. L. 393, 400 (1998) (noting that “EPA failed to list even one HAP within the original ninety day period” required by the 1970 amendments); Patricia Ross McCubbin, The Risk in Technology-Based Standards, 16 DUKE ENVTL. L. & POL’Y F. 1, 30 (2005) (noting that between 1970 and 1990, EPA “only identified eight hazardous air pollutants and adopted ‘ample margin of safety’ standards for a small fraction of the industries emitting those pollutants”). https://www.epa.gov/sites/production/files/2015-07/national_tier1_caps.xlsx. U.S. EPA, PROGRESS CLEANING THE AIR AND IMPROVING PEOPLE’S HEALTH, https://www.epa.gov/cleanair-act-overview/progress-cleaning-air-and-improving-peoples-health. Id. U.S. EPA, REPORT ON THE ENVIRONMENT, http://www.epa.gov/roe/. See infra note 179 and accompanying text. U.S. EPA, THE BENEFITS AND COSTS OF THE CLEAN AIR ACT FROM 1990 TO 2020, 7–8 (Apr. 2011), https:// www.epa.gov/sites/production/files/2015-07/documents/fullreport_rev_a.pdf.

Hannah J. Wiseman

92

35,000 30,000

Thousands of Tons

25,000 20,000 15,000 10,000 5,000 0 1970

1980

1990

2000

2010

2014

Stationary fuel combustion CO

Industrial and other processes CO

Stationary fuel combustion NOX

Industrial and other processes NOX

Stationary fuel combustion SO2

Industrial and other processes SO2

figure 3.1 Selected criteria pollutant emissions from selected stationary sources Source: U.S. EPA, NATIONAL EMISSIONS TOTALS, https://www.epa.gov/sites/production/ files/2015-07/national_tier1_caps.xlsx. Thanks to Kristen McCormack, Resources for the Future, for assistance with this figure.

lower, as a whole, for stationary sources, with a present value of $49 billion in costs of electric utilities and $43 billion for industrial sources compared with $220 billion for on-road vehicles and fuels, all in 2006 dollars.176 Total costs of CAA controls on all sources – stationary and mobile – are estimated to be $380 billion, whereas total benefits are estimated to amount to $12 trillion, again in 2006 dollars, present value.177 This is not to say, however, that all air quality controls have been implemented using least-cost methods for industry. As the National Research Council observed in 2004, standards such as MACT for hazardous air pollutants, which tend to have the effect of requiring the installation of specific technologies despite technically being more flexible, “are not set up to minimize costs across all companies.”178 Despite dramatic progress, stationary sources continue to be major contributors to air quality problems. For example, approximately 77 percent of acid gases, 60 percent of sulfur dioxide, and 62 percent of arsenic emissions come from power plants.179 Coal-fired power plants are the largest single source of US mercury 176 177 178 179

Id. Id. NATIONAL RESEARCH COUNCIL, AIR QUALITY MANAGEMENT, supra note 1, at 188. U.S. EPA, CLEANER POWER PLANTS, https://www.epa.gov/mats/cleaner-power-plants#controls.

Stationary Sources, Movable Rules

93

emissions180 and still cause numerous premature deaths and illnesses in the United States. According to EPA calculations, an Obama administration rule that would limit mercury emissions from these plants (and also particulate matter emissions) would annually avoid approximately 4,200 to 11,000 premature deaths that are currently caused by particulate matter emissions.181 Other impacts that would be avoided by this rule further demonstrate the magnitude of the societal costs imposed by coal-fired plants: EPA estimates that the rule, simply by reducing particulate matter emissions from these plants, in 2016 would avoid “4,700 nonfatal heart attacks, 2,600 hospitalizations for respiratory and cardiovascular diseases, [and] 540,000 lost work days,” among other impacts.182 Particulate matter emissions are a copollutant, meaning that when mercury emissions are reduced, this will also automatically reduce particulate matter due to the technologies and processes required to reduce mercury emissions; many of the estimated benefits arise from reduction of copollutants.183 Major stationary sources such as US electricity generation also continue to contribute to other air quality problems. These sources emitted 2,039.8 million metric tons of CO2 (as measured by CO2 equivalents) in 2013, forming the largest contribution to the total of 5,502 million metric tons of CO2 emitted in the United States in 2013.184 These sources have since been displaced by transportation, which is now the highest emitting sector with respect to CO2.185 Overall, the CAA has substantially reduced emission from stationary sources, but there is much room for progress – particularly from grandfathered and minor sources.

3.4 durability, adaptability and flexibility of the caa’s stationary source provisions Environmental controls must achieve an effective balance between providing relatively firm mandates to ensure durable progress on environmental quality while also requiring and/or allowing adaptation of these controls over time and flexibility for the agencies and states that implement those controls. Adaptability requires agencies that implement congressional directives to update and modify standards to address 180 181

182 183

184

185

Id. (noting that “power plants are currently the dominant emitters of mercury (50 percent)”). U.S. EPA, REGULATORY IMPACT ANALYSIS FOR THE FINAL MERCURY AND AIR TOXICS STANDARDS at ES-1 (Dec. 2011), https://www3.epa.gov/ttnecas1/regdata/RIAs/matsriafinal.pdf. Id. at ES-3. Id. at E1–5 (“In addition to reducing HAP, the emissions control technologies that will be installed on coal- and oil-fired EGUs [energy-generating units] to reduce HAP [hazardous air pollutants, including mercury] will also reduce sulfur dioxide (SO2) and particulate matter (PM)”). U.S. EPA, INVENTORY OF U.S. GREENHOUSE GAS EMISSIONS AND SINKS: 1990–2013, table 2-1, pp. 2–4 (2015), http://www3.epa.gov/climatechange/Downloads/ghgemissions/US-GHG-Inventory-2015Main-Text.pdf. ENERGY INFO. ADMIN., MONTHLY ENERGY REVIEW, ch. 12: ENVIRONMENT, https://www.eia.gov/totale nergy/data/monthly/pdf/sec12.pdf.

94

Hannah J. Wiseman

new risks and benefits – or to update standards as scientific understanding of existing risks and benefits improves. Further, flexibility, depending on how it is wielded by the agencies and states implementing the act, can help give these entities the leeway they need to address changing problems and can also achieve the benefits of environmental controls at lower costs to industry. Inflexible standards, in contrast – those that are brightline or “sharp edged,” as defined here – are clear, hard and fast numerical thresholds that define which sources are or are not regulated under the CAA. Both flexible and inflexible standards have in some respects contributed to beneficial durability in terms of achieving and maintaining CAA goals of cleaner air. Yet in some cases they also have allowed slippage away from those goals. 3.4.1 Adaptive, Flexible and Inflexible CAA Attributes 3.4.1.1 Adaptive Components With respect to adaptability, the features of the CAA that exhibit the most adaptability and that tend to improve air pollution control include, inter alia, the NAAQS, NSPS, NSR and citizen suit provisions. The NAAQS in particular forces continued progress toward air quality regardless of the number of sources emitting pollutants and the individual controls on those sources. It is itself a relatively inflexible rule (setting a strict threshold for the definition of attainment and nonattainment) and causes the entire CAA and air quality under the CAA to be more durable. Additional components of the NAAQS help to ensure adaptation and that air quality achievements are maintained. These include the requirement that EPA periodically consider updating the NAAQS based on new health-based understandings of air quality impacts, that new sources in nonattainment areas more than offset the emissions they will add to these areas and improve air quality and that states demonstrate how attainment will continue to be achieved for ten years after a nonattainment area has been redesignated as an attainment area. A key tool for adaptation under NSPS and NSR involves states’ selection of technologies for new source permitting on a case-by-case basis, including more stringent technologies than those required by federal rules. This, as well as the flexibility permitted in the writing of SIPs to achieve the NAAQS (allowing states to regulate more stringently above a federal floor), allows for and encourages state innovations. When states select technologies for a permit that are more stringent than those required by federal standards, these technologies become part of a clearinghouse, which serves as a substantial guide for other states writing permits. While states are not required to select the most stringent technology in the clearinghouse, they must provide meaningful justifications for not selecting this “top” technology. Thus, as explored earlier, these innovations in some cases become parts of federal guidelines or standards, as exhibited by the oil and gas VOC NSPS guidelines. In some cases, however, there has been extensive litigation addressing

Stationary Sources, Movable Rules

95

whether states’ choices within the SIPs are in fact as stringent as the federal “floor,”186 and it appears that this flexibility sometimes allows states to maintain these less stringent approaches, particularly because EPA lacks the resources to fully monitor and enforce state decisions. Citizen suit provisions within the CAA also help place a check on EPA and states over time even as administrations and priorities change. They contribute to the durability of progress toward clean air by allowing citizens to constantly monitor the Agency’s actions. Nonprofit groups and other interested entities may bring agencies to court to challenge their interpretation of statutory meanings, to require them to perform nondiscretionary duties under the CAA and to spur enforcement of violations of air quality standards that are inadequately enforced by the states.187 In this sense, just like the NAAQS, citizen suits make the CAA itself, and the progress it achieves, more durable by providing a relief valve when agencies and states fail to live up to their duties or lack the resources and time to do so. However, citizen suits have limits. EPA often repeatedly violates court orders and consent decrees that require the issuance of specific stationary source regulations by clear deadlines.188 The CAA also contains provisions that potentially hinder adaptation. Some brightline, relatively rigid standards within the Act have constrained adaptation by limiting EPA’s ability to address some growing air quality risks. This, in turn, creates lasting problems and limits the durability of air pollution reductions achieved under the Act. The hard and fast rule that “existing” sources be subject to less stringent standards and complex regulations with very specific numerical thresholds that allow for minor modifications to be made without triggering new source regulations have allowed old sources of air pollution to continue emitting substantial quantities of pollutants. Further, brightline numerical rules that address which sources count as a minor versus a major source allow many sources to operate just below the major source threshold and enjoy less stringent minor source regulations despite, in some cases, having major impacts, particularly when those sources are considered collectively. 3.4.1.2 Flexible Mechanisms Additional mechanisms within the CAA allow for flexibility in implementation of the Act, and this helps to contribute to adaptation of the Act over time. These include, inter alia, certain broadly worded definitions, EPA’s discretion in requiring state compliance with federal standards, state and EPA discretion with respect to 186

187 188

See, e.g., State of Texas et al. v. EPA (5th Cir. 2012) (striking down EPA’s disapproval of Texas’s flexible federal permit program for NSR). 42 U.S.C. § 7604. See, e.g., Dugan, Sierra Club v. Jackson, The Failed Potential of Judicial Review, supra note 122 (describing many missed deadlines).

96

Hannah J. Wiseman

individual source compliance and provisions for trading of pollutant reductions among sources. The portions of the CAA that are worded in more general terms – such as those broadly defining air pollutants – have given EPA the flexibility needed to effectively address some new risks and to make the Act adapt to new risks. Not all flexibility is beneficial, however, as shown by the failure of EPA to list HAPs and states’ ability to flexibly implement federal standards through their SIPs – sometimes in a manner that seems to violate the requirement that state standards be more stringent than federal standards. But mandates, too, can backfire. Requiring an underfunded and understaffed agency to write standards for listed HAPs does not fully solve the flexibility problem; more regulatory support is needed for the agency to fulfill its mandate. With respect to EPA discretion in requiring state compliance with federal standards, the Agency’s authority to grant states extensions in terms of meeting the NAAQS, and its use of this power, has allowed for certain needed adaptations. As explored in a case study (Section 3.4.2), in Colorado, an Early Action Compact (EAC) with EPA, which was designed to give Colorado more time to try to avoid having the region of the state including Denver listed as a nonattainment area, caused the state to implement relatively aggressive controls on VOC emissions at oil and gas sites. These controls were later incorporated into federal EPA guidance for a new NSPS on oil and gas VOC emissions. Although these controls did not ultimately avoid nonattainment status for the Denver area, they had other positive impacts due to their influence at the federal level and subsequent dissemination to other states. With respect to the discretion of EPA and states in requiring individual source compliance, sources are allowed to net their emissions, meaning that they can increase emissions at one portion of their plant and decrease them elsewhere at another portion of the plant to avoid exceeding the NSR threshold, among other potential effects of increased emissions. This offers important flexibility for industry in terms of allowing creative pollution control strategies to be used where they are most efficient and reducing industry compliance costs. However, netting and similar discretionary mechanisms can also allow sources to make continuous modifications while potentially avoiding more stringent controls that should apply to them given their contributions to air quality problems. As with any statute, discretion must have boundaries in order to avoid exceptions becoming the rule and stymying needed progress. A final flexible aspect of the stationary source portions of the CAA – a feature that has substantially helped achieve the Act’s durable standards – is the trading provisions discussed in detail in Chapter 6. The Acid Rain Program for power plants caused these plants to achieve SO2 reductions more quickly and cheaply than they otherwise would have and generated substantial progress in terms of improved air quality. The NOx Budget Trading Program was similarly effective.189 The following 189

U.S. EPA, 2014 PROGRAM PROGRESS: CLEAN AIR INTERSTATE RULE, ACID RAIN PROGRAM, AND FORMER NOX BUDGET TRADING PROGRAM, https://www3.epa.gov/airmarkets/progress/reports/index.html.

Stationary Sources, Movable Rules

97

case studies highlight some of these positive and negative attributes of flexibility and adaptability under the CAA. 3.4.2 The Benefits and Challenges of Strict, Nonadaptive Standards: Examples from the Oil and Gas Sector and Carbon Regulation Many of the CAA’s mandates are brightline hard-edged rules that help to create the durability of the Act. Yet some of these brightline rules also can block certain needed adaptation of the Act over time. With respect to the benefits of these rules, as discussed in Section 3.2, the CAA directs EPA to identify a specific NAAQS, which is an acceptable concentration of criteria pollutants in the air. This is the target designed to be achieved by a combination of national technology-based standards and state SIPs that both implement these standards and require additional controls. Establishing a NAAQS connects environmental quality goals directly with environmental standards, setting a clear target to be achieved in order to protect human health and the environment and ensuring that as the number of regulated sources grows, this target is still met. The NAAQS also forces certain, typically nonadaptive brightline rules (such as the definition of major and new source and the regulation of these sources) to change. When sources become numerous, individual emissions controls become inadequate for limiting the collective impacts of these sources; the CAA helps to solve this problem by requiring states to regulate certain existing and minor sources under their SIPs for achieving the NAAQS and, once an area fails to achieve the NAAQS, regulating smaller sources and existing sources. As noted earlier, the emissions threshold that a source must meet to be considered major is lower in certain types of nonattainment areas, such as those in severe nonattainment status. While brightline rules can lead to important environmental progress and clarity, they also pose challenges when not adequately balanced with flexible provisions for implementation, as demonstrated by the following two case studies. In the case of carbon regulation, EPA struggled to find a way to fit sources of GHGs within existing definitions of major sources. But the section 111(d) provision, which provides a “catch-all category” for sources not regulated under NSPS and HAPs but that should be subject to NSPS, allowed EPA to regulate those sources – albeit through a rule that might not endure for political reasons. In the case of oil and gas, certain sources with substantial pollution are not defined as major simply because they are slightly separated spatially, and they therefore avoid stringent NSR regulations. 3.4.2.1 Carbon Regulation under the CAA The regulation of carbon dioxide and other GHG emissions under the CAA is one of the clearest examples of the benefits and challenges of the CAA’s sometimes confusing combination of adaptive and nonadaptive standards with flexible and some

98

Hannah J. Wiseman

inflexible implementation provisions. The CAA’s broad definition of the pollutants to be regulated under the Act provides important adaptability in that it allows the act to remain relevant over time and enables EPA to update its regulations to address new risks. However, when EPA identifies a new pollutant to be regulated under the CAA, this can trigger a variety of nonadaptive rules, such as numerical and technological standards that were defined in the 1970s and have not since changed substantially. This sometimes limits EPA’s ability to regulate a pollutant in the most effective way possible, instead forcing it to conform to rigid categories and definitions within the CAA. In the case of carbon and other GHG emissions, the CAA’s broad, adaptive definition of pollutant has given EPA the flexibility needed to address the effects of a pollutant for which scientific understanding of risk has recently improved. Specifically, the US Supreme Court determined in 2007 that GHGs are pollutants, as defined by the CAA’s general provisions,190 in which an air pollutant is “any pollution191 agent or combination of such agents, including any physical, chemical, biological, radioactive . . . substance or matter which is emitted into or otherwise enters the ambient air.”192 Thus the CAA enables EPA to regulate emissions of GHGs from mobile and stationary sources. While EPA need not, and indeed has not, interpreted air pollutant identically within different portions of the Act, the Court since 2007 has noted that this definition is a “sweeping and capacious interpretation,” although it does not give the Agency free rein to regulate GHGs.193 With respect to stationary sources, EPA has controlled GHGs from new and modified sources through both the NSR PSD and NSPS programs, although it appears that some of these regulations will be largely withdrawn by the Trump administration. Despite the fact that some carbon regulations are likely to disappear due to politics, they provide an important example of the CAA’s ability to cover previously unregulated pollutants in an adaptive manner, as well as its limits when it comes to addressing new risks under the CAA. As introduced in Section 3.2, the NSR program has brightline rules that can pose problems from a regulatory perspective. For example, in attainment areas, sources are major only if they are a source listed by Congress that emits 100 tons annually of a regulated pollutant or if they are unlisted and emit 250 tons annually. In regulating GHGs under the NSR PSD program, EPA observed that defining sources of GHGs as major if they met these thresholds would be unworkable because it would subject numerous small sources – those not intended by Congress to be covered by the NSR program – to relatively stringent technology-based emission limitations.194 This 190 191 192 193 194

Mass. v. EPA, 549 U.S. 497, 528–29 (2007). 42 U.S.C. § 7479(1). 42 U.S.C. § 7602. Utility Air Reg. Group v. EPA, 134 S. Ct. 2427, 2441 (2014) (quoting Mass. v. EPA, 549 U.S. at 528). UARG, 134 S. Ct. 2427 at 2442–43 (describing the Tailoring Rule and EPA’s conclusion that the PSD thresholds would be unworkable for GHGs).

Stationary Sources, Movable Rules

99

would place an unreasonable permitting burden on the Agency and would not give it the flexibility to prioritize the truly large sources that were some of the primary contributors to the carbon problem. EPA therefore finalized a Tailoring Rule that substantially raised the threshold for defining a stationary source of GHGs as major for NSR purposes.195 In the first stage of this rule, EPA regulated sources that were subject to NSR anyway (so-called anyway sources) because they were newly built and emitted major amounts of another non-CO2 NSR-regulated pollutant in addition to CO2 or would be modified in a way that would cause them to be major for another non-CO2 NSR-regulated pollutants. For these anyway sources, only those that increased net CO2 emissions by at least 75,000 tons per year were regulated. The second stage of the Tailoring Rule regulated sources that were not considered major for non-CO2 NSR-regulated pollutants and that met thresholds of 100,000 or 75,000 tons per year of CO2 emitted.196 In 2014, the Supreme Court struck down the second stage of this rule, finding that the “Act, in no uncertain terms, requires permits for sources with the potential to emit more than 100 or 250 tons per year of a relevant pollutant” and that “[a]n agency has no power to ‘tailor’ legislation to bureaucratic policy goals by rewriting unambiguous statutory terms.”197 However, the Court upheld the first stage of the rule, allowing EPA to require the technology-based standard applicable under the PSD NSR program (BACT) to apply to GHGs for anyway sources.198 The Court made this determination based on a broad definition within the CAA that requires BACT “‘for each pollutant subject to regulation under this chapter’ (i.e., the entire Act),” as noted by the Court.199 GHGs are regulated under the CAA – including under the mobile source portions of the CAA – and therefore also may be regulated under BACT, according to the Court. The inability of EPA to set different thresholds for defining major sources of GHGs under the brightline definitions within the NSR program and the preservation of most portions of EPA’s rule by relying on other, broader definitions under the CAA demonstrate the contortions that are sometimes required to adapt as needed to fit pollutants that are clearly covered by the Act into the Act’s specific programs. While the combination of brightline rules and broader definitions can enable regulation of pollutants for which new risks have been more recently identified and quantified, it also can make regulation more complicated and difficult to promulgate. The benefits of the broad, relatively discretionary and adaptive nature of some of the CAA’s definitions, in contrast, are demonstrated by the Obama administration EPA’s use of NSPS to additionally regulate GHGs from new stationary sources. 195 196 197 198 199

75 Fed. Reg. 31,514 (2010). Id. at 31,516. UARG, 134 S. Ct. 2427 at 2444–45. Id. at 2437. Id. at 2448.

100

Hannah J. Wiseman

Under the NSPS program, Congress provides a broad directive for EPA to establish a list of stationary sources and standards for these sources and to include a particular source on the list if the source in the EPA administrator’s judgment “causes, or contributes significantly to, air pollution which may reasonably be anticipated to endanger public health or welfare.”200 Thus EPA – which had already listed and established standards for power plants under the NSPS program for other, nonGHG pollutants – used this portion of the Act to regulate GHGs from newly constructed and modified plants.201 The broadly worded language of the CAA also enabled the Obama administration EPA to use NSPS to regulate GHGs from existing sources of pollutants through the CPP.202 As discussed in Section 3.2, Congress in the 1970 amendments provided in section 111(d) that existing sources of pollutants that are not criteria pollutants or HAPs must also be regulated under the NSPS if an NSPS would apply to those sources but for the fact that they were not new. States must then write plans similar to SIPs to implement NSPS regulations for existing sources. The meaning of section 111 (d) is in dispute due to two competing House and Senate amendments to the CAA in 1990 that were not fully reconciled. But setting aside this interpretational dispute, the regulation of existing sources of GHGs under the CAA also sheds light on the challenges of fitting new pollutant risks into CAA provisions without imposing unnecessary costs on regulated entities or needlessly complicating regulation. As discussed in Section 3.2, the technology-based standard to which NSPS applies is the “best system of emission reduction” (BSER).203 Although the CAA provides that in establishing BSER EPA must consider the cost of achieving such reduction and any non–air quality health and environmental impact and energy requirement,” it does not define BSER.204 In setting BSER emission guidelines for existing sources of GHGs – specifically coal- and natural gas–fired plants – the Obama administration EPA assumed that power plants would, as they already do, switch among different sources of generation in order to provide electricity while also reducing GHG emissions. For example, a utility might reduce generation at its coal-fired power plant and rely more on generation from its other natural gas–fired or renewable plants or other plants. EPA took the most conservative “fuel switching” estimate from each region that it studied in order to set the final emissions guideline. Specifically, it investigated the three different electrical grids that exist within the United States – the Eastern, Western and Texas interconnections. Within each of these interconnections, where transmission lines are relatively well connected among the states in the interconnection, utilities often operate numerous plants and draw electricity from different plants at different times. EPA chose the 200 201 202 203 204

42 U.S.C. 7411(b)(1)(A). 80 Fed. Reg. 64,510, 64,529 (Oct. 23, 2015). 80 Fed. Reg. 64,662 (Oct. 23, 2015). See supra note 69 and accompanying text. Supra note 70 and accompanying text.

Stationary Sources, Movable Rules

101

interconnection in which it was most difficult to achieve GHG reductions through fuel switching and other measures as the national guideline.205 After establishing emission guidelines for coal- and natural gas–fired plants (annual tons of CO2 that could be emitted by each type of plant), the Obama administration EPA established total state goals for emission reductions based on the number of regulated plants in each state and how frequently they run.206 The Agency then included three optional building blocks that states could include in their plans to achieve these state goals. (Although these were optional, they were likely to be selected by many states if the CPP were implemented because the Agency used them to establish the stringency of the goals, in that the building blocks demonstrated which technologies were adequately demonstrated and thus the level of emission reductions that could realistically be achieved.) The building blocks included making coal-fired power plants more efficient, relying more on existing natural gas–fired plants and less on coal-fired power plants for electricity generation and relying more on new renewable plants and less on coal-fired power plants.207 Under the Obama administration EPA’s version of the CPP, states are not required to use these building blocks; they also may rely on measures such as energy efficiency, the operation of existing nuclear power plants or other alternative means of reducing GHGs from existing coal- and natural gas–fired plants.208 Further, states do not have to require that each individual plant meet the national emission guideline. Rather, states can combine reductions at different plants to achieve an overall state goal. But some of this important flexibility afforded to the states could be problematically constrained under the CPP as written by the Obama administration EPA. Some litigants and scholars argued that these building blocks go beyond the definition of best system of emissions reduction and that BSER should be limited to the “fenceline” of the power plant – in other words, that BSER only refers to the emission reductions that can be achieved at one electric utility’s power plant, excluding the possibility of running the plant less often and relying more on other generation, such as renewable or natural gas–fired plants operated by the same or other utilities at different locations.209 In the CPP, EPA noted that this construction of BSER contradicts the very operation of power plants, since utilities routinely rely on a variety of generating plants to provide electricity when it is demanded. Specifically, the Obama administration EPA believed that its interpretation of BSER is reasonable in large part because “operators themselves treat increments of generation as interchangeable between and among sources in a way that creates options for relying on various 205 206 207 208 209

80 Fed. Reg. 64,662, at 64,730. Id. at 64,743. Id. at 64,667. Id. at 64,729–30. For a detailed exploration of the “beyond the fence line” debate and potential interpretations of the meaning of BSER, see Nathan Richardson, Playing without Aces: Offsets and the Limits of Flexibility under Clean Air Act Climate Policy, 42 ENTVL. L. 735 (2012).

Hannah J. Wiseman

102

utilization levels, lowering carbon generation, and reducing demand as components of the overall method for reducing CO2 emissions.”210 Indeed, forcing states to comply with the CPP only by requiring that each individual power plant reduce CO2 emissions would be expensive and difficult to achieve and would ignore the realities of the electric grid – the fact that power plants rarely operate in isolation and instead switch among a variety of generation sources at different locations. And the fact that Congress did not specifically define BSER, combined with legislative history suggesting a relatively broad intent behind BSER, suggests that states will not be limited to needlessly difficult and expensive “within the fenceline” solutions for CPP compliance. But a narrow interpretation of BSER could lead to this inflexible and problematic result. Other flexibility within the CPP also creates a potential coordination problem. The CPP’s allowance for states to use either of two different ways of measuring compliance, called mass- and rate-based approaches – and its prohibition on sources in rate-based states trading with those in mass-based states, or vice versa, to meet CPP requirements – limits trading options and the ability to achieve CPP goals in the most efficient manner possible. It also requires states, which have very different types of sources in some cases as well as different political climates, to somehow agree on one approach or another if they wish to trade, and this agreement in some cases simply will not be achieved. But if the CPP or portions of it are preserved and are validated by courts, these coordination problems could be overcome; regulatory tools could be developed to allow rate- and mass-based states to trade. 3.4.2.2 Regulation of Air Pollutant Emissions from the Oil and Gas Industry Another area in which the CAA demonstrates a combination of both problematic nonadaptive features within some provisions and beneficial flexibility and adaptability in others is the oil and gas context. Oil and gas development in the United States has recently boomed due to the expansion of a new type of hydraulic fracturing to areas such as the Marcellus Shale in Pennsylvania and the Bakken Shale in North Dakota. With this expansion, emissions of air pollutants have increased. The oil and gas sector causes air pollutant emissions at numerous points, including at the wellhead (during the drilling and fracturing stages, as well as once the well is “shut in” and producing), the distribution lines that carry oil and gas from wells to processing points and pipelines, processing plants, compressor stations that compress natural gas to send it through pipelines, pipelines and refineries, among other points. Before and during this boom in oil and natural gas production, EPA used the CAA – at times more successfully than others – to limit pollutants from some of these stages. As with the carbon regulation example, NSPS, which applies to a broad range of 210

80 Fed. Reg. 64,662 at 64,703.

Stationary Sources, Movable Rules

103

pollutants and does not apply only to major sources, has proven to be one of the most flexible programs that has allowed EPA to address increased air pollutant emissions from the oil and gas industry. EPA has used this program to write standards limiting emissions from, among other sources, VOCs emitted from the hydraulic fracturing of natural gas wells,211 compressor stations,212 pneumatic controllers both between wellheads and processing plants and in processing plants (controllers that use natural gas to open and close valves),213 tanks and other vessels that store wastewater, condensate (a light form of oil) and other substances at oil and natural gas sites214 and natural gas processing plants.215 Further, EPA regulates particulate matter (PM), carbon monoxide (CO), NOx and sulfur oxides from petroleum refineries through NSPS standards,216 and in 2016 it finalized NSPS for methane emitted from newly fractured oil and gas wells.217 Some provisions of the CAA also have enabled EPA to regulate new and expanded risks that have emerged from increased oil and gas production and refining. As introduced in Section 3.2, when addressing HAPS under the CAA, Congress – after determining that efforts to established purely health-based standards had largely failed – set technology-based requirements for HAPS. But, to ensure that these requirements did not result in adverse health effects over time, Congress also included a residual risk provision within the Act218 that required EPA to study health impacts remaining after the application of technology-based standards and the need to further regulate specific HAP sources to address those impacts. EPA conducts Risk and Technology Reviews (RTRs) to fulfill its responsibilities under this residual risk provision,219 and a recently completed RTR of petroleum refineries resulted in tighter emissions standards for HAPs emitted from these refineries. As a result of its RTR for petroleum refineries, EPA determined that there were “developments in practices, processes, and control technologies” that warranted more stringent controls on emissions of HAPs from some types of vessels at facilities, flaring (burning off) of excess fuels at the facilities and other emissions at refineries, and it accordingly finalized a rule with more stringent limits in 2015.220 EPA gained knowledge of these changes in practices and technologies – including the ability to reduce emissions from flares – in part from a consent decree that it had previously entered into with a large refinery charged with previous CAA violations, and EPA incorporated this 211 212 213 214 215 216 217 218 219 220

40 C.F.R. § 60.5375. 40 C.F.R. §§ 60.5380–60.5381. 40 C.F.R. § 60.5390. 40 C.F.R. § 60.5395. 40 C.F.R. § 60.5400–60.5407. 40 C.F.R. §§ 60.100–60.109a. 81 Fed. Reg. 35,824. Supra note 118 and accompanying text. U.S. EPA, RISK AND TECHNOLOGY REVIEW, https://www3.epa.gov/airtoxics/rrisk/rtrpg.html. 80 Fed. Reg. 75,178, 75,182, 75,185 (Dec. 1, 2015).

104

Hannah J. Wiseman

knowledge in its 2015 rule.221 EPA then revised the consent decree to require the petroleum refinery to further reduce emissions.222 The Agency estimates that as a result of the consent decree, VOCs emitted from the refinery, including HAPs, will decrease from 678 tons per year in 2015 to 10 tons per year in 2019.223 Although this RTR is an example of increased stringency of standards, the Agency’s reliance on locating the best technology found anywhere in order to set standards and achieve further progress over time can have unintended effects. Officials identify the best technologies through the RACT/BACT/LAER clearinghouse maintained by EPA (with LAER applying to major sources that emit HAPs), and vendors that have developed new pollution control technologies have an incentive to sell these technologies and cause their technology to be required within at least one source permit. Once this technology enters the database, it could set the new standard for all similar sources, thus substantially improving vendor profits. There are industry concerns that vendors sometimes exaggerate the performance of their technologies due to these incentives, yet the technologies are still diffused as a result of each LAER requirement for each source relying on other, supposedly best technologies. In some cases, EPA’s ability to update rules to address new risks based both on the adaptability of the act, such as requirements for reviewing whether technology-based standards are adequately covering health risks, and on the broad definitions under NSPS cause tensions from a political economy perspective. As demonstrated by the consent decree just described, EPA often learns about new technologies and new abilities to reduce pollutants from industry sources. Regulated actors that provide these examples complain, at times, that innovative practices – particularly those not required by a formal decree or rule but are rather voluntarily implemented – become costly mandates. This type of complaint arose when EPA updated NSPS and HAPs for many components of the oil and gas industry, including requiring that operators of newly fractured gas wells capture most emissions of VOCs during the fracturing process.224 This rule relied in part on EPA estimates of achievable emission reductions based on the Natural Gas STAR Program, which is a “flexible, voluntary partnership that encourages oil and natural gas companies – both domestically and abroad – to adopt cost-effective technologies and practices that improve operational efficiency and reduce emissions of methane, a potent greenhouse gas 221

222 223

224

United States v. Marathon Petroleum Co., First Amendment to Consent Decree, Case No. 2:12-cv1154 at 3 (E.D. Mich. 2016), https://www.epa.gov/sites/production/files/2016-06/documents/firsta mendmentmarathonpetro-catlettsburgrefining-cd.pdf. See also JODY FOSTER & ROB BRENNER, CLEAN AIR AND TECHNOLOGY INNOVATION: WORKING CONCEPTS FOR PROMOTING CLEAN TECHNOLOGY INNOVATION UNDER THE CLEAN AIR ACT 19 (2013) (discussing the consent decree and the innovations that it generated). Id. First Amendment to 2012 U.S. v. Marathon Petroleum Co. Clean Air Act Consent Decree, https:// www.epa.gov/enforcement/first-amendment-2012-us-v-marathon-petroleum-co-clean-air-act-con sent-decree#provisions. 77 Fed. Reg. 49,490 (Aug. 16, 2012).

Stationary Sources, Movable Rules

105

and clean energy source.”225 In commenting on the proposed rule, the oil and gas industry objected that “[d]ata from optimal scenarios from the Natural Gas STAR Program should not be used as a basis for a rulemaking.”226 Indeed, this demonstrates the fine line that EPA must walk in taking advantage of the adaptability of the CAA – which allows for needed updates to rules – while also encouraging industry to achieve improvements beyond those in the formal CAA regulations. Too much reliance on industry achievements when writing new standards could in some cases discourage voluntary industry action, for fear that the improvements from these actions will result in mandates. Yet EPA also must investigate and understand new and improved practices and update regulations when new risks demand such updates. The NSPS and HAPS provisions for oil and gas provide a positive example of the CAA’s ability to adapt – albeit with some repercussions that some regulatory targets find to be objectionable. One CAA provision that could help to address this difficult balancing act is the little-used section 111(j), which allows a source to avoid new NSPS requirements for several years if the source is working to perfect an innovative new pollution control technology.227 EPA has also issued regulations that apply this waiver in the BACT context.228 However, both provisions are rarely used.229 The oil and gas sector also provides a stark example of the problematic nonadaptive aspects of the CAA and how rigid brightline standards can prevent EPA from updating rules when it needs to. Although EPA has been able to use the relatively flexible NSPS provisions to address air pollutant emissions from the booming oil and gas industry, it has had more difficulty applying more stringent emission standards to some portions of the industry, in part due to the inflexibility of the definition of major source. For example, in 2009, EPA found that a natural gas sweetening plant (which removes hydrogen sulfide from natural gas) was a major source when combined with the many natural gas wells that sent gas to the plant.230 In other words, EPA made its major source determination by aggregating the sweetening plant with the natural gas wells. The plant, gas wells and distribution lines between the plants and wells were all owned by one company. The plant itself emitted almost 100 tons of several criteria pollutants, but none reached the 100 ton per year threshold without the addition of emissions from flares at the nearby wells that burn off excess methane.231 225 226

227

228 229 230 231

U.S. EPA, NATURAL GAS STAR PROGRAM, http://www3.epa.gov/gasstar/. INTERSTATE NATURAL GAS ASSOCIATION OF AMERICA, COMMENTS REGARDING THE PROPOSED RULE, OIL AND NATURAL GAS SECTOR: NEW SOURCE PERFORMANCE STANDARDS AND NATIONAL EMISSION STANDARDS FOR HAZARDOUS AIR POLLUTANTS REVIEWS (Nov. 22, 2011), http://www.ingaa.org/File .aspx?id=17141. See FOSTER & BRENNER, CLEAN AIR AND TECHNOLOGY INNOVATION, supra note 221, at 15 (discussing the waiver and the flexibility that it would offer if used more often). Id. at 15 (citing 45 Fed. Reg. 52,676, 52,727 (Aug. 7, 1980)). Id. at 15–16. Summit Petroleum Corp. v. EPA, 690 F.3d 733, 739 (6th Cir. 2012). Id. at 736.

106

Hannah J. Wiseman

In this case, a major source determination would have required that the sweetening plant obtain a Title V operating permit under the CAA – a permit that combines various emissions standards and compliance obligations from other parts of the Act, such as NSR, into one document. But in order to be major under Title V of the CAA, the source had to emit at least 100 tons per year of a pollutant. The Sixth Circuit Court of Appeals found that EPA’s interpretation of its own regulation regarding source aggregation was inconsistent with the regulation and EPA’s previous interpretations and thus reversed the finding that the plant was a major source.232 One of the criteria for determining whether facilities should be aggregated together and count as one source for the purpose of defining a major source was whether the facilities were “adjacent,” and EPA interpreted adjacent to mean that the facilities had to be “functionally related, irrespective of the distance that separates them.”233 The court disagreed with this interpretation. Thus, even though the sweetening plant itself came very close to meeting the 100 ton per year threshold and it exceeded this threshold when combined with the nearby wells, the plant and wells were not considered a major source despite having the same air impacts as one, single facility that would have emitted the same amount of pollution. This rigid 100 ton per year cutoff for defining a major source for Title V purposes could lead to perverse results, with numerous sources with substantial air impacts avoiding CAA permitting requirements. This problem demonstrates more broadly the difficulty of drawing bright lines between major and minor or new and old sources and how these lines – often designed for reasons of political economy and predictability for regulated actors – can cause air pollution problems to persist, creating negative durability. If an industry with relatively small sources of pollution, such as thousands of oil and gas well sites, is booming and creating many new air emissions, EPA needs ways to address these sources that were previously considered minor and inconsequential. Fortunately, the diversity of the CAA’s provisions and the inclusion of both states and the federal government in air pollution control efforts can help to mitigate this inflexibility. For example, as described earlier, states are required to have minor source permitting programs in their SIPs,234 and some states have used these and similar programs to address new pollution risks. In Colorado, when oil and gas production steadily increased and tanks containing condensate (a light form of oil) at oil and gas sites became the second largest contributor to ground-level ozone problems,235 the state issued a special SIP that tightened up air quality controls on these tanks236 in order to try to avoid reaching nonattainment status. States also may 232 233 234 235

236

Id. at 744. Id. See supra note 165 and accompanying text. DALE WELLS, COLORADO DEPARTMENT OF PUBLIC HEALTH AND ENVIRONMENT, CONDENSATE TANK EMISSIONS 2, http://www.epa.gov/ttnchie1/conference/ei20/session6/dwells.pdf. STATE OF COLORADO, GENERAL CONSTRUCTION PERMIT, OIL AND GAS INDUSTRY CONDENSATE STORAGE TANK BATTERIES (May 15, 2013), https://www.colorado.gov/pacific/sites/default/files/AP_GeneralPermit-GP01-Condensate-Storage-Tank-Batteries.pdf. See also State of Colorado Comments, supra

Stationary Sources, Movable Rules

107

include more stringent requirements within their SIPs even when not required to do so for nonattainment reasons. Beyond the oil and gas context, the negative aspects of the durability of the CAA created by certain brightline definitions are exhibited most potently by grandfathered stationary sources, which avoid making upgrades or modifications in order to retain their status as existing sources. The CAA rule that existing sources are not subject to NSR – while providing a predictable standard for regulated entities and avoiding imposing draconian, difficult-to-meet standards on some plants – provides a strong disincentive to make much-needed upgrades to plants. Another provision that has stymied progress under the CAA is the netting provision, which offers flexibility to regulated sources but also has negative results. Under this provision, sources that would otherwise count as major when making modifications avoid this categorization by showing that, on the whole, emissions from their plants remain below the major source threshold. For example, even if one smokestack at a plant will emit more than 100 or 250 tons per year of a pollutant, if the plant operator can demonstrate that at another smokestack there will be emissions reductions, provided that the overall net emissions from the plant are below the major threshold, the plant can avoid NSR.

3.5 conclusion Clean Air Act regulation of stationary sources is in many respects a paradox. Parts of the Act that apply to stationary sources are quite nonadaptive, yet other parts inspire innovation and updating of pollution controls over time to adapt to changing circumstances. On the whole, the stationary source program represents a remarkable achievement. Indeed, the lessons gleaned from its rules and application could improve other environmental programs that suffer from similar challenges of addressing numerous, disparate, increasing numbers of sources and updating standards as new risks are discovered. A broad-brush set of lessons would include the following. First, a statute aiming to achieve durable environmental quality goals over time should provide an overall metric toward which individual regulations are aimed, such that limits on individual sources are not dwarfed by the expansion of those individual sources. Second, the statute should incorporate adaptive mandates – requiring the entities implementing the rules to periodically assess and consider updating standards in order to address new risks. Third, it should include catch-all categories such as section 111(d) of the CAA and broad definitions to account for the fact that certain problems cannot be fully understood or predicted when a statute is written. Fourth, the statute should include standards that are worded to require ratcheting up over time, such as best control technology, but word these goals in a way that encourages flexibility and innovation in note 151 (noting a 2008 Ozone Action Plan (SIP) that addressed oil and gas emissions of ozone precursors).

108

Hannah J. Wiseman

meeting those standards. This is the key difference between technology-based emissions standards – which allow for different technologies to meet a specific emissions limit – and technology standards, which command a particular type of technology. A related, important aspect of these standards is the incorporation of case-by-case permitting, which implements the required ratcheting up of standards over time combined with an informational mechanism such as the clearinghouse, which provides unusually detailed information on the best technologies implemented nationally and globally and supports this gradual leveraging of standards. A final consideration for writing durable, effective statutes is to allow adequate flexibility for entities implementing the statute. As other chapters in this book discuss, involving different levels of entities in the regulatory project – in the case of the CAA, states and the federal government – can encourage important innovation and allow for needed flexibility. There are challenges with this approach. On the one hand, too much flexibility can allow entities politically opposed to the regulatory project or those that lack adequate resources to perform at a lower level than the statute anticipates. Therefore, flexibility must be accompanied by adequate review mechanisms and resources. In the case of the CAA, EPA should have an option for enforcing state underperformance that is less of a “nuclear option” than the imposition of a federal implementation plan (FIP) in lieu of a SIP. Citizen suit provisions, too, allow for important review opportunities. Improving the resources available to entities implementing statutes has always been a challenge, but at the state level, relying primarily on legislative appropriations for agency funding poses difficulties. Funding agencies through permit fees paid by industry can be more effective, although this, too, risks encouraging the agencies to issue more permits simply to obtain the funds they need. On the other hand, states under the CAA sometimes have too little flexibility. The NSR provisions and NAAQS mandates might unduly constrain the states, forcing them to focus most of their energies on sources that are no longer the primary causes of certain air pollution problems. Indeed, Professor David Adelman proposes giving both EPA and the states more flexibility through a system that would allow EPA to set different SIP compliance deadlines for different states and to condition SIP approval on the implementation of specific programs with proven success, including programs that would address sources not covered by NSR.237 As this chapter has discussed, EPA, too, sometimes lacks the flexibility needed to regulate risks previously unaddressed by the CAA, such as climate change, largely due to brightline rules such as the definition of a major source. Providing alternative catch-all categories such as section 111(d) offers important flexible alternatives for EPA. There is no silver bullet for writing an effective environmental statute or regulation, including for climate issues. But the lessons from the stationary sources portion of the CAA provide useful hints as to a potentially successful combination of mechanisms. 237

Adelman, Environmental Federalism, supra note 11.

Stationary Sources, Movable Rules

109

appendix 3.1 new source performance standards (nsps) decision tree

Stationary source? “building, structure, facility, installation which emits or may emit any air pollutant”

NO

no NSPS

CAA § 111(a)(3); 40 C.F.R. § 60.2

YES Emits air pollutant subject to NSPS? broad definition, but standards generally apply only to pollutants regulated somewhere in the CAA; see standards in 40 C.F.R. § 60.30b through approximately 40 C.F.R. § 60.5165 (typically address criteria pollutants from various sources); but see CAA § 111(d) (other air pollutants that are not criteria pollutants or HAPs but still should be regulated under NSPS)

NO

no NSPS

CAA § 302(g); CAA § 111(b)(1)(A)

YES New? modification = increase in hourly emission rate of the pollutant construction = “fabrication, erection, or installation of an affected facility” (affected facility = facility/activity to which standard applies, such as “each exterior base coat operation, each overvarnish coating operation, and each inside spray coating operation” at a beverage can surface coating plant – see 40 C.F.R. 60.490 for this example)

NO

no NSPS

construction or modification must commence after publication or proposal of NSPS (CAA § 111(a)(2)) CAA § 111(a)(2); CAA § 111(a)(4); 40 CFR § 60.14(a)-(b); 40 C.F.R. § 60.2

YES

NSPS applies if EPA has written the NSPS standard for the source category. A best demonstrated technology (BDT) emission limitation for the pollutant at issue therefore applies. CAA § 111(a)(1)

Hannah J. Wiseman

110

appendix 3.2 prevention of significant deterioration (psd) new source review (nsr) decision tree

Source located in attainment area for pollutant at issue?

NO

no PSD NSR

YES Stationary source? [“emitting facility?”] “any building, structure, facility, or installation which emits or may emit a regulated NSR pollutant”

NO

no PSD NSR

NO

no PSD NSR

CAA § 165; CAA § 302(j); 40 C.F.R. § 51.166(b)(5)

YES Emits regulated NSR pollutant? broad definition, but generally applies only to pollutants regulated somewhere in the CAA CAA § 302(g); 40 C.F.R. §§ 51.166(b)(49), 52.21(b)(50)

YES Major source of air pollution? major sources = certain listed sources with potential to emit (PTE) 100 tpy, all others with PTE 250

NO

no PSD NSR unless modification itself is major

CAA § 169(1); 40 C.F.R. § 52.21(b)(1)i)

YES New? modification= significant net emissions increase 100 tpy CO, 40 tpy NOx, 40 tpy SO2, etc. construction = “any physical change or change in the method of operation (including fabrication, erection, installation, demolition, or modification of an emissions unit) that would result in a change in emissions”

NO

no PSD NSR

CAA § 302(g); 40 C.F.R. §§ 51.166(b)(49), 52.21(b)(50)

YES The PSD program (one type of NSR regulation) applies. A best available control technology (BACT) emission limitation for the pollutant at issue therefore applies. CAA § 165(a)(4)

Stationary Sources, Movable Rules

appendix 3.3 nonattainment nsr decision tree

Source located in nonattainment area for pollutant at issue?

NO

no nonattainment NSR

YES Stationary source? building, structure, facility, or installation that emits regulated NSR pollutant

NO

no nonattainment NSR

NO

no nonattainment NSR

NO

no nonattainment NSR

CAA § 172(c)(5); CAA § 302(j); 40 C.F.R. § 51.165(a)(1)(i)

YES Emits regulated NSR pollutant? generally applies only to pollutants regulated somewhere in the CAA CAA § 302(g); 40 C.F.R. § 51.165(a)(1)(xxxvii)

YES Major source of air pollution? PTE 100 or more tpy of any regulated NSR pollutant, except, e.g., 50 tpy VOCs serious and severe ozone nonattainm. areas, 10 tpy NOx extreme ozone nontattainm. area, etc.

unless modification itself is major

CAA § 302(j); 40 C.F.R. § 51.165(a)(1)(iv)(A)

YES New? construction, OR modification with significant net emissions increase 100 tpy CO, 40 tpy NOx, 40 tpy SO2, etc.

NO

no nonattainment NSR

CAA § 171(4); CAA § 111(a)(4); 40 CFR § 51.165(a) (1)(v)(A); (a)(1)(x)(A); (a)(1)(xviii)

YES The nonattainment program (one type of NSR regulation) applies. A lowest achievable emission rate (LAER) emission limitation for the pollutant at issue therefore applies. CAA § 173(a)(2)

111

1965

1970

Clean Air Act Amendments of 1970 direct EPA to set NAAQS and identify and write standards for HAPs, establish New Source Performance Standards, and direct states to write SIPs

CAA Amendments of 1977 1975 establish New Source Review program and direct EPA to write technology-based standards for HAPs 1980

1985

1990

CAA Amendments of 1990 establish list of HAPs, set more stringent limits for SO2 and NOx emissions and establish SO2 allowance trading program

1995

2000

Clean Air Interstate Rule largely replaces NOx Budget Trading Program

NOx Budget Trading Program begins 2005

2010

EPA issues CPP Final Rule under Section 111(d) of the CAA EPA issues CPP Repeal Proposed Rule

2015

Cross-State Pollution Rule replaces CAIR due to North Carolina v. EPA, 531 F.3d 896 (D.C Cir. 2008)

California Air Pollution Control Act signed into law

1945

1950

1955

1960

1965

Clean Air Act Amendments of 1970

Air Quality Act of 1967 introduces California waiver process

1970

1975

CAA Amendments of 1977 expand waiver to allow other states to adopt California standards

1980

1985 CAA Amendments of 1990 include new federal standards for mobile sources directly linked to 1990 existing California standards 1995

California’s AB 1493 launches new focus on GHG emissions in transportation

2000

NLEV program allows states to adopt standards more stringent than federal standards while similar to CA LEV standards

2005

Federal CAFE standards build off CA LEV 2010 standards to address GHG emissions 2015

EPA issues Endangerment Finding

4 Leveraged Federalism and the Clean Air Act The Case of Vehicle Emissions Control Barry G. Rabe*

The tailpipe emissions program of the Clean Air Act (CAA) demonstrates the unique possibilities to achieve durability and foster adaptability through policy innovation that designates distinct roles for both the federal government and a single state, California. It built on pioneering policy engagement by California in the decades before federal involvement while creating a unique mechanism to allow that state to continue to leverage subsequent innovation at the same time that the federal government maintained a central regulatory role that preempted policies by other states. This federalism framework has proved quite durable across multiple decades. It was sustained through subsequent legislative reauthorizations and major executive branch reforms despite repeated shifts in the partisan control of the presidency and the California governorship. Adaptability that facilitates incorporation of new scientific and technological information was driven in large part through the exclusive delegation to California of the opportunity to bring ongoing waiver requests to the federal government with exact statutory criteria guiding federal agency review and approval. This was highly unusual among other intergovernmental programs that employ some form of waiver and was used with consistency and frequency by California across multiple decades. The state’s need to address profound and enduring air quality problems and an economy with only limited engagement in vehicle manufacturing and assembly created ongoing political incentives to routinely pursue use of these adaptability mechanisms. Waivers were crafted and advanced by committed governors and legislators from both political parties and a formidable administrative agency with unusual staff depth and technical skills, the California Air Resources Board. These forces converged to enable California to make regular use of the adaptability provisions as new information about air quality risks and technological options emerged across multiple decades. Its receipt of a federal waiver could then allow *

I am grateful to Dallas Burtraw, Sanya Carley, Ann Carlson, Amelia Keyes, Michael Kraft, Kristin McCormack and Kathryn Harrison for thoughtful comments on earlier versions of this chapter. I also appreciate the excellent research assistance of Paul Sheridan, Hannah Smith, Emily Upton and Tom Van Heeke.

114

Barry G. Rabe

other states to join forces with California through a bandwagon effect that frequently served to pressure, or leverage, the federal government into accepting the waiver as a model for federal policy. In turn, the policy allowed some compliance flexibility through such provisions as fleet averaging, which made it possible to avoid imposing uniform technology in every vehicle that was manufactured. These provisions demonstrated continued durability and adaptability through a historic merger between tailpipe air quality and vehicle fuel economy standards adopted in 2009. This was established without new legislation and forged a partnership between a pair of federal executive agencies with minimal prior collaborative experience. California continued to play a central role both in prompting this step and subsequently in pushing for added regulatory rigor as the reduction of air contaminants was formally linked to efforts to reduce greenhouse gas (GHG) emissions. This program thus remained a vital element of air quality strategy while also evolving into a central plank in the federal government’s approach to mitigate climate change. By the mid-2010s, however, this policy would face a series of new challenges to its durability and continuing ability to use its flexibility provisions to regularly leverage added regulatory stringency. This chapter reviews the half-century odyssey of the California–federal governmental alliance on vehicular emissions control. It examines this policy’s sustained record of durability, long after the demise of its initial supportive coalition and across many partisan shifts of political control in Sacramento and Washington. It also demonstrates ongoing capacity for adaptability through continued use of its waiver provisions. This chapter begins with a political history of this air quality issue in California and the highly durable policy created to address it. The chapter reviews key design features, including flexibility provisions, of major legislative steps taken in Sacramento and Washington from the 1940s to the present. This analysis also considers the key factors that have propelled and sustained this policy, including its waiver system and subsequent methods for intergovernmental negotiation. This chapter will more fully define the distinctive and unique elements of leveraged federalism that have guided decades of policy in this area. The chapter will also consider the environmental, economic and political impacts of this experiment, introduce a series of significant challenges that stand ahead and consider prospects for future diffusion of this approach to other policy areas or nations.

4.1 political history and the emergence of leveraged federalism Mounting evidence from the 1950s and 1960s indicated that motor vehicle emissions posed substantial and expanding environmental health risks, moving this issue to center stage in the accelerating debate over how the federal government might respond to air pollution. The rapid expansion of the interstate highway system that had been launched as a national priority in the 1950s underscored the American

Leveraged Federalism and the Clean Air Act

115

affinity for cars and trucks.1 Any efforts to discourage driving through steep gasoline taxes or vehicle operation restrictions were political nonstarters. Instead, the abiding focus turned to engine design. This entailed policy that steadily required cleaner engines, necessitating a technological transformation that would allow unrestricted vehicle use while attempting to mitigate some of the environmental impacts of increased driving. By the mid-1960s, the United States appeared to be headed toward a national regulatory strategy for cleaner vehicles through performance standards imposed via federal government preemption of existing or emerging state and local policies. During this period, most vehicles used in America were built domestically, particularly in states of the industrial Midwest. Vehicle emissions were generated in every state, county and legislative district, but neither vehicles nor their pollution were confined within any jurisdictional boundaries. All evidence indicated that these emissions were a national problem that was only likely to worsen as the number of vehicles continued to increase and total mileage driven soared annually. Any policy that allowed for different standards and procedures in various states or regions seemed highly problematic politically, environmentally and economically given the national scope of vehicle manufacturing and the frequency of cross-border driving.2 Vehicle emissions climbed onto the American political agenda at a point where leaders from both political parties and branches of government recognized that some form of bold federal action was necessary. Political officials could claim credit for these steps, even if any demonstrable evidence of environmental improvement would likely be delayed for some time. Focusing exclusively on newly manufactured vehicles rather than used ones meant that citizens would only face added product cost if and when they purchased a new vehicle. Any pricing increase would be far less apparent to consumers than fuel taxation. This meant that political champions of the policy could largely avoid blame for increasing any costs of one of the more expensive purchases most citizens make. There was, of course, considerable opposition to such policy steps, particularly from major manufacturing firms. An industry often boastful of its technical and engineering prowess became quite animated in articulating the likely limits of its ability to modify vehicle and engine design to achieve significant emissions reductions. Ford Motor Company executive Lee Iacocca insisted in the 1960s, for example, that any policy that would mandate installation of pollution control devices such as the catalytic converter “will cause Ford to shut down.”3 This view was shared by a formidable base of political supporters, including Michigan’s stalwart 1 2

3

HENRY PETROSKI, THE ROAD TAKEN: THE HISTORY AND FUTURE OF AMERICA’S INFRASTRUCTURE (2016). WILLIAM R. LOWRY, THE DIMENSIONS OF FEDERALISM: STATE GOVERNMENTS AND POLLUTION CONTROL POLICIES chaps. 1 and 4 (rev. ed. 1996). Quoted in THOMAS FRIEDMAN, HOT, FLAT, AND CROWDED: WHY WE NEED A GREEN REVOLUTION – AND HOW IT CAN RENEW AMERICA 275 (2008).

116

Barry G. Rabe

Democratic Congressman John Dingell.4 But Dingell and his auto industry allies never mustered the votes to counter the political pressure to produce cleaner vehicles in the 1960s or in subsequent decades. A federal strategy of preemption of state standards followed, beginning with the 1967 Air Quality Act and continuing through subsequent federal amendments in 1970, 1977 and 1990. All of this was preceded by earlier policy efforts by California that reflected its experience with early and intensive impacts from vehicle-generated emissions that followed explosive population and economic growth directly linked to expansive use of cars and trucks. California’s unilateral actions from the 1940s through the 1960s included creation of local air quality districts, funding for research on health effects, expansion of public health capacity to monitor and respond to risks from exposure and development of legislation and regulations to produce cleaner vehicles. Indeed, the first elected American executive to sign significant air pollution control legislation linked to vehicle emissions was not Lyndon Johnson or Richard Nixon but rather Earl Warren, governor of California in the 1940s.5 The existence of substantial early policy engagement by one state posed some challenges to the design of subsequent federal policy. There was ample precedent for Congress to pursue vertical diffusion, simply folding early state policies into a new federal strategy that was binding across the nation.6 Emerging federal policy might draw from the lessons of prior state experience, as would occur with wetlands regulation in the 1970s and the development of a sulfur dioxide emissions trading scheme in the 1990s.7 There might even be particular credits or preferred status granted to pioneering or politically influential states, including latitude to sustain some aspects of program oversight.8 But in the area of vehicle emissions, Congress would go much farther in not only recognizing the early role and stake of California on this issue but also in designing through statute a unique flexibility mechanism that allowed the Golden State to drive subsequent policy formation through an enduring waiver process. This provision was introduced into 1967 federal legislation and sustained through three subsequent rounds of amendments. In 1977, it was expanded to allow other states to adopt California standards once a waiver request had been federally approved. 4

5

6

7 8

On Dingell’s political constituency and engagement on air quality issues, see RICHARD E. COHEN, WASHINGTON AT WORK: BACK ROOMS AND CLEAN AIR 31–39 (2nd ed. 1995). Warren signed the California Air Pollution Control Act into law in 1947, authorizing the creation of an air pollution control district in every California county. See NOGA MORAG-LEVINE, CHASING THE WIND: REGULATING AIR POLLUTION IN THE COMMON LAW STATE ch. 7 (2003). As former EPA Deputy Administrator Bob Perciasepe noted in 2014, federal environmental statutes adopted during the 1970s were “all borne out of state activity” and “modeled on what some states were already doing.” But none received specialized treatment through a waiver system as in the mobile source case. American Academy of Arts & Sciences, Comments on the Workshop on Durability and Adaptability in Energy Policy, Cambridge, MA (December 15, 2014). LOWRY, THE DIMENSIONS OF FEDERALISM, supra note 2, at 45; also see Chapter 6 of this volume. VIVIAN E. THOMSON, CLIMATE OF CAPITULATION: AN INSIDER’S ACCOUNT OF STATE POWER IN A COAL NATION (2017).

Leveraged Federalism and the Clean Air Act

117

This form of leveraged federalism is likely without parallel in the United States, both in environmental policy and in other arenas of regulatory federalism. It formally allows a single state to adopt its own regulatory standards through legislation or administrative provisions that are more demanding than federal standards. California cannot unilaterally put these standards into operation but is instead provided a formal channel to seek a review and waiver of federal policy. This includes very specific criteria for federal waiver request decisions, contrary to other waiver processes that give considerable interpretative latitude to federal agencies and can vary markedly in their use across administrations.9 It leads to the possibility that federal policy not only can be continually modified by future Congresses but also can be leveraged by the actions of the one state that, in this instance, has the largest population, number of vehicles and annual vehicle miles driven. Such a waiver approach creates the possibility of a two-part national market for vehicles (California and those states that follow its standards, on the one hand, and the remaining states and territories, on the other). It also triggers a pattern whereby the federal government might not only approve the waiver for the requesting state but ultimately also elevate it into federal policy to sustain nationwide consistency in response to this state-led leveraging process. For nearly a half-century, this unique system of intergovernmental regulation has proven durable and triggered a transformation in US vehicle emissions reduction technology. Dozens of policy modifications have been made in past decades through its adaptability mechanisms, manifested in California waiver requests that ultimately received federal approval and were eventually applied in some form on a national basis. This basic policy framework has endured far-reaching changes of executive branch political control in both California and Washington, DC (Table 4.1). The most prolonged period of intergovernmental conflict over policy direction during the George W. Bush presidency was ultimately broken by a combination of continued state pressure on the federal government, expanded federal court engagement and a significant modification of the policy through a political bargain struck by President Barack Obama some forty-two years after the adoption of the 1967 Air Quality Act. The policy has thus proven durable, but it has done far more than just survive and drift. The leveraged federalism mechanism has delivered sustained policy adaptability that reflects evolving scientific understanding and technological capacity. As a result, American vehicles have continually been redesigned in past generations to reduce their environmental imprint in ways almost unthinkable when weighed against the technology employed during the 1960s. The California mode, then, is in many respects a poster child for adaptive policy in a federal system. However, this 9

In the area of educational policy, see Kenneth K. Wong, Federal ESEA Waivers as Reform Leverage: Politics and Variation in State Implementation, 45(3) PUBLIUS: THE JOURNAL OF FEDERALISM 405–26 (2015). In the area of health policy, see FRANK J. THOMPSON, MEDICAID POLITICS: FEDERALISM, POLICY DURABILITY, AND HEALTH REFORM (2012).

Barry G. Rabe

118

table 4.1 Political Leadership and Partisan Control of US Federal and State of California Executive Branches, 1966–2016

Year US president 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003

Lyndon Johnson Lyndon Johnson Lyndon Johnson Richard Nixon Richard Nixon Richard Nixon Richard Nixon Richard Nixon Richard Nixon/ Gerald Ford Gerald Ford Gerald Ford Jimmy Carter Jimmy Carter Jimmy Carter Jimmy Carter Ronald Reagan Ronald Reagan Ronald Reagan Ronald Reagan Ronald Reagan Ronald Reagan Ronald Reagan Ronald Reagan George H. W. Bush George H. W. Bush George H. W. Bush George H. W. Bush Bill Clinton Bill Clinton Bill Clinton Bill Clinton Bill Clinton Bill Clinton Bill Clinton Bill Clinton George W. Bush George W. Bush George W. Bush

US president party affiliation

California governor

California governor party affiliation

Democrat Democrat Democrat Republican Republican Republican Republican Republican Republican

Pat Brown Ronald Reagan Ronald Reagan Ronald Reagan Ronald Reagan Ronald Reagan Ronald Reagan Ronald Reagan Ronald Reagan

Democrat Republican Republican Republican Republican Republican Republican Republican Republican

Republican Republican Democrat Democrat Democrat Democrat Republican Republican Republican Republican Republican Republican Republican Republican Republican Republican Republican Republican Democrat Democrat Democrat Democrat Democrat Democrat Democrat Democrat Republican Republican Republican

Jerry Brown Jerry Brown Jerry Brown Jerry Brown Jerry Brown Jerry Brown Jerry Brown Jerry Brown George Deukmejian George Deukmejian George Deukmejian George Deukmejian George Deukmejian George Deukmejian George Deukmejian George Deukmejian Pete Wilson Pete Wilson Pete Wilson Pete Wilson Pete Wilson Pete Wilson Pete Wilson Pete Wilson Gray Davis Gray Davis Gray Davis Gray Davis Gray Davis/Arnold Schwarzenegger

Democrat Democrat Democrat Democrat Democrat Democrat Democrat Democrat Republican Republican Republican Republican Republican Republican Republican Republican Republican Republican Republican Republican Republican Republican Republican Republican Democrat Democrat Democrat Democrat Democrat/ Republican (continued)

Leveraged Federalism and the Clean Air Act

119

table 4.1 (continued)

Year US president 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

George W. Bush George W. Bush George W. Bush George W. Bush George W. Bush Barack Obama Barack Obama Barack Obama Barack Obama Barack Obama Barack Obama Barack Obama Barack Obama

US president party affiliation

California governor

California governor party affiliation

Republican Republican Republican Republican Republican Democrat Democrat Democrat Democrat Democrat Democrat Democrat Democrat

Arnold Schwarzenegger Arnold Schwarzenegger Arnold Schwarzenegger Arnold Schwarzenegger Arnold Schwarzenegger Arnold Schwarzenegger Arnold Schwarzenegger Jerry Brown Jerry Brown Jerry Brown Jerry Brown Jerry Brown Jerry Brown

Republican Republican Republican Republican Republican Republican Republican Democrat Democrat Democrat Democrat Democrat Democrat

process has not been seamless nor eliminated environmental risks from vehicle use. Indeed, its considerable successes may serve to further discourage other approaches, such as significant increases in gasoline prices through taxation that might be more efficient but would be far more politically contentious. Nonetheless, this case offers significant evidence that this approach to federalism has largely worked as intended, raising the specter of its possible diffusion to other policy areas or other federal systems of government.

4.2 the environmental problem posed by vehicle emissions Emissions from vehicles lagged behind those from smokestacks as an air quality concern in the United States, reflecting the massive pace of industrialization and unrestricted use of coal for electricity and manufacturing in many regions of the nation.10 There was, however, a gradual recognition that the dramatic expansion in the numbers of motor vehicles on American roads and the explosive growth in miles driven annually were creating additional air quality and public health challenges. California emerged as ground zero in this environmental epiphany, particularly Los Angeles County and surrounding territory. The unique combination of topography, air currents and temperatures within this basin creates conditions to capture groundlevel smog or haze for extended periods.11 10 11

J. CLARENCE DAVIES, THE POLITICS OF POLLUTION (1975). For a more expansive history of this era, see DAVID VOGEL, Protecting Air Quality, in CALIFORNIA GREENIN’: HOW THE GOLDEN STATE BECAME AN ENVIRONMENTAL LEADER (2018).

120

Barry G. Rabe

Any natural proclivity toward adverse air quality was exacerbated by stunning increases in population and vehicle usage during the mid-1900s. Los Angeles County became home to more than 5 million people, greater than the population of forty-two individual states, by 1960. More than 1.2 million additional residents were added between 1950 and 1956 alone. This population growth coincided with rapid development of new housing and commerce across Los Angeles County, featuring “sprawling low-density single family homes, monoculture communities . . . long commutes and the addiction to gas.”12 By 1956, approximately 5 percent of all motor vehicles registered in the United States were based in Los Angeles County. This reflected nearly 3 million vehicles packed into a 1,600-square-mile topographic zone that was uniquely at risk from entrapped emissions. There were many early warning signs of mounting air quality issues in this area but considerable uncertainty and debate over whether vehicles were really a significant source of these problems. Los Angeles County suffered through a widely publicized “five day siege of smog” in 1953, as well as other similar episodes during the decade.13 Vehicle industry representatives actively campaigned to raise doubts, arguing that industrial sources were the cause rather than vehicles.14 The Los Angeles City Council devoted substantial resources to investigate whether a synthetic butadiene plant operated by the Southern California Gas Company might be the literal smoking gun in the area’s “peculiar atmospheric condition” that would ultimately be defined as its “smog problem.” But periodic plant closures and the installation of pollution control devices ultimately led to the recognition that the smog problem persisted and must be attributable to other sources.15 Comparable problems began to surface in other regions with heavy vehicle use, most notably New York City and the Northeast, increasingly suggesting that motor vehicles might be significant factors. The search for answers produced multiple strands of research supported by different institutions that provided a definitive and mutually reinforcing message that smog was principally linked to motor vehicles and constituted a growing public health threat in Los Angeles County and beyond. First, California Institute of Technology biochemist Arie Haagen-Smit made a powerful discovery of how ozone, a primary component of smog, was produced, which was not previously understood, including identification of a central role for automotive exhaust. 12

13

14

15

WYN GRANT, AUTOS, SMOG AND POLLUTION CONTROL: THE POLITICS OF AIR QUALITY MANAGEMENT IN CALIFORNIA 40 (1995). George Gonzalez, Urban Growth and the Politics of Air Pollution: The Establishment of California’s Automobile Emission Standards, 35(2) POLITY 213−36 (2002). For example, the Ford Motor Company responded to a Los Angeles County inquiry on its research on motor vehicle emission control by noting that “[t]he Ford engineering staff, although mindful that automobile engines produce exhaust gases, feels that these vapors are dissipated in the atmosphere quickly and do not represent an air pollution control problem.” Quoted in DAVID CARLE, INTRODUCTION TO AIR IN CALIFORNIA 135 (2006). VOGEL, in CALIFORNIA GREENIN’, supra note 11.

Leveraged Federalism and the Clean Air Act

121

Second, local leaders in Los Angeles funded and promoted research demonstrating that motor vehicles were central to expanding smog problems. A series of regional business leaders joined forces to create the Air Pollution Foundation of Los Angeles, which sponsored and disseminated findings of numerous studies examining motor vehicle impact on air quality. This included an influential 1956 report that concluded that “motor vehicles were the principal contributors to smog in Los Angeles.”16 Third, the federal government complemented research funding from the state of California with its own research program on national air quality. This began in the mid-1950s with grants to state and local governments, including significant resource transfers to California. A prominent focus of this effort was expanding research on possible links between motor vehicle emissions and public health, particularly respiratory risks. This led to the publication of a 1962 Surgeon General’s Report on the impact of motor vehicle emissions that deemed them a substantial and growing public health threat, which reached a broad audience given extensive media coverage and thereby elevated issue salience nationally.17 Vehicle emissions were thus confirmed to play a central role in a range of urban air maladies, many of which can combine to contribute to a range of respiratory problems and premature death. Vehicle engines routinely produce significant quantities of carbon monoxide (CO), hydrocarbons, nitrogen oxides, sulfur oxides, particulate matter (PM) and volatile organic compounds (VOCs). This is most commonly linked to smog, representing a blending of the words smoke and fog, although this is only one visible aspect of the public health risks entailed.18 Emissions from American vehicles increased steadily over the course of the twentieth century, particularly after World War II when vehicle use expanded dramatically. Between 1940 and 1970, vehicle emissions of CO, nitrogen oxides and VOCs more than doubled.19

4.3 the environmental policy response: foundational state and federal statutes Air quality linked to motor vehicle use transformed from a fringe concern to a leading agenda item with spasms of legislative activity in both Sacramento and Washington, DC. These began two years after the end of World War II and continued into the twenty-first century. They tended to follow a pattern whereby California would take a lead role, beginning in a period when there was no 16 17

18 19

Gonzalez, Urban Growth and the Politics of Air Pollution, supra note 11, at 227. For an excellent overview of this issue, see CHARLES O. JONES, CLEAN AIR: THE POLICIES AND POLITICS OF POLLUTION CONTROL 61–68 (1975). DOUGLAS S. EISINGER, SMOG CHECK: SCIENCE, FEDERALISM, AND THE POLITICS OF CLEAN AIR 6–7 (2010). U.S. EPA, NATIONAL AIR POLLUTANT EMISSIONS ESTIMATES, 1940–1988 58–62 (1990), https://hero.epa .gov/hero/index.cfm/reference/download/reference_id/5674; GARY C. BRYNER, BLUE SKIES, GREEN POLITICS: THE CLEAN AIR ACT OF 1990 AND ITS IMPLEMENTATION ch. 4 (rev. ed. 1995).

122

Barry G. Rabe

consequential state or federal policy nor federal provision to preempt states from imposing restrictions on vehicular emissions.20 But the federal government ultimately engaged, not just replicating California’s steps but also creating a unique adaptation mechanism to allow for mutual leveraging of efforts. This continued to recognize the outsized impact of these emissions on Californian life while also confirming that it was likely not feasible economically to allow each state to set its own policy and standard. The evolving approach would not involve market mechanisms such as fuel taxes to increase the costs of driving and thereby reduce emissions. Such policy options surfaced periodically on the political fringes, reflecting the extreme political sensitivity to restrictions on American vehicle usage and access. Instead, both state and federal governments would opt for a form of technology forcing, one that political scientist Charles Jones characterized as “speculative augmentation.” Jones’s landmark study of American air quality policy described such augmentation as embracing policy solutions that lie beyond what is technically or administratively feasible at the time of policy adoption.21 However, this policy to alter vehicle design in order to reduce their emissions would allow for considerable compliance flexibility. Rather than mandate use of identical emissions abatement technology for every car and truck, fleet averaging was authorized. This provided manufacturers with considerable latitude in applying new technology to vehicle fleets, as long as average emission targets were achieved. This helped mitigate concerns that various waves of new standards might undermine the viability of certain engines, vehicles or even manufacturing firms. Under this regulatory regime, both state and federal authorities insisted on a technological transformation that could ultimately produce vastly cleaner engines. The resulting policy delegates regulatory authority to the US Environmental Protection Agency (EPA), mandating the creation of emissions standards for certain vehicle and engine classes. It addresses both tailpipe emissions during vehicle operation and evaporative emissions during vehicle fueling. In some instances, actual numeric standards are specified in legislation. EPA is required to consider environmental and health impacts, costs and technological feasibility in setting emission limits. EPA has authority under section 202 of the CAA to promulgate standards for any class of vehicle or engine that is judged to generate air emissions that cause or contribute to air pollution that may reasonably be anticipated to endanger public health or welfare.22 The Agency may classify vehicles based on their gross weight, horsepower, type of fuel used for operation or additional factors. It has historically 20 21 22

LOWRY, THE DIMENSIONS OF FEDERALISM, supra note 2, ch. 4. JONES, CLEAN AIR, supra note 17. U.S.C. at 7521; portions of this paragraph are drawn heavily from Paul Sheridan, Emission Standards for Mobile Sources (unpublished paper, University of California Los Angeles Law School, October 14, 2015; on file with author).

Leveraged Federalism and the Clean Air Act

123

distinguished vehicles intended for highway use into two broad categories. Lightduty vehicles have a gross vehicle weight rating up to 8,500 pounds; heavy-duty vehicles exceed the 8,500-pound level. Manufacturers of these vehicles must demonstrate that they can pass emission tests that simulate both short- and longrun usage in order to obtain certificates of conformity to permit sale in the United States. Monetary penalties for noncompliance are specified in the legislation, although they were used modestly and sparingly until the mid-2010s. Manufacturers must also provide a warranty to vehicle purchasers that guarantee effective operational performance of required pollution control technology over specified travel distances. These provisions were designed for implementation on a national basis by the lead federal agency, but there was one important exception. California was not only an early policy architect that inspired federal policy design, but it also retained an ongoing role in developing its own policies alongside federal ones.23 This entailed a unique waiver provision granted only to California, specified in 1967 federal legislation and retained in all subsequent amendments. This waiver allows California to propose policies that would be more rigorous than those of the federal government, provided that EPA approves the waiver request based on formal criteria specified in statute. Such a waiver makes possible a leveraged federalism process that will be discussed extensively later in this chapter. This process emerged through the give-and-take of California and the federal government over multiple decades. Legal scholar Ann Carlson has outlined a series of steps of “iterative federalism,” whereby state and federal authorities not only take turns in redefining policy over time but influence each other along the way, with California almost always taking the first step.24 This analysis will not revisit each step of that iterative process but rather note that there have been clusters of significant statutory adoptions in both Sacramento and Washington, DC, over more than a halfcentury. These have served to sustain an approach toward motor vehicle emissions that has proven both durable and adaptable. California as Policy Pioneer. The state of California essentially invented air quality policy for motor vehicles through a series of statutes adopted between 1947 and 1960. These foundational pieces not only served the United States but also constituted “the world’s first auto emission regulations.”25 The statutes were, however, influenced by prior experience of local governments in California, which sustained an 23

24

25

ROGER KARAPIN, POLITICAL OPPORTUNITIES FOR CLIMATE POLICY: CALIFORNIA, NEW YORK AND THE UNITED STATES ch. 6 (2016). Ann Carlson, Iterative Federalism and Climate Change, 103 (3) NORTHWEST. U. L. REV. 1097–161 (2009), http://webshare.law.ucla.edu/Faculty/bibs/carlson/Carlson-IterativeFederalism.pdf. Also see DAVID VOGEL, TRADING UP: CONSUMER AND ENVIRONMENTAL REGULATION IN A GLOBAL ECONOMY (1995). DAVID VOGEL, THE POLITICS OF PRECAUTION: REGULATING HEALTH, SAFETY, AND ENVIRONMENTAL RISKS IN EUROPE AND THE UNITED STATES 105 (2012).

124

Barry G. Rabe

ongoing local role in policy implementation. During each of the first five decades of the twentieth century, Los Angeles County adopted air pollution control laws and established some local air governance capacity, setting the stage for more ambitious policies that would subsequently be launched on a statewide basis. These earlier policies generally focused on industrial pollution sources, but an expansion to consider mobile sources began in 1947 when California adopted its Air Pollution Control Act.26 Passed with unanimous support in both chambers of the California legislature and signed into law by Republican Governor Earl Warren, this legislation authorized the creation of local air pollution control districts. Funding was provided to hire professional staff and develop expertise in air quality issues of particular concern to various local jurisdictions around the state. As political scientist David Vogel noted, “[a] new emphasis on scientific research and highly trained investigators began altering the status of smog workers.”27 By 1961, California was spending more money on state and local air quality research and staff than all other states combined.28 That expansion was especially evident in the Los Angeles Air Pollution Control District (LAAPCD), which received nearly $10 million from state grants between 1947 and 1958 and became the “best funded air pollution control agency in the nation.”29 California’s 1947 air quality legislation did not establish statewide vehicle emission standards, but its financial support for the LAAPCD enabled it to confront air quality issues and play a longer-term role in policy development and implementation as the emerging national epicenter for these concerns. The LAAPCD formally endorsed statewide standards for motor vehicle emissions and the mandatory installation of pollution control equipment in 1958. One year later, additional state legislation required the state Department of Public Health to establish statewide standards that were adequate to protect public health and improve visibility. However, this legislation deferred to local authorities on the particulars of any such standards, raising questions about the viability of allowing potentially dozens of different standards for vehicles that routinely crossed local jurisdictional boundaries.30 This further prompted state elected officials to go one step farther in 1960. The Motor Vehicle Pollution Control Act created a lead state agency, the California Motor Vehicle Pollution Control Board, to approve new technologies used to reduce vehicle emissions and meet California Department of Public Health emission standards. This bill passed with broad bipartisan support, although it did not go 26

27 28 29

30

For an excellent review of this legislation and era in California air quality policy, see MORAG-LEVINE, CHASING THE WIND, supra note 5, at 124–133. VOGEL, in CALIFORNIA GREENIN’, supra note 11, at 5–6. JONES, CLEAN AIR, supra note 17, at 113–15. SCOTT DEWEY, DON’T BREATHE THE AIR: AIR POLLUTION AND U.S. ENVIRONMENTAL POLITICS, 1945–1970 44 (2000); Gonzalez, Urban Growth and the Politics of Air Pollution, supra note 11, at 224–25. Matthew Weinbaum, Implications of Automotive Emission Restrictions in California, 4 FEDERALISM-EJOURNAL, 37–48 (February 2004).

Leveraged Federalism and the Clean Air Act

125

as far as LAAPCD had wanted in mandating vehicle inspection and testing. This led to a series of state mandates for vehicle emission control technology beginning in the 1963 model year with positive crankcase ventilation, followed in the next year by requirements for various types of catalytic converters. The first standards for hydrocarbons and carbon monoxide were established for model year 1966.31 Legislation adopted in 1967 sustained this model and gave the newly created California Air Resources Board (CARB) broader authority to implement these provisions and sustain an active local government role. As Democratic Governor Edmund Brown noted in the 1970s, these policies were designed to make his state “the ecological gold standard.” Brown declared that “our problems apply to every state in the country” and that California was “a good place for these problems to be studied and their solution found.”32 The Federal Response. California launched the search for solutions to the challenge of motor vehicle emissions but was not acting in isolation. Fourteen other states had adopted some form of air pollution control by 1960, although most of these were quite modest and focused exclusively on industrial emissions.33 The federal government also began to define its role and fashion a response, not quite in tandem with California but along somewhat parallel and interactive lines. This evolved substantially through the 1970s, ultimately producing a leveraged federalism relationship between California and the federal government. This intergovernmental partnership was launched in 1955 when Congress adopted and President Dwight Eisenhower signed into law the federal Air Pollution Control Act. This has commonly been viewed as largely a symbolic step, formalizing the standing of air pollution on the federal environmental agenda. It also provided funding to California for research and staff development that boosted the state’s unilateral efforts to build policy analysis and development capacity. This initial federal foray into air quality expanded markedly in subsequent years to consider vehicle emissions, reflected in an accelerating pace of congressional hearings on the topic. Many of these were convened in California, framing the salience of this localized issue for a national audience. In these hearings, federal legislators learned about new technologies used only in California through industry response to state policy, leading them to wonder aloud why they were not available on a wider basis.34 In 1963, the first version of the modern CAA was adopted, albeit with a modest focus. This included assigning the US Department of Health, Education and Welfare (HEW), the lead federal environmental protection unit at that point, to encourage industry to find ways to reduce emissions from motor vehicles. Two years later, Congress went farther through amendments to the 1963 legislation 31 32 33 34

EISINGER, SMOG CHECK, supra note 18, at 23. Quoted in VOGEL, in CALIFORNIA GREENIN’, supra note 11, at 174. JONES, CLEAN AIR, supra note 17, at 68. Id. at 63.

126

Barry G. Rabe

empowering HEW to create federal standards for carbon monoxide, hydrocarbons and nitrogen oxides. The resulting standards were largely based on those already in place in California, promulgated in 1966 and operational in the 1968 model year. The legislation did not expressly address the issue of federal preemption of other states in the future, raising concerns among California officials over their future role in this area. The 1965 legislation allowed the rest of the nation to begin to keep pace with what California was already doing. It also set the stage for full definition of how intergovernmental duties would be allocated over the longer term. This issue dominated congressional debate in 1967 and led to adoption of the CAA amendments. At that point, the National Conference of State Legislatures reported that approximately 100 bills had been introduced into various state legislatures on air quality, many of which deviated from key elements of the CAA and the existing California legislation. An industry leader contended that “[a] multiplicity of differing State standards would create chaotic conditions in this Nation’s largest industry.”35 This view contributed to industry demands that the federal government should preempt this field, precluding all states, including California, from considering any unilateral policies on motor vehicle emissions. Many legislators found this argument appealing, particularly representatives from both political parties in states with a considerable vehicle manufacturing presence. However, California’s legislative leaders proved formidable. These included Republican Senators Thomas Kuchel and George Murphy, as well as House leaders who secured considerable uniformity among the state’s thirty-seven-member delegation to the House of Representatives, a group that had expanded in recent decades given California’s steady population growth in comparison with national averages. This delegation actively engaged in debates and lobbying for some form of designation that would preserve California’s authority to address motor vehicle emissions as an exception in any federal preemption strategy. They emphasized California’s unique risks from these emissions and historic leadership role in addressing them. At the suggestion of New York Democratic Senator Robert Kennedy, they also advanced the case that this should be framed as a federalism issue, whereby states with particular issues or needs should receive greater latitude to fashion their own policy responses. These types of messages were targeted toward southern legislators. Such officials had little direct experience with vehicle manufacturing at home, but they resonated with the idea of a “states’ rights” strategy that protected the interests of California against pressures from states from the industrial North, given their own ongoing battles with federal authorities over issues related to racial justice.36

35 36

Quoted in id. at 68. VOGEL, in CALIFORNIA GREENIN’, supra note 11, at 177; JAMES KRIER & EDMUND URSIN, POLLUTION AND POLICY: A CASE ESSAY ON CALIFORNIA AND EXPERIENCE WITH MOTOR VEHICLE AIR POLLUTION, 1940–1975 182 (1977).

Leveraged Federalism and the Clean Air Act

127

The California case for special status was ultimately addressed in 1967 legislation that generally established federal preemption for motor vehicle emissions. However, the legislation also allowed EPA to “waive application” of preemption to “any State” that had adopted standards “for the control of emissions from new motor vehicles or new motor vehicle engines prior to March 30, 1966, if the State determines that the State standards will be, in the aggregate, at least as protective of public health and welfare as the applicable Federal standards.”37 The statute did not explicitly name any state but was solely applicable to California, given its unique role in policy development prior to the March 1966 cutoff point. The legislation also offered specifics about the criteria EPA was to use in evaluating waiver requests. This differed markedly from conventional practices in other intergovernmental waiver programs, whereby only broad terms were outlined in statute, giving agencies enormous discretion to grant or reject waiver requests and to define the criteria and review process.38 In contrast, the motor vehicle waiver process for California “would seem to give EPA little latitude to turn down the waiver request,” although the statute did specify that No such waiver shall be granted if the Administrator finds that:

(A) the determination of the waiver is arbitrary and capricious, (B) such State does not need such State standards to meet compelling and extraordinary conditions, or (C) such State standards and accompanying enforcement procedures are not consistent with section 202(a) of this part.39 These provisions established the key elements of the leveraged federalism system that would endure subsequent rounds of amendments and implementation. They also set the stage for another major political battle over the particulars of federal standards and other key elements of the federal vehicle emission control strategy. The political salience of air pollution from motor vehicles reached its peak in 1970, reflecting continued episodes of local emission concentrations, substantial media attention and intensified public concern about the topic.40 This also collided with electoral incentives, with the possibility that the anticipated 1972 presidential race between incumbent Richard Nixon and a Democratic challenger would feature a race-to-the-top opportunity to demonstrate environmental credentials 37 38

39

40

Clean Air Act, 42 U.S.C. § 7543 (1967). Wong, Federal ESEA Waivers as Reform Leverage, supra note 9; Kenneth K. Wong, Education Governance in Performance-Based Federalism, in EDUCATION GOVERNANCE FOR THE 21ST CENTURY: OVERCOMING THE STRUCTURAL BARRIERS TO SCHOOL REFORM (Paul Manna and Patrick McGuinn eds., 2013), ch. 8; THOMPSON, MEDICAID POLITICS, supra note 9. Clean Air Act, 42 U.S.C. § 7543 (1967). For a thoughtful overview, see JAMES E. MCCARTHY, CALIFORNIA’S WAIVER REQUEST TO CONTROL GREENHOUSE GASES UNDER THE CLEAN AIR ACT (2008), https://www.hsdl.org/?view&did=482829. DAVIES, THE POLITICS OF POLLUTION, supra note 10; RICHARD N. L. ANDREWS, MANAGING THE ENVIRONMENT, MANAGING OURSELVES: A HISTORY OF AMERICAN ENVIRONMENTAL POLICY (2nd ed. 2006).

128

Barry G. Rabe

through tough stances on air quality. The California waiver process was revisited in subsequent reauthorizations, particularly in the House, where it was challenged again by members from vehicle manufacturing districts but ultimately sustained.41 Other key sections of the CAA were changed markedly through amendments signed into law by Nixon in 1970. These included the creation of the Federal Motor Vehicle Control Program, which established national emissions standards for carbon monoxide, nitrogen oxides, VOCs, particulates and fuel evaporation.42 The program required reductions of 90 percent in carbon monoxide and hydrocarbon emissions from 1970 levels by the 1975 model year. Comparable nitrogen oxide emission reductions were required from 1971 levels by the 1976 model year. This program was backed by testing and enforcement provisions, including fines on manufacturers for noncompliance and vehicle owners who removed pollution control devices after purchase. Citizens were authorized to bring suits, either against polluters who violated standards or against EPA for failure to take specified actions.43

4.4 durability meets adaptability: leveraging change across decades These foundational state and federal statutes established a policy base for regulating vehicle emissions that has proved remarkably durable and flexible in succeeding decades. The ongoing operations of the CAA continue to reflect many of these early design features despite subsequent legislative steps in both Sacramento and Washington. In particular, this intergovernmental sharing of responsibilities allows for continued cross-pressures to leverage further steps to steadily mandate evercleaner vehicle engines. This transition was aided by the near-instant jolt of credibility accorded to this policy by its early success in facilitating rapid deployment of catalytic converters in motor vehicle fleets within a few years of policy adoption. Despite considerable industry adamance in the late 1960s that such technology was neither reliable nor ready for mass application, the transition was accomplished relatively swiftly, with significant impact on vehicle emissions and only modest impact on vehicle purchase prices.44 This technological phase-in was coordinated with a phase-out of leaded gasoline. The gasoline transition was heavily influenced by the growing concern about public health risks linked to lead exposure and the rapid expansion in the supply of nonleaded gasoline that facilitated extended catalytic converter operation, since leaded fuel use would harm their functioning45 (see Chapter 5 for an extensive 41 42 43 44 45

JONES, CLEAN AIR, supra note 17, at 186–88. BRYNER, BLUE SKIES, GREEN POLITICS, supra note 19, at 157–58. Id. at 100–1. EISINGER, SMOG CHECK, supra note 18, at 25. VOGEL, THE POLITICS OF PRECAUTION, supra note 25, at 1.

Leveraged Federalism and the Clean Air Act

129

discussion about the federal phasing out of lead from fuel). Despite early and enduring industry opposition, which continued through legislative amendment processes in the 1990s, it became routine for policy proponents to cite this early triumph as a model that would likely be replicated in subsequent rounds of advancing technology through speculative augmentation. As political scientist Gary Bryner noted, this early success led to ongoing demands that “even more stringent controls be placed on tailpipe emissions.”46 This early impact of California’s waiver provisions contributed to its political durability. This jump-start lent prompt legitimacy to the foundational CAA statutes and would routinely be invoked at subsequent intervals to deflect concerns about the feasibility or cost of further technological expansion. But the underlying design of the policy kept open the possibility of sustained modification and expansion, either through administrative adjustments or through subsequent rounds of legislation. Both paths were followed with great frequency in the ensuing decades, allowing for a remarkable durability of the policy framework alongside considerable capacity for adaptability.

4.5 when waivers work: california’s enduring leadership role Subsequent decades provided a vital test of the viability of a leveraged federalism system through a structured waiver process. This offered an unusual intergovernmental mechanism whereby one state might not only influence federal policy in its formation but also over an extended period and in ways that were difficult to envision at the outset. It was the key design feature in fostering state-driven adaptability. As political scientist Paul Posner demonstrated, there is ample precedent for vertical policy diffusion in energy and environmental policy, whereby the actions of one or more states ultimately compel a federal policy response.47 This response may include some form of preemption while possibly modeling that preemption on best practice from the states. Vertical diffusion often follows some form of horizontal diffusion, under which policy innovation in a single state is replicated by other states, ultimately diffusing across regions. There is also ample precedent for such horizontal diffusion to create a “tipping point,” whereby there is so much state engagement and possible variation from state to state that federal political economy concerns emerge and prompt serious consideration of a federal role that might limit or constrain early state actors.48 This inevitably leads to questions about the subsequent status of those early adopters, including questions of whether they might receive 46 47

48

BRYNER, BLUE SKIES, GREEN POLITICS, supra note 19, at 158. Paul Posner, The Politics of Vertical Diffusion: States and Climate Change, in GREENHOUSE GOVERNANCE: ADDRESSING CLIMATE CHANGE IN AMERICA 73–98 (Barry G. Rabe ed., 2010). Paul Teske, Checks, Balances, and Thresholds: State Regulatory Re-enforcement and Federal Preemption, 38(3) POL. SCI. & POLITICS 367–70 (2005); PAUL TESKE, REGULATION IN THE STATES (2004).

130

Barry G. Rabe

special treatment or credits for their efforts, although this often leads to intergovernmental tension and resentments and broader concerns about equity and fairness within a federation.49 California has routinely emerged at the epicenter of these kinds of diffusion processes, ranging from early academic studies in the field in the 1960s through the mid-2010s.50 It is common in American diffusion studies examining social and economic policy adoption to find California in a lead role across the twentieth century and into the current one, most commonly with counterparts on the opposite coast (such as Massachusetts and New York) in the early stages and with subsequent followers in neighboring states such as Oregon and Washington. But there is no precedent for actions like California’s on motor vehicle emissions to receive permanent single-state access to a clear and enduring federal waiver mechanism spelled out in exact form in statute. This has allowed the state to continue to pursue state policy development that might lead to further diffusion nationally, albeit with a process for federal review before approval. However, there was no certainty on creating the waiver process that California would remain a robust intergovernmental partner. California clearly passed that test with aggressive use of waiver requests that began almost immediately after creation of the process in 1967 and with continued use in the following decades (Table 4.2). Waiver requests can be driven solely by administrative review and resulting requests by the CARB. They can also stem from new California legislation, which commonly leads to a formal CARB waiver request based on that statute. The federal legislation places no limits on the number of such requests that can be made in a particular time period and imposes no clear bounds on the potential topics or approaches that might be addressed as long as they fall under the broader provisions established in federal legislation. Waiver requests then lead to an EPA Notice of Decision or Summary of Determination. Between 1968 and 2017, 126 waiver requests were considered, with at least one in all but three years. Five or more such decisions were made during 1978, 1980, 1982, 1986, 1988, 2006, 2011 and 2012, including a peak of nine in 1978 (see Table 4.2). Sixty-nine of these decisions were made in the twenty-nine years in which EPA and the executive branch were headed by a Republican president (Nixon, Ford, Reagan, Bush I, Bush II and Trump). Fifty-seven were made in the twenty-one years in which they were headed by a Democratic president (Johnson, Carter, Clinton and Obama). The pace of federal decisions on these requests frequently accelerates toward the end of a presidential term, although this may also coincide with other factors, such as timing of federal or state legislation.

49 50

JENNA BEDNAR, THE ROBUST FEDERATION: PRINCIPLES OF DESIGN (2008). Alka Sapat, Devolution and Innovation: The Adoption of State Environmental Policy Innovations by Administrative Agencies, 64(2) PUB. ADMIN. REV. 141–51 (2004); Jack L. Walker Jr, The Diffusion of Innovations among the American States, 63(3) AM. POL. SCI. REV. 880–99 (1969).

Leveraged Federalism and the Clean Air Act table 4.2 EPA Decisions on California Waiver Requests, 1968–2017

Year

Number of accepted waiver requests

Number of denied waiver requests

1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

1 1 0 3 2 0 2 1 1 3 7 2 5 3 5 1 3 3 5 1 8 1 2 0 3 1 2 1 1 0 3 1 2 1 1 2 2 2 5 1

0 0 0 0 0 1 0 1 0 0 2 1 1 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 (continued)

131

Barry G. Rabe

132

table 4.2 (continued)

Year

Number of accepted waiver requests

Number of denied waiver requests

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 Total

1 2 1 5 6 4 2 3 1 4 117

1 0 0 0 0 0 0 0 0 0 9

The pace of decisions may also be influenced by California’s consideration of emission control options that were added in later rounds of federal legislative amendment. For example, urban buses, locomotives and nonroad vehicles were included in federal legislation through the 1990 amendments. Highway motorcycles were added in the 1977 amendments. This triggered a series of separate California requests for each new vehicle category, although these new sources collectively remained a very small portion of the total emissions produced by motor vehicles. This added factor may particularly inflate the frequency with which waiver requests have been made in the last two decades as opposed to earlier periods. The overwhelming pattern in federal decisions was to approve California’s requests and sustain an ongoing process of state policy development. There were frequent differences over particular terms, such as possible delays in timetables for implementation, but the regular outcome was a federal stamp of approval. As a 2007 review of this process by James McCarthy of the Congressional Research Service concluded, EPA has “repeatedly found . . . that California faces compelling and extraordinary conditions and needs its own standards to meet them. EPA has also generally deferred to the state’s judgment regarding consistency with Section 202(a).” This pattern reflects continued adherence to a 1975 statement by EPA that was made in one of the earlier waiver determinations: “These provisions must be read in the light of their unusually detailed and explicit legislative history . . . Congress meant to ensure by the language it adopted that the Federal government would not second-guess the wisdom of state policy here . . . Sponsors of the language eventually adopted referred repeatedly to their intent to make sure that no ‘Federal bureaucrat’ would be able to tell the people of California what auto emission standards

Leveraged Federalism and the Clean Air Act

133

were good for them, as long as they were stricter than Federal standards.”51 This statement is consistent with those found in other waiver approvals, including earlier affirmations of California’s unique authority by founding EPA Administrator William Ruckelshaus.52 This approval process was not seamless. Of the 126 waiver requests submitted between 1967 and 2017, nine requests have been denied. However, four of these requests were denied because the Administrator determined that a waiver was not required in order for California to enforce its new regulations. Furthermore, one was denied for restrictions based on the originally requested model year but was accepted under the following model year. Of the final four, two received partial denials because some proposed portions were accepted, whereas others failed to meet the requirements of the CAA preemption provision. Another had been approved previously, but the decision was brought before the court and ruled illegal, so the Administrator needed to revoke approval. All in all, only one was officially denied based on the failure of California to meet the necessary requirements for waiver acceptance.53 This suggests a continued pattern of federal approval of California waiver requests over a period of nearly a half-century. It reflects ongoing EPA acceptance of the waiver process and the terms for approval set forth in legislation. But it also indicates a remarkable process whereby a single state not only made an early foray into a policy arena but has also managed to sustain a national leadership role over multiple decades, despite numerous election cycles and changes in partisan control of state government institutions (see Table 4.1). Consequently, California’s early commitments in the 1950s and 1960s to race to the top in establishing air quality standards for motor vehicles has not flagged in subsequent decades and is reflected in active use of the waiver request mechanism and a willingness to repeatedly push the bounds of policy. As we shall see, this has included ventures into issues not on the agenda when the waiver process was created, including requests to create entirely new categories of low- through no-emission vehicles and to add carbon dioxide to the list of controlled vehicle emissions. The federal government signaled continuing support for California’s efforts to continually push the envelope on what might be possible technically to reduce vehicle emissions. During the 1990s, this took the form of CAA amendments as well as a series of waivers that embraced intensifying efforts in California to build on earlier successes. While many leading analyses of the 1990 amendments tend to focus on the politics surrounding industrial source emissions and the expansion of emissions trading for sulfur dioxide releases, mobile sources also received 51

52 53

40 Fed. Reg. § 23,103 (1975); quoted and examined in MCCARTHY, CALIFORNIA’S WAIVER REQUEST, supra note 39, at 10–11. 49 Fed. Reg. § 23,103 (1975). EMILY UPTON & THOMAS VAN HEEKE, FEDERALISM AND CALIFORNIA’S ROLE IN U.S. VEHICLE EMISSIONS STANDARDS (2017).

134

Barry G. Rabe

considerable scrutiny, with many new federal steps directly linked to previous ones taken by California. The legislation established or tightened “more than ninety emission standards for motor vehicles,” including reductions of 60 percent of nitrogen oxides and 35 percent of hydrocarbons for all new vehicles sold by model year 1996. In turn, auto manufacturers were required to produce fleets of experimental cars for sale in California, beginning with 150,000 vehicles by 1996 and then expanding to 300,000 by 1998.54 California paved the way for these steps, including 1988 legislation and CARB regulations two years later based on new vehicle categorizations. These reflected the significant improvements in vehicle technology achieved in prior decades and a strategic decision to divide vehicles into four separate categories to propel improvements in each area. Each of these new categories featured separate emission standards with intensifying degrees of stringency. These so-called low-emission vehicle (LEV) standards were known as LEV I in their first round and then were expanded in 1998 under LEV II. In the latter program, light-duty trucks would join passenger cars, and nitrogen oxide standards would be tightened significantly. California established four separate categories within the LEV I framework: Transitional Low-Emissions Vehicles (TLEVs), Low-Emissions Vehicles (LEVs), Ultra-Low-Emissions Vehicles (ULEVs), and Zero-Emissions Vehicles (ZEVs). The TLEV category was phased out under LEV II and replaced by a Super-LowEmitting Vehicle (SLEV) category. Federal embrace of California’s efforts led to significant transitions in many areas of vehicle development. However, the ZEV program proved quite difficult to implement, requiring such adjustments as softening purchase targets and including hybrids for partial credit under the zeroemissions umbrella.55 The Bandwagon Option. The LEV and ZEV experience also demonstrated the possibility that other states might further propel adaptability in vehicle emissions. Despite the singular delegation of waiver authority to California, the 1977 CAA amendments created a second-stage state adoption process. This allows any state to seek EPA approval to replicate California’s standards once its waiver requests are approved. This provision was retained in the 1990 amendments, albeit with the added stipulation that no state could adopt more stringent standards than those established by California. As a result, alongside the option that the federal government might ultimately adopt California standards on a national basis emerged the possibility that one or more states or even a cluster of regional states might adopt them. Northeastern states were particularly interested in the LEV and ZEV approaches. This reflected their own intensive population concentrations and enduring air 54

55

BRYNER, BLUE SKIES, GREEN POLITICS, supra note 19, at 147–48; Weinbaum, Implications of Automotive Emission Restrictions in California, supra note 30, at 40. EISINGER, SMOG CHECK, supra note 18, at 27, 37.

Leveraged Federalism and the Clean Air Act

135

quality concerns linked to motor vehicle use as well as their envy over seeing California frequently get the first crack at purchasing vehicles deploying new technologies. This region was the most active in exploring possible state legislation during federal preemption and California exceptionalism in the 1960s and frequently sought to adopt California standards once they were approved federally. This established the possibility of a regulatory “bandwagon effect,” one where individual states might join California, gain the benefits from engagement and also possibly further tip the federal government into national adoption of California standards.56 While Massachusetts and New York were particularly vocal and active in this area, the northeastern states tended to work collaboratively on a regional basis in air quality. This reflected the rather small territory represented by many such states and the penchant for substantial cross-border air pollutant migration. Such regional organizations as the Mid-Atlantic Pollution Control Compact, the Mid-Atlantic Regional Air Management Association and the Northeast States for Coordinated Air Use Management (NESCAUM) led this coordination effort. They routinely encouraged the federal government to treat northeastern states with greater deference given their particular air quality problems and interest in following the path set by California.57 However, these organizations had little binding authority, focusing primarily on research, interstate collaboration and advocacy for consideration of regional needs at the federal level. Congress expanded and formalized regional authority in the 1990 CAA amendments through the creation of the Northeast Ozone Transport Commission (OTC). The OTC was not given formal regulatory authority but received a charge to provide EPA with advice on how to improve regional air quality. The OTC included twelve states and the District of Columbia.58 Within a year of its creation, the OTC tested its wings with the adoption of a memorandum of understanding that endorsed California’s LEV program and sought EPA permission to adopt this program. Some OTC states, led by New York and Massachusetts, adopted California LEV and ZEV requirements for themselves. They petitioned EPA for regional approval, and this was authorized in 1995, requiring all OTC states and the District of Columbia to comply with California standards. However, OTC members were not unanimous in their views, and four states voted against the memorandum of understanding and opposed emulation of California.59 Virginia was particularly outspoken 56

57

58

59

On the bandwagon effect in the case of international relations, see STEPHEN M. WALT, Keeping the World “Off-Balance” in AMERICA UNRIVALED (G. John Ikenberry ed., 2002). JOSEPH P. ZIMMERMAN & DEIRDRE A. ZIMMERMAN EDS., THE POLITICS OF SUBNATIONAL GOVERNANCE (1992); BRYNER, BLUE SKIES, GREEN POLITICS, supra note 19, at 221, 234. OTC state members include Connecticut, Delaware, Maine, Maryland, Massachusetts, New Hampshire, New Jersey, New York, Pennsylvania, Rhode Island, Vermont and Virginia. For an extended review of this intergovernmental process, see BRYNER, BLUE SKIES, GREEN POLITICS, supra note 19, at 220–26.

136

Barry G. Rabe

on this issue and challenged the EPA decision in federal court. In 1997, the DC Circuit Court ruled in Commonwealth of Virginia v. EPA that the federal agency could not impose the California standards on the entire OTC region.60 The federal government attempted to resolve this issue through the establishment of a voluntary program in 1998. Known as the National Low Emission Vehicle program (NLEV), NLEV allowed individual states and manufacturers to adopt standards that were somewhat more stringent than federal ones while similar to California LEV standards. Nine states and twenty-three manufacturers opted for this approach, whereas other OTC states chose to adopt standards identical to California’s.61 By 1999, thirteen states and the District of Columbia, representing approximately one-third of the national vehicle market, had adopted emission standards that were nearly identical to those required by California. However, the NLEV process was subsequently overtaken by federal adoption of a fleetwide approach that adhered to California’s LEV standards. Subsequent rounds of California LEV standards, known as LEV II and LEV III, have not only received federal waivers but also have been adopted on a national basis after other states also sought to emulate California in the interim. The LEV III standards, now in parallel with federal Tier 3 standards, include continuous fleetwide emission reductions through 2025, zero evaporative emissions and a combined hydrocarbon and nitrogen oxide standard. The ZEV program has also evolved, albeit without a federal application. The current California program requires that 15 percent of new car sales in the state be ZEVs, and nine other states have either adopted or expressed intent to adopt this program.62 This case demonstrates the added complexity of opening up supplemental state engagement opportunities following approval of a California waiver. It confirms the interest by multiple states in following California in cases where the federal government has not established a national standard or has adopted one weaker than California’s. It also underscores the challenges of building a broader coalition across states, reflecting significant political and policy differences even in a case where there are common concerns about air quality and some broad desire for regional collaboration.63 The possibility of other states, particularly those from the Northeast, creating an alliance with California on motor vehicle air quality was further illustrated by a subsequent case that would prove transformational. 60 61 62 63

Commonwealth of Virginia v. EPA, 108 F.3d 1397 (D.C. Cir.1997). Sheridan, Emission Standards for Mobile Sources, supra note 22, at 9. Vogel, in California Greenin’, supra note 11, at 214–215 For a more extensive analysis of the LEV and ZEV programs, see Carlson, Iterative Federalism, supra note 24, at 1119–25, 1138.

Leveraged Federalism and the Clean Air Act

137

4.6 the next mobile source frontier: expanding the definition of air pollutants and the judicial role in resolving intergovernmental conflict Leveraged federalism proved remarkably durable and adaptable over its first thirtyfive years of operation for mobile source emissions. This reflected core design elements that endured subsequent rounds of state and federal legislative amendment, with adaptation made possible through regular use of the California waiver option. There were no further legislative revisions to the CAA after the 1990 amendments, although this did not mean that the mobile source program drifted into complacency. Instead, a new California effort to pursue a transformational expansion of the definition of air pollutants that could be covered under its waiver provision provided an unprecedented test of its power. This ultimately led to conflict with one federal executive unwilling to grant the requested waiver and an ensuing battle in the courts that resulted in a landmark case. All of this culminated in a historic agreement reached by the succeeding federal administration that embraced the original California waiver request while also making major changes to the federal program. This new agreement, however, is in jeopardy under the Trump administration. This challenge to the prevailing policy order reflected the mounting concern over climate change and the contribution made to that problem by transportation-sector carbon dioxide emissions. This was not a new issue because both California and the federal government had recognized this problem and taken preliminary steps to develop policy options even before adoption of the 1990 CAA amendments. However, climate change reached particular salience in California during the 1990s and early 2000s due to significant localized signs of climate change consequences, sobering concerns over longer-term threats, and abiding California interest in once again taking the lead in policy development and federal government inaction. By 2002, California adopted a number of policies focused in part on the goal of reducing GHG emissions.64 These included renewable energy and energy efficiency mandates as well as investments in research and state policy capacity.65 But the recognition that more than 40 percent of the state’s GHG emissions came from motor vehicles led to questions about how to achieve significant reductions. One possible step was to mandate increased vehicle fuel economy, given the direct link between fuel efficiency performance and carbon dioxide releases from fuel combustion. But fuel economy represented an area of policy where federal preemption had been established firmly in 1975 and included no exceptions for state waivers.66 64

65 66

On the challenges of sustaining regional environmental governance strategies, see Barry G. Rabe, The Durability of Carbon Cap-and-Trade Policy, 29(1) GOVERNANCE 103–19 (2016); MARTHA DERTHICK, BETWEEN STATE AND NATION: REGIONAL ORGANIZATIONS OF THE UNITED STATES (1976). KARAPIN, POLITICAL OPPORTUNITIES FOR CLIMATE POLICY, supra note 23, at ch. 7. BARRY G. RABE, Governing the Climate from Sacramento in UNLOCKING THE POWER OF NETWORKS: KEYS TO HIGH-PERFORMANCE GOVERNMENT (Stephen Goldsmith & Donald F Kettl eds., 2009).

138

Barry G. Rabe

Instead, California legislators focused on reducing carbon emissions from vehicle tailpipes, where the state had a potential role if CAA waiver requests were approved. An air quality standard could be translated into comparable fuel economy performance but not formally encroach on that preempted policy territory. This led to the adoption of AB 1493 in July 2002. The legislation was drafted by an environmental group, the Bluewater Network, and sponsored by Democratic assemblywoman Fran Pavley, a first-term state assembly member and former civics teacher. It was backed by a coalition including environmental groups, Silicon Valley and other corporate leaders and California municipalities. The legislation also received support from entertainment celebrities, former President Bill Clinton and international leaders, many of whom attended multiple signing ceremonies led by Democratic Governor Gray Davis.67 But nothing about this bill was straightforward, including its adoption. Environmental groups and vehicle manufacturers battled “in a clash reminiscent of their first collisions more than three decades ago.”68 More recent patterns of broad bipartisan support for tighter vehicle emission standards evaporated into intense and highly partisan debates in both legislative chambers. In the Senate, Democrats supported the bill by a 23-to-3 margin, but all sixteen Republicans were opposed. In the House, Republican opposition was unanimous, but strong Democratic majorities prevailed.69 “This bill represents the worst form of environmental extremism,” said Republican Assembly Leader David Cox in leading opposition.70 Governor Davis did not signal his views on the legislation until after passage, when he offered his belated support. Shortly after California adopted AB 1493, a number of the same northeastern states active in OTC on vehicle emissions signaled interest in launching a multistate bandwagon effort through their own emulation. In 2003, for example, New York’s Republican Governor George Pataki said in his State of the State Address, “Let’s work to reduce greenhouse gases by adopting the carbon dioxide emission standards for motor vehicles which were recently proposed by the State of California.”71 The adoption of AB 1493 began a prolonged path to implementation, one that took nearly a decade. The legislation launched a process whereby CARB was assigned to develop regulations by January 2005 that achieved “the maximum feasible, cost-effective and technologically achievable reduction of greenhouse gas pollution emitted by new passenger vehicles” by 2005.72 These regulations went into 67

68

69 70

71 72

Portions of this section are drawn from BARRY G. RABE, STATEHOUSE AND GREENHOUSE: THE EMERGING POLITICS OF AMERICAN CLIMATE CHANGE POLICY ch. 4 (2004). Carl Ingram, Senate Votes to Require Cleaner-Running Cars, Light Trucks, L. A. TIMES, May 3, 2002, http://articles.latimes.com/2002/may/03/local/me-exhaust3. Weinbaum, Implications of Automotive Emission Restrictions in California, supra note 30, at 43. Gary Polakavic & Miguel Bustillo, Assembly Passes Bill to Control Emissions of Greenhouse Gases, L. A. TIMES, Jan. 31, 2002, http://articles.latimes.com/2002/jan/31/local/me-green31. RABE, STATEHOUSE AND GREENHOUSE, supra note 68, at 141–44. For fuller discussion, see Weinbaum, Implications of Automotive Emission Restrictions in California, supra note 30, at 37.

Leveraged Federalism and the Clean Air Act

139

effect one year later, after a period of legislative review and amendment. Vehicle manufacturers were expected to implement changes for their 2009 models, although they could secure credits for any emission reductions attained between 2000 and the beginning of the 2009 model year. The legislation, however, provided only limited detail to guide CARB in crafting the regulations. It did not establish any specific emission reduction targets or goals, much less guidance on how reductions might be achieved. However, the legislation did prohibit CARB from considering several possible steps toward compliance, including imposing new taxes on vehicles or fuel, reducing speed limits, banning certain types of vehicles, mandating any reduction in vehicle weight or establishing any restrictions on driving behavior. Despite a partisan shift in the California governorship within one year of AB 1493 adoption, CARB remained on pace to produce initial GHG emission standards for vehicles for 2009–13, with a second and more stringent set of standards to apply for 2013–16. The recall of Governor Davis and his replacement with Republican Arnold Schwarzenegger did not in any way impair CARB’s latitude or weaken political support for a waiver request to EPA that was necessary to allow this to move forward. However, the George W. Bush administration’s EPA had rejected previous petitions from environmental groups that it use its CAA authority to set GHG emission standards for vehicles, saying it lacked authority to apply those provisions to GHGs. The Agency also contended that climate science uncertainties about the human contributions to climate change legitimized their decision to delay any regulatory actions until greater certainty was achieved. In 2007, the US Supreme Court offered a strong response, particularly on the latter point, in a five-to-four decision in Massachusetts v. EPA. As legal scholar Kirsten Engel noted, “the Court used its bully pulpit to undercut the argument of scientific uncertainty by showing respect for the work and conclusions of climate scientists.”73 This decision meant that EPA possessed the authority under the CAA to regulate GHG emissions because they fell within the legislation’s definition of an air pollutant, but it did not force the agency’s hand in addressing the question of whether to grant California a waiver request to regulate vehicle GHG emissions. The Bush administration completed its full second term without ever designating that GHG emissions from motor vehicles would trigger an endangerment finding and thus rejected the state’s waiver request. This set the stage for a dramatic federal reversal in the early months of the Barack Obama presidency. An April 2009 endangerment finding was followed within weeks by a historic Rose Garden announcement of a far-reaching agreement to accept California’s waiver proposal and weave it into a major new policy to address GHG emissions from motor vehicles. Under this agreement, the federal government would integrate GHG emissions from motor vehicles with Corporate Average Fuel Economy (CAFE) standards through an inter-agency partnership between 73

KIRSTEN ENGEL, Courts and Climate Policy in GREENHOUSE GOVERNANCE 241–42 (2010).

140

Barry G. Rabe

EPA and the National Highway Traffic Safety Administration (NHTSA) of the US Department of Transportation. These units had never previously collaborated in this way, and the CAFE program had a considerably more troubled performance record, raising inevitable questions on how this partnership would fare. Nonetheless, an agreement was reached between these federal units, the state of California and major vehicle manufacturers. This was reflected in a White House ceremony featuring environmental supporters, Governor Schwarzenegger, Massachusetts Governor Deval Patrick, Michigan Governor Jennifer Granholm and auto industry leaders. It occurred at a point when the federal government had maximum leverage over Michigan officials and American auto manufacturers such as General Motors and Chrysler because the federal government was assuming a majority stake in them in response to their severe fiscal woes.74 The resulting multiagency agreements were finalized in April 2010, requiring vehicles such as passenger cars, minivans, sport-utility vehicles (SUVs) and other light trucks to meet combined emissions levels averaging 250 grams per mile of carbon dioxide for model year 2016, comparable to a 30 percent reduction in emissions from 2010 levels. This would be commensurate with CAFE standards set at 34.1 miles per gallon by 2016. The standards were phased in beginning in 2012 and were intended by 2025 to reduce vehicle emissions by about 50 percent during that period and exceed 50 miles per gallon, although this was to be revisited through a major midterm review.75 Additional standards have moved forward for other types of vehicles, including provisions for heavy-duty trucks approved in August 2016, although these have generally lagged behind the far larger category (measured in total vehicles and emissions) of cars and light trucks. This represents one area where the federal government has frequently led rather than California, in part due to the fact that “many heavy-duty trucks travel across state lines and are registered in states other than California.”76 California, however, pursued a series of waivers related to heavyduty vehicles during the 1980s and has intensified its focus on this category in more recent years as it has increasingly come to view heavy-duty vehicle emissions reduction as a significant component in achieving its overall GHG emission goals. This reflects the growing impact of heavy-duty vehicles on total emissions in the state, given emission reductions in other sectors amid continued heavy reliance on heavy-duty truck fleets that can take multiple decades to turn over.

74 75

76

Barry G. Rabe, Contested Federalism and American Climate Policy, 41(3) PUBLIUS 494–521 (2011). JAMES E. MCCARTHY & BRENT YACOBUCCI, CARS, TRUCKS, AND CLIMATE: EPA REGULATION OF GREENHOUSE GASES FROM MOBILE SOURCES (2014), https://www.fas.org/sgp/crs/misc/R40506.pdf; JAMES E. MCCARTHY, JONATHAN L. RAMSEUR, JANE A. LEGGETT, ALEXANDRA M. WYATT & ALISSA M. DOLAN, EPA’S CLEAN POWER PLAN FOR EXISTING POWER PLANTS: FREQUENTLY ASKED QUESTIONS (2016), https://www.fas.org/sgp/crs/misc/R44341.pdf. UPTON & THOMAS VAN HEEKE, FEDERALISM AND CALIFORNIA’S ROLE IN U.S. VEHICLE EMISSIONS STANDARDS, supra note 53.

Leveraged Federalism and the Clean Air Act

141

4.7 environmental impacts Scholarly analysis of the motor vehicle emissions control program has been consistently enthusiastic about the performance of the program over its multiple decades of operation. This reflects the sustained pattern of reducing such major emissions as hydrocarbons, carbon monoxide and nitrogen oxides from newly minted vehicles. Even before the expanded reduction goals set forth in the 2010 California-EPA-NHTSA agreement that runs through 2025, this is reflected in vehicle emission patterns that have shifted dramatically in recent decades. Table 4.3 demonstrates the dialectic between state and federal standards over time, through the point at which California adopted AB 1493 and launched a new focus on GHG emissions. This demonstrates that California has often acted first, usually through a federally approved waiver. But this state-specific standard is often reflected in subsequent years on a federal basis, indicating either congressional or EPA emulation of the steps previously taken in Sacramento. These reductions were most dramatic during the first two decades of the program, reflecting the advent of the catalytic converter and related technologies that produced fundamental changes in engine function. But they have been sustained over multiple decades and are expected to continue as the new blending of vehicle emission reductions with fuel economy improvements moves into advanced stages of implementation. Emerging scholarship that examines vehicle emission policy performance cuts across disciplinary lines, reaching similar and mutually supportive conclusions. Legal scholar Ann Carlson noted that “it is worth highlighting the extraordinary success the state and the federal government have had in dramatically reducing tail pipe emissions.”77 Political scientist Gary Bryner chronicled the ongoing pattern of progress, reflected in both the significant emission declines through 1990 and the subsequent wave of expanded reduction commitments.78 An influential paper published by a team of economists at the University of California, Davis, concluded that average vehicles sold in the 2000s were between 90 to 99 percent cleaner in terms of lower emissions than comparison vehicles produced in the 1960s prior to the advent of regulation.79 This study concluded that emissions improvement occurred almost exclusively because of persistent and aggressive government regulation. Market factors and consumer behavior played almost no role.”80 In turn, there was no definitive scholarship through the first decade of the 2000s that found the added price of new vehicles from refined emission control technology to be a significant deterrent to new purchases. Daniel Sperling and colleagues 77 78 79

80

Carlson, Iterative Federalism, supra note 24, at 1125. BRYNER, BLUE SKIES, GREEN POLITICS, supra note 19, at 158, 198 DANIEL SPERLING, DAVID BUNCH, ANDREW BURKE, ET AL., ANALYSIS OF AUTO INDUSTRY AND CONSUMER RESPONSE TO REGULATIONS AND TECHNOLOGICAL CHANGE, AND CUSTOMIZATION OF CONSUMER RESPONSE MODELS IN SUPPORT OF AB 1493 RULEMAKING 3 (2004), http://www.arb.ca.gov/research/ apr/past/02–310_part6chen.pdf. Id. at 24.

Barry G. Rabe

142

table 4.3 US Federal and California Light-Duty Vehicle Emission Standards, 1966–2010 Federal Year 1966 1968 1970 1971 1972 1973 1974 1975 1977 1978 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1993 1994

HC

CO

6.3 4.1

51.0 34.0

3.0

NOx Evap. PM

28.0

2007

HC

CO

6.3

51.0

4.1

34.0

2.9

34.0

0.9 0.4

9.0

NOx

Evap.

PM

6.0 4.0 3.0

3.0

3.0 2.0 1.5

15.0

3.1 2.0

2.0

1.5

6.0 0.4

7.0 3.4

0.39 1.0

7.0 2.0

1.0 0.7 0.4

6.0 2.0

0.6 0.4 0.2 0.08

0.25 (0.31)

3.4 0.4 (4.20) (0.6)

1995 1996 2003 2004

California

0.08 (0.10) 2 (0.05)

0.125 (0.156)