Home » Articles » Volume 44 » Issue 2 » Have Judges Gone Wild? Plaintiffs’ Choices and Success Rates in Litigation Against Federal Administrative Agencies


Have Judges Gone Wild? Plaintiffs’ Choices and Success Rates in Litigation Against Federal Administrative Agencies


Topics: ,



Our colleague, Professor Joe Feller of the Sandra Day O’Connor College of Law at Arizona State University was well-known and well-regarded in the community of environmental and natural resources law professors. He stood out for the quality of his scholarship, as well as his active and helpful participation in our community and his community at large. Tragically, a car struck and killed Professor Feller on the evening of April 8, 2013, while he was returning home from work.

Joe had been the inspiration for a natural resources law teachers conference scheduled for the following month in Flagstaff, Arizona. Following that conference, Professors Appel and Squillace conversed and agreed that one of many suitable tributes to Joe would be to arrange to publish one of the last articles that Joe drafted. The subject was wilderness. Professor Squillace received the Article in draft with Joe’s request that Squillace make comments on it. Professor Appel—who had never met Joe in person but had spoken to him on the phone and had exchanged emails several times with him—received the Article in draft because the Article directly and pointedly critiques some of Appel’s own work. Despite his near-complete rejection of Appel’s arguments, Joe nevertheless suggested that, as the Article matured with revision, he and Appel could be co-authors on the piece. Appel responded in a lengthy email, which is reproduced after the main body of the Article. Unfortunately, that collaboration or friendly debate never occurred. Nevertheless, the offer to co-author the Article is a striking testament to Joe’s generosity, willingness to cooperate, and his focused dedication to get the answer to a problem right. As those who have delved in the law know, getting the right answer is not as easy as it sounds.

As the reader will see, this Article involves important issues regarding wilderness specifically and federal environmental litigation generally. Despite the fact that it does not represent Joe’s finished thoughts on the subject, we nevertheless believe that this Symposium issue of Environmental Law offers an ideal forum for this fine piece of scholarship. With the generous agreement of the journal’s editors, as well as the permission of Joe’s family and his community at ASU Law, we present this Article to join its rightful place in the legal literature about wilderness.

A note on editing: Because we wanted this Article to represent Joe’s views and not our own, we have taken a very light touch in editing this Article. We have taken the liberty to correct typographical errors and citations to conform to the Bluebook and to Environmental Law’s style guidelines. Where we believed that a thought needed expansion or clarification, we have done so and noted the emendation in square brackets. We tried to channel Joe as best as we could to keep these thoughts in the spirit of Joe’s thinking, not our own.

In closing, we wish to express deepest gratitude once again to Professor Joe Feller’s family for granting us permission to reproduce Professor Feller’s Article posthumously. We hope that this serves as a fitting tribute and memorial to the life of a great scholar, passionate advocate, good friend, and all-around mensch.

I. Introduction

In March of 2010, there appeared in the Stanford Environmental Law Journal an empirical study of litigation between private parties and the federal government over management and use of federal wilderness areas. Entitled Wilderness and the Courts,[1] the study included a few different types of cases: judicial review of administrative decisions by the federal agencies that manage wilderness areas,[2] tort and takings claims against these same agencies arising from such management, and criminal prosecutions for violations of the Wilderness Act of 1964[3] and related statutes. While the results of the study may have warmed the hearts—or confirmed the views—of wilderness advocates and other environmentalists, they were likely alarming to wilderness opponents and to students of the relationship between administrative agencies, courts, and the body politic.

The study, by Professor Peter Appel of the University of Georgia, classified wilderness related cases into two categories.[4] In the first category, which Professor Appel called suits seeking “more protection,” were “cases in which the challenge [against a federal agency] was brought by an environmental organization seeking greater protection for fewer uses within a wilderness area.”[5] In the second category, labeled “less protection” suits, were “cases in which a plaintiff sought more uses within a wilderness area or more protection for private rights within a wilderness area.”[6] The study revealed a striking imbalance between the courts’ reception of these two types of suits: Environmentalists won about half (52%) of their cases against the federal government, while those asserting private rights or more freedom to use wilderness areas won only about one in seven (14%) of their cases.[7] Professor Appel conceded that several alternative explanations of the data were possible and called for further research,[8] but stated:

The foregoing evidence suggests that, in the context of protecting wilderness areas, courts are not acting as much like courts as one would predict from the doctrines of deference that apply in these cases. An implication of this suggestion is that, in this context, judges behave more like policy makers than neutral arbiters.[9]

Professor Appel was not the first to use tallies of victories and defeats to divine the influence of judges’ ideologies on their decisions, or to measure the degree of judicial deference—or lack thereof—to the expertise and policy judgments of administrative agencies, or to reach conclusions about the extent to which such agencies are, or are not, conforming their decisions to the dictates of the law. Several studies have examined the success rate of court challenges to agency decisions as a function of several variables, including, among other things: time; whether the agency decision reflected a new or old, or consistent or inconsistent interpretation of a statute; and the substantive content of the agency decisions.[10] The results of these studies have been presented as evidence for or against temporal trends in agency compliance with statutory requirements or in the degree of judicial deference to agencies,[11] for or against various models of judicial review of agency action,[12] and for or against ideological biases on the part of judges.[13]

Unfortunately, some of these studies suffer from a methodological flaw that renders them at best suspect and at worst meaningless. That flaw is the implicit assumption that the cases that are presented to the courts for decision represent an unbiased sample of federal administrative agency actions.[14] Under this implicit assumption, for example, a decrease over time in the success rate of anti-agency litigation is taken as evidence that agencies are increasingly compliant with the law or that courts are increasingly deferential to agencies.[15] Under the same implicit assumption, a difference in success rate depending on the substantive content of the decisions under review is taken as evidence of judicial bias.

In fact, however, courts see a highly selected sample of federal agency decisions. The selection is performed not by the courts themselves but by the affected parties who choose whether or not to seek review of the decisions. This selection is likely to have a substantial, and in some cases overwhelming, effect on the success rate of such litigation. For example, if the number of affected parties who are ready, willing, and able to bring agencies to court changes over time, then the success rate of such litigation will also change over time. More specifically, the more parties there are with the resources and motivation to file lawsuits challenging agency action, the more likely it may be that such lawsuits will be filed even when the prospects for success are limited. And the more low-odds challenges to agency actions are brought to court, the higher will be the affirmance-reversal ratio. Similarly, if parties on one side of an ideological divide are more likely to litigate than are their ideological opponents, then a differential success rate depending on the substantive content of agency decisions may simply reflect the greater tendency of the former than the latter to seek judicial review of adverse agency decisions even when their chance of success is uncertain or unlikely.

In the area of private litigation seeking monetary damages—e.g., torts or contract disputes—substantial empirical and theoretical work suggests that the selection by plaintiffs—or potential plaintiffs—of which cases to take to trial and which to settle is the dominant factor influencing the plaintiffs’ success rate in those cases that do proceed to trial.[16] In other words, the ratio of plaintiffs’ verdicts to defendants’ verdicts tells us more about plaintiffs’ strategic choices than it does about the substance of legal rules, the attitudes or biases of judges or juries, or the propensity of Americans to commit torts or breach contracts. Similar research in criminal law reveals that the conviction-acquittal ratio in criminal trials is largely determined by prosecutors’ choices as to which cases to try, which cases to plea bargain, and which cases to walk away from.[17]

Surprisingly, empirical studies of judicial review of administrative agency actions have largely overlooked the influence of parties’ selection of which agency actions to challenge and which to let stand, either ignoring this factor altogether or treating it as an aside.[18] However, as will be elaborated in this Article, there is no reason to believe that such selection is any less at work in judicial review of agency actions than it is in private damages litigation or criminal prosecutions, and there is some reason to believe it is even more so.

This Article has two objectives. First, I would like to show that the suggestion that courts may be biased in favor of more protection of wilderness areas might not be justified; Professor Appel’s data can easily be explained by the differing selection mechanisms at work in the two types of cases he studied. Second, I hope to encourage others to further investigate the selective forces at work in other areas of judicial review of administrative action. Administrative action encompasses such a broad range of subjects and interests that it is unlikely that a single model can encompass all of them. Some administrative agencies, such as the Social Security Administration, award or deny monetary benefits to individuals,[19] and it may well be that the dynamics driving parties to accept or seek review of such decisions are the same as those driving parties in private damages litigation. Other agencies, such as those managing federal wilderness areas that were the object of Professor Appel’s investigation, make decisions that may be of concern to thousands or even millions of people, yet have direct economic impact on only a few, or even no, individuals.[20] The dynamics driving individuals or organizations to seek review of these agencies’ decisions are likely much different than for agencies disbursing economic benefits and are likely to be highly asymmetric depending on the nature of the interests involved.

Part II of this Article briefly reviews several empirical studies of judicial review of administrative agency action that have been published over the last three decades and then describes in greater detail the most recently published study, that of Professor Appel concerning wilderness litigation.[21] Part III reviews the leading quantitative model of private damages litigation, that of Priest and Klein, that has attempted to show how plaintiffs’ selection of which cases to litigate determines their success rate in those cases that do proceed to judgment.[22] While the Priest-Klein model cannot be directly applied to nonmonetary litigation, I extract from this model the qualitative dynamics that can be so applied and suggest a very crude variation of the model that gives some indication of how public interest organizations’ choices may determine their success rate in challenging administrative agency decisions.

Part IV argues that criminal prosecution, where prosecutors must allocate limited resources by selecting which cases to prosecute among a large universe of potential cases, is, at least in some circumstances, a better analogue to public interest organizations’ judicial challenges to administrative agencies’ decisions than is private damages litigation. In criminal law, the legal burden of proof—beyond a reasonable doubt—is heavily weighted against the prosecution, yet prosecutors win upwards of 80% of their cases,[23] demonstrating either 1) that judges and juries are tacitly ignoring the legal burden of proof, 2) that Americans are shockingly likely to commit crimes, or 3) that case selection by prosecutors significantly, and likely overwhelmingly, influences their success rate. While the first and second explanations cannot be ruled out, surely the third factor must be an important one. Prosecutors do not prosecute randomly selected individuals; they prosecute those who they think they can convict. A recently developed quantitative model of prosecutorial decision making helps to explain how prosecutors’ budgets and strategies determine their conviction rate.[24] I argue that a similar selection mechanism can explain why a public interest organization can have a high success rate in challenging federal administrative agencies even if such agencies usually follow the law and even if courts are usually deferential to agency decisions.

Part V revisits Professor Appel’s study of wilderness litigation and examines the nature and interests of the private parties who have opposed the federal government in the cases that the study considered. That examination reveals an enormous difference between the government’s opponents in the “more protection” cases and those in the “less protection” cases. Ninety percent of the “more protection” cases involved national or regional nonprofit environmental organizations with an interest in many issues in many wilderness locations, while the vast majority of the “less protection” cases were brought by individuals or businesses with an interest in a single issue or group of issues in a single wilderness area.[25] I argue that the differences in resources and incentives between these two groups of litigators lead to differing litigation choices that can explain why the former has a substantially greater success rate than the latter. For this reason, Professor Appel’s conclusion that there may be a judicial bias in favor of wilderness protection is unjustified.

Part VI considers the application of the models discussed in Parts III and IV to the cases in Professor Appel’s study. I argue that the “more protection” cases are best modeled by the variation of the Priest-Klein model that I propose in Part III or by the criminal model described in Part IV. The “less protection” cases can also be described by a modified Priest-Klein model, but not by the criminal model. Finally, the last Part considers some of the other empirical studies of judicial review of administrative agency decisions and attempts to identify which are, and which are not, suspect because of their failure to consider the influences of plaintiffs’ choices as to which cases to litigate.

II. Empirical Studies of Judicial Review of Administrative Agency Actions

From the mid-1970s to the present, a series of empirical studies have attempted to measure, and analyze as a function of several variables, the rate of success of parties who have sought to use the courts to overturn decisions of federal administrative agencies.

A. Pre-Chevron Studies (1975–1986)

The first such published study, by Warner W. Gardner, appeared in the Columbia Law Review in 1975 and examined decisions of the United States Supreme Court and federal courts of appeals that were published in the 1974 calendar year.[26] The study found that the Supreme Court affirmed agencies in twelve out of seventeen cases (71%).[27] While acknowledging the small size of the sample, Gardner noted that this affirmance rate “offer[s] confirmation of the intuitions of the bar that the Supreme Court is disposed to accord considerable deference to the procedures and judgments of the administrative agency.”[28] In the same year, Gardner found that the Courts of Appeals affirmed agency decisions in 330 out of 506 reported cases (65%).[29] Despite this much larger sample size, Gardner concluded “little is proved” by the 65% affirmance rate, since “[o]ne would expect that the reviewing court would agree with the initial tribunal more often than not.”[30] In treating judicial review as simply a second look at a sample of agency decisions, Gardner gave no consideration to the possibility that, because the sample was chosen by parties’ decisions to litigate, it was a highly selective, and possibly quite biased, sample.

The next similar study concerned the National Environmental Policy Act of 1969 (NEPA)[31] and was published in 1986 by Paul G. Kent and John A. Pendergrass.[32] Kent and Pendergrass identified 1,067 reported federal cases where plaintiffs had challenged agency decisions on NEPA grounds.[33] Analysis of these cases revealed, among other things, “that plaintiffs have generally been less successful in the 1980’s than the 1970’s” and that environmental organizations and state governments had a higher success rate than business groups, local governments, or property owners.[34] Kent and Pendergrass attributed the overall declining success rate to agencies’ increasing willingness and ability to comply with NEPA as their experience with the statute increased.[35] With regard to the environmental groups’ better performance than others, Kent and Pendergrass gave a nod to the possible of role of differential selectivity: “Environmental groups do very well, yet generally have fewer economic resources available than business groups or certain property owner groups. One possible explanation is that, partly because of these economic limits, environmental groups chose [sic] the cases they litigate very carefully.”[36]

B. Post-Chevron Studies (1990–2006)

The Supreme Court’s 1984 decision in Chevron, USA v. Natural Resources Defense Council,[37] triggered a series of empirical studies of judicial review of decisions by federal administrative agencies. The Chevron case required a reviewing court to defer to an agency’s interpretation of an ambiguous statute so long as that interpretation is “permissible” or “reasonable,” rather than “simply impos[ing the court’s] own construction on the statute.”[38] A number of empirical scholars attempted to ascertain the nature and extent of Chevron’s impact on judicial decisions and the extent to which judges were, or were not, heeding its command.

The first such study was published in 1990 by two Yale law professors, Peter H. Schuck and E. Donald Elliott, one of whom, Elliott, was also serving at the time as General Counsel of the U.S. Environmental Protection Agency (EPA).[39] Schuck and Elliott examined several thousand reported federal cases reviewing the decisions of dozens of different federal administrative agencies.[40] Schuck and Elliott attempted, among other things, to test the accuracy of the widespread belief “that judicial review of administrative action was deferential in the 1960s, that it became more stringent during the ‘hard look’ era in the 1970s, and then settled back to a moderately deferential stance during the 1980s.”[41] They found that, contrary to this belief, their data showed:

a long-term trend toward increasing rates of affirmances by courts in administrative law cases, from 55.1% in 1965, to 60.6% in 1975, to 76.6% in 1984–85. There are corresponding decreases in the rates of reversals and remands. These data contradict the conventional wisdom, which would predict a higher rate of reversals and remands in 1975 than in 1965 or 1984–85 . . . . Based on an analysis of the opinions published in the Federal Reporter, we were prepared to conclude that the conventional wisdom was wrong . . .[42]

Schuck and Elliott concluded that either courts were steadily becoming more deferential to agency decisions or the agencies were becoming better at “conforming to the dictates of the law.”[43] They then went on to discuss why parties continued to challenge administrative agencies in court despite their apparently low chances of success.[44] Schuck and Elliott did not, however, consider the possibility that parties’ decisions about whether to litigate might be determining, rather than responding to, the success rate.

Schuck and Elliott also found a marked difference between agencies in the plaintiffs’ success rate.[45] They conclude that one group of agencies—the Merit Systems Protection Board, the Immigration and Naturalization Service, the Patent and Trademark Office, and “Other Departments”—enjoyed a much higher rate of affirmance by the courts than did another group, comprising health and environmental agencies, other “regulatory” agencies, and the Department of Labor.[46] They infer that the “data suggest (though they certainly do not prove) that agencies may be less likely to be affirmed in cases that involve broad policy questions and multiple parties . . . as opposed to cases that involve only individual litigants.”[47] They briefly mention that the explanation for this differential rate of affirmance may reflect a number of factors, including “different incentives to litigate [and] different kinds of litigation adversaries.”[48] I will attempt to show below that case selection by these agencies’ adversaries is likely to be an overwhelmingly influential explanatory factor.[49]

In 1997, Professor Richard Revesz of New York University became the first to empirically study the impact of judges’ political ideology on their tendency to affirm or overturn decisions of administrative agencies.[50] Looking at review of EPA decisions by the District of Columbia Circuit from the 1970s through the early 1990s, and using the political party of the appointing President as a proxy for the political leaning of each judge on the circuit, Professor Revesz found that, in the 1980s and the 1990s, Republican appointees were significantly more likely than Democratic appointees to overturn EPA decisions at the behest of polluting industries, whereas Democrats were significantly more likely than Republicans to overturn EPA at the behest of environmental organizations.[51]

Looking at a broader range of agencies and courts over a narrower time period, Orin Kerr also found a link between judges’ party affiliations and their willingness to overturn the decisions of administrative agencies.[52] In an examination of the affirmance rate of administrative agencies in decisions published by the federal courts of appeals in 1995 and 1996, Kerr found that judges appointed by Democratic presidents were more likely than Republican appointees to side with applicants for economic entitlement benefits in challenges to the denial of such benefits and with immigrants in immigration appeals.[53] He also found, conversely, that Republican-appointed judges were more likely than Democrats to overturn economic regulatory decisions dealing with commerce, trade, and taxes.[54] Kerr found little support in his data for the “contextual model” that posits that the chances of judicial acceptance of an agency’s construction of a statute depends on whether the construction is consistent or inconsistent with the agency’s previous application of the same statute, how long the agency has adhered to the same construction, and how closely contemporaneous the interpretation is with the passage of the statute.[55] He also found little evidence that rates of affirmance were influenced by judges’ philosophies of statutory construction—the “interpretive model.”[56]

The conclusion of Revesz and Kerr that judges’ propensity to overturn agency decision is significantly influenced by their political ideologies was challenged in a 2001 analysis by University of Akron Professor William Jordan.[57] Jordan, like Revesz, looked exclusively at decisions by the District of Columbia Circuit reviewing actions of EPA.[58] Breaking down decisions by individual legal issues rather than simply overall victory or defeat by the challengers, and separating cases that present “policy[making] opportunities” from those presenting technical or procedural issues,[59] Jordan found little tendency for Democratic appointees to side more with environmentalists than with industries or for Republican appointees to do the opposite.[60]

Weighing in favor of the influence of ideology in judicial review of agency action was a larger study conducted by Professors Thomas J. Miles and Cass R. Sunstein of the University of Chicago and published in 2006.[61] Miles and Sunstein reviewed all published decisions of the federal courts of appeals reviewing actions by EPA or National Labor Relations Board—253 decisions in all.[62] They classified each agency decision under review as “liberal” or “conservative” depending on the nature of the party challenging the decision.[63] Where the plaintiff opposing the agency decision was an industrial corporation or association of such corporations, the decision was classified as “liberal;” agency decisions challenged by public interest organizations or labor unions were classified as “conservative.”[64] Miles and Sunstein found a significant connection between judges’ political affiliations and a differential tendency to affirm liberal or conservative agency decisions.[65] Democratic appointees were more likely to affirm liberal agency decisions than to affirm conservative ones; Republican appointees were more likely to affirm conservative decisions.[66]

C. The Appel Study (2010)

In March 2010, Professor Peter Appel of the University of Georgia published the first empirical study of litigation under the Wilderness Act of 1964[67] and subsequent legislation designating wilderness areas on federal public lands.[68] The Wilderness Act created the National Wilderness Preservation System and prescribed the principles and rules governing the use and management of wilderness areas within that system.[69] The Wilderness Act generally instructs agencies managing wilderness areas to “preserve [their] wilderness character”[70] and specifically prohibits, with some exceptions, roads, commercial enterprises, motor vehicles and other forms of “mechanical transport,” motorized equipment, motorboats, the landing of aircraft, and “structure[s] or installation[s]” within such areas.[71]

Under the Wilderness Act, only Congress can designate particular areas of federal public land for inclusion in the wilderness system.[72] The Wilderness Act designated approximately nine million acres of land in fifty-four units on National Forests in thirteen states as the initial components of the system.[73] The Wilderness Act instructed the Secretary of Agriculture to review additional areas in the National Forests and make recommendations to the President, who in turn was to make recommendations to Congress, as to which areas were suitable for addition to the system.[74] The Wilderness Act also instructed the Secretary of the Interior to review roadless lands in National Parks and National Wildlife Refuges and make recommendations, through the President to Congress, as to their suitability or unsuitability for designation as wilderness.[75] The Federal Land Policy and Management Act of 1976 (FLPMA)[76] extended the wilderness system to include lands managed by the U.S. Bureau of Land Management (BLM) and instructed the Secretary of the Interior to review roadless BLM lands for their suitability for designation as wilderness.[77] Numerous subsequent acts of Congress,[78] acting either pursuant to or in some cases without[79] recommendations by the executive branch, have expanded the National Wilderness Preservation System to its current size of over 109 million acres, comprising 757 individual wilderness areas in forty-four states and Puerto Rico.[80] Most of these 757 units are subject to the same protective statutory restrictions as the first wilderness areas created by the Wilderness Act itself in 1964. However, in designating additions to the system, Congress has occasionally exercised its power to include special provisions altering those restrictions as applied to the newly designated areas.[81]

The 1964 Wilderness Act and subsequent legislation designating additional wilderness areas have generated a slow but steady trickle of litigation, which, after nearly a half-century, has accumulated a substantial body of case law. In his study, Professor Appel searched for all reported federal court decisions referring to the Wilderness Act, and then eliminated cases in which the Wilderness Act was mentioned only incidentally.[82] To avoid double counting, Professor Appel bundled together multiple reported decisions in the same case (such as a reported appeals court decision affirming or reversing a reported trial court decision) unless the decisions addressed different issues.[83] After this winnowing, he was left with a sample of ninety-four “principal” reported cases, or an average of just over two cases per year from 1965 through 2009.[84]

In all of these ninety-four cases a nonfederal party was pitted against the United States or a federal agency, but the cases involved many different substantive issues in a variety of procedural settings. The greatest number, about half, were suits for judicial review of decisions by the four federal land management agencies that administer wilderness areas—National Park Service, U.S. Forest Service, the Bureau of Land Management, and U.S. Fish & Wildlife Service—to undertake, allow, disallow, or limit particular activities in the wilderness areas that they manage.[85] Other types of cases included criminal prosecutions for violations of the Wilderness Act’s prohibitions,[86] tort claims against the federal government for injuries suffered in wilderness areas,[87] and “takings” claims by persons allegedly deprived of property rights by the Wilderness Act’s restrictions.[88]

Professor Appel classified the cases in his sample into two categories.[89] In the first category, which Professor Appel called suits seeking “greater protection,” were “cases in which the challenge was brought by an environmental organization seeking greater protection for or fewer uses within a wilderness area.”[90] In the second category, labeled “less protection” suits, were “cases in which a plaintiff sought more uses within a wilderness area or more protection for private rights within a wilderness area.”[91] This second category included the tort and takings claims as well as the criminal prosecutions, in which the “plaintiff” was not actually a plaintiff but rather a criminal defendant.[92] Of the ninety-four cases in Professor Appel’s sample, fifty were “more protection” cases and forty-four were “less protection” cases.[93]

Professor Appel found that the plaintiffs in “more protection” cases prevailed over the federal government 52% of the time (twenty-six out of fifty cases), while the government lost approximately 14% of the “less protection” cases (six out of forty-four).[94] Another way of looking at the same data is that in the total of ninety-four cases, courts rendered restrictive “pro-wilderness” decisions—i.e., decisions upholding disallowances of wilderness-impairing uses or overturning allowances of such uses—sixty-four times (68% of the total) and rendered permissive “anti-wilderness” decisions—upholding allowances of uses, overturning disallowances of uses, or sustaining tort or takings claims against the government—only thirty times (32% of the total).[95]

Professor Appel finds in these data, evidence for a substantive bias on the part of the courts deciding these cases.[96] He writes:

The foregoing evidence suggests that, in the context of protecting wilderness areas, courts are not acting as much like courts as one would predict from the doctrines of deference that apply in these cases. An implication of this suggestion is that, in this context, judges behave more like policy makers than neutral arbiters.[97]

Professor Appel then goes on to consider several alternative explanations for what he calls this “one-way ratchet” in judicial review of wilderness management decisions.[98] The alternative explanations he considers include the terms of the Wilderness Act itself,[99] widespread support for wilderness protection among politicians as well as the public,[100] a risk-averse judicial tendency to disallow activities that may have irreversible adverse impacts on wilderness values,[101] superior attorneys representing pro-wilderness plaintiffs,[102] and judicial correction of an anti-wilderness bias on the part of the agencies whose decisions are under review.[103] For the most part, Professor Appel rejects these alternative explanations, although, with respect to the agency bias hypothesis, he concludes simply that it is a very difficult hypothesis to test.[104]

One possible explanatory factor that Professor Appel does not discuss, and that the other studies of judicial review of administrative decisions also mostly overlook, is the choices made by private parties as to which cases to litigate and which not. If the pro-wilderness and anti-wilderness litigators are making different types of choices, such a difference might explain the variance in their success rates. Since such choices have been found to be an overwhelming factor influencing the success rates of plaintiffs in private damages litigation and prosecutors in criminal cases, we now turn to studies of these types of litigation.

III. Plaintiffs’ Choices in Private Damages Litigation

Outside the context of judicial review of administrative agencies, scholars of the judicial process have long recognized that plaintiffs’ decisions as to which cases to litigate are a determinative factor in any accounting of the rate of plaintiffs’ success in those cases that are litigated.[105] A significant amount of the research on the dynamics of the selection process has focused on private damages litigation,[106] and the acknowledgements that such selection may also be a factor in judicial review of administrative agencies generally refer to that literature.[107]

A. The Priest-Klein Model

In a pathbreaking article published in 1983, George Priest and Burton Klein developed a model to predict which cases would be tried and which would be settled.[108] The model was built on the simple economic assumption that each party to a lawsuit will seek to maximize its expected monetary return—or minimize its expected loss.[109] For this reason, it is not directly applicable to suits for review of administrative agency decisions where both the plaintiff, often a nonprofit organization, and the defendant, a government agency, are likely to be motivated by factors other than monetary gain or loss.[110] Nonetheless, the model is worth considering here, because some of its qualitative features, though not necessarily its quantitative predictions, are likely to have broader application.

According to the Priest-Klein model, the choice between litigation and settlement depends on whether the defendant is economically motivated to offer an amount in settlement that will be more economically attractive to the plaintiff than a judgment at trial.[111] A plaintiff will proceed to trial when the plaintiff’s expected economic return from a trial exceeds the economic return of a settlement, that is, when:


Ep – Cp > M,


whereEP is the plaintiffs’ observed probability of victory times the expected amount of judgment, CP is the plaintiffs’ expected cost of trial, and M is the maximum amount the defendant is willing to offer in settlement.[112] Where the amount of damages is fixed, and the trial will result in a simple “yes” or “no” decision, Ep will simply be the predetermined amount of damagestimes the plaintiffs’ estimation of his own chance of victory.[113] The maximum amount that a defendant would be willing to offer in settlement would be the defendants’ expected economic loss from a trial:


M = Ed + Cd,


where Ed is the defendants’ expectation value of the judgment—defendants’ estimate of plaintiffs’ chance of victory times amount of damages—and Cd is the defendant’s expected cost of trial.[114] Thus, under the model, a case will proceed to trial only when


Ep – Cp > Ed + Cd




Ep – Ed > Cp + Cd.[115]


In other words, the only cases that will be tried are those in which the parties’ expectations of the outcome of a trial differ by an amount greater than the total cost to the parties of the trial—Cp + Cd. This prediction makes intuitive sense. Where the parties’ expectations of the outcome of a trial are the same, there is no point in having the trial; they will settle for an amount close to their mutual expectation. Where both parties expect the plaintiff to fare poorly at trial, the amount of the settlement will be small; where both parties expect the plaintiff to do well at trial, the settlement amount will be large. Only cases where the parties’ expectations differ substantially are likely to be tried.

Each party’s expected outcome at trial can be expressed as that party’s estimate of the probability of a plaintiff’s victory times the amount that the party stands to gain or lose from such a victory:


Ep =Pp (Jp); Ed = Pd (Jd),


where Pp represents the plaintiff’s estimate of the likelihood that he will succeed at trial, Pd is the defendant’s estimate of the plaintiff’s likelihood of success, Jp is the amount that plaintiff stands to gain from a trial victory, and Jd is the amount that the defendant stands to lose. The parties will proceed to trial only when:


Pp (Jp) – Pd (Jd) > Cp + Cd.


In the simplest case, where only money damages are at stake, Jp = Jd, that is, the amount that the plaintiff stands to gain from a trial victory is the same as the amount that the defendant stands to lose, namely, the amount of damages. As Priest and Klein recognized, however, the parties may have more at stake than just the amount of damages.[116] For example, if the defendant is a corporation facing many potential similar lawsuits, a loss in one lawsuit may set a precedent that will adversely affect the corporation’s position in future cases. Or a loss in say, a product liability case, may adversely affect future sales of that product.

To clarify the dependence of the likelihood of litigation on the shared expectations, as well as the differing expectations, of the parties, Priest and Klein reexpressed Pp, Jp, Pd , and Jd in terms of the averages of, and the differences between, Pp and Pd and Jp and Jd:


P = (Pp + Pd) / 2; ΔP = Pp – Pd




J = (Jp + Jd) / 2; ΔJ = Jd – Jp.[117]


Using these new variables, the condition for litigation can be rewritten as:


Pp – Pd > C / J + P (ΔJ / J).


If ΔJ is positive—that is, if the defendant stands to lose more from a plaintiff’s victory than the plaintiff stands to gain—then the magnitude of the right-hand side of the inequality will increase as P, the average of the parties’ expectations of the likelihood of a plaintiff’s victory, increases. Therefore, in situations where the defendant has more at stake than the plaintiff, and all other things being equal, one would expect the likelihood of a case proceeding to trial would decrease as the plaintiff’s likelihood of success increases. Thus, according to Priest and Klein, “where the stakes are greater to defendants than to plaintiffs, relatively more defendant than plaintiff victories ought to be observed in disputes that are litigated. The results are reversed where the stakes are greater for the plaintiff.”[118]

B. Application of the Priest-Klein Model to Judicial Review of Agency Decisions

What are the implications of the Priest-Klein model, developed for private damages litigation, for cases of judicial review of environmental agency decisions? At first glance, it may appear that the two types of cases have little in common. In a typical environmental review case, an industry association or an environmental public interest organization does not seek money damages but rather seeks to overturn a rule or other decision of the agency.[119] The association or organization ostensibly exists to promote the interests of its members, not to maximize its own economic gain.[120] Settlement may be possible, but any settlement will typically take the form of a commitment by the agency to reconsider, revoke, or revise the decision, not a cash offer.[121] In principle, one could imagine quantifying in monetary terms the values to the organization or its members of a victory or a settlement and assume that the organization compares those values in deciding whether to proceed to litigation. But such quantification would be extraordinarily difficult and of dubious value in attempting to explain the complex of political, social, economic, bureaucratic, and legal factors that motivate an administrative agency’s decisions. Thus, the assumption that plaintiffs’ and defendants’ decisions are determined by simple quantitative comparison of the costs and benefits of litigation and settlement, which is the foundation of Priest and Klein’s model,[122] is not readily applied to environmental administrative agencies and the organizations that sue them.

Nonetheless, the qualitative conclusion—that where plaintiffs have more “at stake” than defendants, one ought to expect more plaintiffs’ than defendants’ victories in litigated cases—may apply to judicial review of environmental agency decisions.[123] To determine whether this is so, requires consideration of the dynamics underlying the last equation above.[124] In the Priest-Klein model, the settle/litigate decision is driven by the competing forces of the defendant’s fear, and the plaintiff’s hope, of the outcome of a trial.[125] The former force determines the amount that the defendant is willing to offer in settlement; the latter force determines the amount that the plaintiff will demand. Both forces increase as the likelihood of the plaintiff’s victory at trial (in the estimation of the parties) increases.[126] That is why, when the stakes for both sides are equal, the settle/litigate decision depends only on the difference between the parties’ estimates of the plaintiff’s chances of victory. Cases where both parties believe the plaintiff is likely to win are no more likely to be settled than cases where both parties give the plaintiff a slim chance.[127] When the stakes are unequal, however, the dynamics change. If the defendant has more at stake, then, as the odds of a plaintiff’s victory increase, the defendant’s interest in settlement, and hence his settlement offer, will increase more rapidly than the plaintiff’s expected return from a trial, and hence her settlement demand. Thus, the greater the plaintiff’s likelihood of trial success, the more likely a settlement will be.[128] Cases that go to trial will tend to be those where the plaintiff’s likelihood of success is least. Conversely, if the plaintiff has more at stake, then, as the plaintiff’s chances of victory increase, the plaintiff’s expected return from a trial, and hence her settlement demand, will increase more rapidly than the defendant’s settlement offer.[129] Thus, the greater the plaintiff’s likelihood of success, the less likely a settlement will be. In this situation, the cases that go to trial will tend to be those where the plaintiff’s likelihood of success is greatest.[130]

In short, where the defendant has the most at stake, the defendant’s interests will drive the settlement/trial dynamics predominantly and defendants will tend to avoid trying cases where they expect the plaintiff to win. Where the plaintiff has the most at stake, the plaintiff’s interests will drive the dynamics predominantly, and plaintiffs will tend to choose to try the cases where they expect to win.[131]

How, if at all, do these dynamics translate to judicial review of agency decisions? First, one must identify the counterpart of a “dispute” in the administrative review context. Priest and Klein define a “dispute” as “any occasion in which a plaintiff asserts a claim for some injury against a defendant.”[132] If one identifies the agency as the defendant and the potential challenger of the agency’s final decision as the plaintiff, it becomes apparent that there are multiple points at which one might deem a “dispute” to have arisen. The simplest definition is that a dispute arises when an agency makes a final decision and an affected party, dissatisfied with the decision, considers seeking judicial review of the decision.[133] However, a strong argument could be made that this simple definition overlooks much of the dynamics that precede the issuance of the agency’s decision. An agency’s interaction with affected parties generally begins long before the agency renders a final decision that is reviewable in court.[134] In a rulemaking proceeding, for example, the doctrine of exhaustion of administrative remedies requires potential litigants to present their “claims” to an agency, in the form of comments on a proposed rule, before the agency issues a final rule.[135] More generally, interested parties often express their interests, desires, and legal arguments to agencies in an ongoing process of communication—which may include letters, petitions, phone calls, personal contacts, and meetings—that may extend for years or even decades before the agency renders a decision on a particular matter.[136]

Therefore, the agency’s decision could be seen not as the beginning of a dispute but rather as a sort of settlement offer attempting to resolve a long ongoing dispute. Acceptance of this offer, however, requires no affirmative act. Rather, a party “accepts” the offer simply by doing nothing further, thereby allowing the decision to go into effect unchallenged. The party can “decline” the offer by going to court for judicial review of the agency’s decision. Settlement, in the conventional sense of settling a lawsuit after it has been filed, may still occur after a party takes the agency to court, but the universe of “settled” cases should be understood to include the—potentially much larger—class of cases where an interested party simply chooses to walk away rather than taking the agency to court.

If an agency’s final decision is conceptually equated with a settlement offer, it is best analogized with a final, bottom line, “take it or leave it” offer rather than a step in an ongoing process of negotiation. While agencies generally have power to revisit and revoke or revise their final decisions, and sometimes exercise that power in order to settle judicial challenges to those decisions,[137] an agency’s short-term likelihood of changing course voluntarily decreases drastically once it has issued a final decision for at least four reasons. First, unlike a tort or a breach of contract that may have occurred inadvertently, or without consideration of its legal consequences, an agency’s final decision generally represents the culmination of a process in which the agency, at least to some extent, heard and considered the evidence and arguments of interested parties and considered the possibility that one of those same parties might bring the agency to court if dissatisfied with the result.[138] Therefore, it may be relatively unlikely that the agency will learn anything postdecision that would cause it to change course. Second, there are generally substantial procedural barriers to such a change of course.[139] To take again the example of a rulemaking, in order to revise a final rule an agency must, at a minimum, re-initiate the rulemaking process by publishing a new proposed rule and accompanying rationale;[140] take, consider, and respond to public comments on the new proposed rule;[141] and publish a new final rule with accompanying rationale.[142] It may also have to develop and publish new supporting documentation such as a new or supplemental Environmental Impact Statement,[143] a Regulatory Impact Analysis,[144] or an analysis under the Paperwork Reduction Act.[145] All of these requirements create a substantial disincentive for an agency to adjust its action in order to settle a case. Third, an agency can often expect substantial political fallout from a last minute change of course. If the issue before the agency is of any interest to the press and the public, it may be difficult to explain why, having reached a final decision, the agency promptly decided to reverse that decision.[146] Fourth and finally, when an agency changes course it can expect to face heightened judicial scrutiny if a court reviews its actions.[147] Courts often look carefully at the reasons for a reversal and require that the agency provide an adequate explanation.[148]

There are, however, three important distinctions between an administrative agency’s final—but judicially reviewable—decision and a final settlement offer in private litigation. First, if an interested party takes the agency to court, and loses, the losing party still gets the benefit of the “settlement”; i.e., the agency’s final decision, including whatever concessions to the unsuccessful plaintiff’s interests it may have included, will remain in effect.[149] Second, the “settlement” itself will affect the plaintiff’s likelihood of success in litigation. Presumably, the more the agency’s final decision responds to the evidence and arguments presented by a party—either by refuting them or bending to them—the lower will be the same party’s chance of convincing a court that the decision was unlawful, unjustified, or unreasonable. Of course, where multiple parties with conflicting interests are involved, the agency’s bending to the will of one party—say an industry—may well increases the likelihood of a successful legal challenge by another party—say an environmental organization.

Finally, a third distinction between judicial challenges to agency decisions, particularly in the environmental field, and private damages litigation is that parties to the former are likely to be “repeat players” with a potential interest in challenging many such decisions but having limited resources to pursue such challenges.[150] Such a party filing a lawsuit will incur not only the out-of-pocket costs of litigation but also an opportunity cost, which may be much larger, of other potential cases foregone. For this reason, such a party may be better analogized to a criminal prosecutor, allocating a limited staff and budget, than to a plaintiff in private damages litigation. Such an analogy will be developed in Part IV;[151] for the meantime we will simply note that this opportunity cost may exist and may be large.

One can modify the Priest-Klein model to make a crude description reflecting the observations above. Assume that the value to an affected party of an agency’s, fixed, decision is D: D may be negative if the decision is harmful to the party. The party hopes that a trip to court will modify the agency’s decision to a more favorable decision with a value to the party of Dp and a likelihood of success of P: Dp may be zeroif the party hopes to simply nullify the decision. If the party fails in court—the likelihood of which is 1 – P—the party will be left with D. Litigation is in the party’s interest if the expectation value of litigation is greater than D, which is:


P(Dp) + (1 – P) D – C – O > D,


where C is the out-of-pocket cost of the litigation and O is the opportunity cost of other litigation foregone. Solving for P:


P> (C + O) / (Dp – D).


In other words, a rational plaintiff would look at the ratio of the total cost—direct plus opportunity—of litigation to the improvement in the party’s position, Dp – D, that would result from victory. Only if the likelihood of success exceeds that ratio would the plaintiff sue; otherwise the plaintiff would simply walk away, accepting D.

This crude model reflects the intuitive notion that a party, such as an environmental organization, with limited resources, and facing a large number of agency decisions that it might challenge, will be unlikely to pursue cases in which it has a small chance of success. For such an organization, the opportunity cost of bringing a lawsuit, O, will be large, and therefore the threshold that the organization’s likelihood of success must reach before it will sue will be large. On the other hand, if the opportunity cost is low, either because the organization is concerned with relatively few agency decisions or because it has ample resources, then the threshold will be lower, and the organization could be expected to bring some lawsuits in which its likelihood of success is relatively low.

IV. Criminal Trials as an Alternative Analogue

The need to insert a loosely defined opportunity cost into the Priest-Klein model in order to adapt it to the administrative context suggests that it is not the best model for the purpose. A better model would more fully incorporate the dynamics of a party’s selection of which cases to pursue and which to simply walk away from when the party has limited resources.

Without attributing any nefarious character to administrative agencies or their decisions, one can still see how an organization’s selection of which agency decisions to litigate may be analogous to a prosecutor’s decision of which cases to prosecute. Like a prosecutor facing more alleged or suspected crimes than she has resources to prosecute, an advocacy organization with limited resources must choose from the universe of thousands of agency decisions those relatively few that it will take to court.

Criminal trials also demonstrate the overwhelming effect that a party’s selection can have on its success rate. In 2004, 79% of all federal criminal trials resulted in convictions.[152] The conviction rate in state criminal trials is even higher.[153] This conviction rate is remarkable, especially when one considers that the standard of proof in criminal cases—namely, beyond a reasonable doubt—is the most demanding of all judicial standards.[154] Using the logic employed by some of the studies of judicial review of agency decisions, which took the affirmance rate as evidence of some combination of the degree of agency compliance with the law and the degree of judicial deference to agency discretion,[155] one could only conclude that either a) the American public is extraordinarily inclined to violate criminal laws, or b) the courts are grievously deviating from the standard of proof that they are supposed to be applying, or both. While a) or b) is possible, it should be obvious that a much more likely explanation of the high conviction rate is that prosecutors are choosing to try only cases that they feel they are likely to win.

A. The 3R Model

Eric Rasmusen, Manu Raghav, and Mark Ramseyer have recently elaborated a quantitative model of the exercise of prosecutorial discretion, which I shall call the “3R” model in recognition of their shared last initial.[156] In the 3R model, a prosecutor is faced with an inventory of potential cases to prosecute, which she ranks in order of her assessment of the strength of the evidence to support a conviction.[157] Given this ranking, a prosecutor with a fixed budget has two choices to make. First she must choose how far down the rank list she wishes to go in prosecuting cases. She is assumed to establish a cutoff below which she deems cases to be too weak to be worth prosecuting, given her finite resources.[158] Second, she must choose how to distribute her budget among those cases she does prosecute.[159] The model assumes that there is a certain minimum fixed cost to bringing any case and that the discretionary amount beyond the fixed cost that the prosecutor will allocate to each case is a function of the strength of the case.[160] This function is constrained by the requirement that the total expenditure on all cases above the cutoff not exceed the prosecutor’s budget.[161] The model further assumes that, for each case, the probability of the prosecutor achieving a conviction is the strength of the case times a factor P, which in turn depends on the amount of discretionary resources—E, for effort—invested in the case.[162] P is presumed to be zero when E is zero; i.e., simply filing a case and investing no effort whatsoever in it will not yield a conviction. P increases with E but with decreasing slope, i.e., each extra dollar spent on a case yields some increase in the likelihood of a conviction, but with diminishing return.[163] Finally, P is presumed to be less than one, that is, no amount of prosecutorial effort can yield a likelihood of conviction greater than the underlying strength of the evidence.[164]

The three Rs considered the effect of the prosecutor’s budget on the conviction rate and showed that an increase in the budget can either increase or decrease the conviction rate, depending on how the increase is allocated.[165] The conviction rate will increase if most of the new money is allocated to increase the level of effort expended on cases that would have been brought even without the increase.[166] On the other hand, if the increase is used to reach deeper into the barrel of potential defendants and prosecute cases previously considered too weak to pursue—i.e., to lower the cutoff, then the increased budget could result in a decrease in the conviction rate.[167]

How an increase in budget will be allocated depends on the prosecutor’s objectives. In their first iteration of the model, the three Rs assume that a conscientious, apolitical prosecutor will attribute a value to each successful prosecution that is an increasing function of the sentence imposed, which is in turn is an increasing function of the resources expended, and that she will seek to maximize the sum of the values of all successful prosecutions.[168] They consider an alternative version in which the prosecutor also seeks to maintain a high conviction rate—by avoiding hard cases—in order to please voters or superiors, or seeks to otherwise dedicate available resources to promoting goals other than conviction of criminals.[169]

How can one quantitatively characterize the objectives of an environmental organization choosing which administrative agency decisions to challenge in court? At first glance, it would seem to be an extraordinarily daunting task. The goals of such organizations are likely to be diverse and often nonmonetary, including protection of human health, preservation of natural landscapes, maintenance of populations of wildlife, mitigation of global warming, or protection and enhancement of opportunities for outdoor recreation.[170] Further complicating the picture is that a judicial victory or loss for such an organization may have greater symbolic, precedential, or fundraising impact than its direct impact on the organization’s environmental objectives.[171]

I suggest, however, that, at least for the purpose of an initial stab at modeling the litigation decisions of an environmental organization, it is worth considering a very simple assumption, namely that such an organization will attempt to maximize the number of cases that it wins.[172] Though crude, this assumption is not completely divorced from reality. Every court victory, regardless of the scope or the nature of the issues involved, brings benefits to a nonprofit organization in the business of litigation. Each victory enhances the organization’s reputation, raises its public profile, and provides concrete, tangible evidence of the organization’s effectiveness that can be reported to members, donors, and potential donors.[173] Within the organization, each victory gives attorneys a favorable result to report to their colleagues and superiors. Each victory also presents an opportunity to apply for an award of attorney fees from the government, which are available to a prevailing party under the federal Equal Access to Justice Act[174] as well as several more specific attorney fees provisions in such statutes as the Endangered Species Act.[175]

The assumption of win maximization would make an environmental organization’s objectives different not only from a private individual or corporation seeking to optimize its economic position but also from those of a prosecutor who seeks to advance his career by maximizing his conviction rate, or percentage of successful prosecutions. The three Rs presented anecdotal evidence that prosecutors publicize their conviction rates and argued that an unsophisticated public may fail to realize that prosecuting only the easier cases and avoiding the harder ones can achieve a high conviction rate.[176] However, based on my personal observation and experience as an environmental lawyer, I do not believe that an attorney employed by an advocacy organization would please his superiors, nor would the organization please its members and donors, by maintaining a small docket with a high success rate. Each win is an opportunity for the organization to demonstrate its importance and effectiveness; the percentage of wins is of little consequence. A case brought but lost is a detriment to the organization because of the resources that it uses up, but not because it dilutes the organization’s batting average.

Despite these arguments in defense of the win maximization assumption, it is clearly a crude gloss on the multiple strategic, political, economic, and even personal factors that go into an environmental organization’s choices of which cases to litigate. Nonetheless, the predictions that flow from such an assumption are worth considering in that they can shed light on how and why an environmental organization’s success rate may differ from that of an economically driven private litigant or a prosecutor. In the real world, where such an organization’s motives are almost certainly mixed, the organization’s success rate will likely fall somewhere between that predicted by such an assumption and that predicted by the purely economic or prosecutorial models.

V. Plaintiffs’ Selection in Wilderness Litigation

Each year, the four federal agencies managing units of the National Wilderness Preservation System—the National Park Service, the U.S. Forest Service, the Bureau of Land Management, and the U.S. Fish & Wildlife Service[177]—issue scores of formal decisions concerning management of the wilderness areas under their jurisdiction.[178] In addition to these formal decisions, they also take countless other informal steps that could potentially be construed as agency action subject to judicial review under the Administrative Procedure Act.[179] Yet of these numerous agency actions, only ninety-four, or an average of less than three per year since the creation of the system, were challenged in lawsuits that led to judicial decisions reported in Professor Appel’s study.[180] Clearly, litigation over wilderness management is highly selective.

Professor Appel classified each principal case in his study as seeking either more protection—and therefore fewer uses or greater restrictions on uses, or less protection—and therefore more uses or less restrictions on uses—of wilderness areas.[181] This classification was based on the substance of the claim or defense asserted against the federal government in each case.[182] Professor Appel did not categorize the cases according to the types of entities opposing the federal government.[183]

Examination of the cases, however, reveals that the types of entities involved in the two types of suits were distinctly different.[184] Of the fifty suits seeking more protection, forty-five (90%) were brought by nonprofit, public interest organizations created wholly or partially for the purpose of protecting the environment or natural resources.[185] Of the remaining five lawsuits seeking “more protection” for wilderness areas, just three (6%) were brought by private individuals and two (4%) were brought by Native American governments.[186]

Repeat players were common among the litigants seeking greater protection of wilderness areas. The Sierra Club, Wilderness Society, and Wilderness Watch were each lead plaintiff in six lawsuits; the Izaak Walton League of America was lead plaintiff in three.[187] Together, these four national organizations accounted for 40% of the “more protection” cases.[188] Two regional organizations, the High Sierra Hikers Association and Friends of the Boundary Waters Wilderness were lead plaintiffs in five and four cases, respectively. Altogether, these six organizations were plaintiffs in over half of the “more protection” cases.[189]

In contrast, only seven out of forty-four, or about 16%, of the “less protection” cases pitted nonprofit associations against the government, while nearly half of these cases, twenty-one, were brought by private individuals.[190] Another eleven of the “less protection” cases involved for-profit business corporations, so that altogether nearly three quarters of these cases—thirty-two of forty-four—involved what could be characterized as essentially private interests.[191] The remaining five “less protection” cases were brought by state or local governments.[192]

In contrast to the repeat players involved in the majority of the “more protection” cases, repeat players were rare in the “less protection” cases. Only two individuals, and no nonprofit organizations, corporations, or governmental entities, were named parties in more than one case.[193] Moreover, although each of these two individuals was involved in more than one case, each was a private landowner concerned only with the effects of federal wilderness management on his or her use and enjoyment of his or her own property.[194] Kathy Stupak-Thrall was involved in three separate actions against restrictions on her use of Crooked Lake in Michigan’s Sylvania Wilderness Area, where she owned lakefront property;[195] Ned Fixel twice sought compensation—once through a takings claim[196] and once through a tort claim[197]—for his inability to work his mining claims on a creek tributary to Idaho’s Wild and Scenic Salmon River.

Furthermore, while all of the “more protection” cases were attempts to alter, in one way or another, the management of wilderness areas or potential wilderness areas, fourteen of forty-four, or almost one third, of the “less protection” cases were not, on their face, about land management.[198] Six were claims for monetary compensation under the Federal Tort Claims Act;[199] four were claims for monetary compensation for alleged takings of private property;[200] four were criminal prosecutions.[201]

The stark difference in the nature of the litigants and their claims between the “more protection” and the “less protection” cases suggests an equally stark difference in the selection processes that caused these, rather than other, cases to end up in court. The national and regional conservation organizations that brought the vast majority of the “more protection” cases are concerned with the management of many, if not all, of the nation’s 794 designated wilderness areas as well as large additional areas of public land that they would like to see designated as wilderness.[202] For example, The Wilderness Society is a national organization with over 500,000 members and ten regional offices, spanning the United States from Maine to California and Georgia to Alaska.[203] It advocates for designation and protection of wilderness areas in all areas of the country. The management of most, if not all, of the nation’s 758 designated wilderness areas, spanning over 100 million acres, is therefore of concern to The Wilderness Society.[204] Even the smaller Friends of the Boundary Waters Wilderness is concerned with numerous issues including logging, mining, invasive species, and air pollution on over one million acres of the Boundary Waters Canoe Area Wilderness as well as roads and development on tens of thousands of acres of surrounding lands.[205] For these organizations, the fifty reported court decisions in which they challenged the government’s management of wilderness areas and potential wilderness areas are just the tip of the iceberg of the government decisions with which they are concerned.[206] If each wilderness area were the subject of just one management decision per year for the last thirty years, these fifty cases would represent less than one quarter of one percent of all such decisions. Even allowing for several lawsuits filed for each reported decision, it is still clear that these organizations take to court only a small fraction of the government decisions that concern them. If they select to litigate only the cases in which they stand a relatively high chance of success, then such selection can be expected to yield a high success rate.

On the other hand, the private individuals and businesses who opposed the government in over 70% of the forty-four “less protection” cases were generally not in a position to pick and choose their battles among all of the government decisions affecting hundreds of wilderness areas and tens of millions of acres across the country.[207] Typically, these private parties were each concerned with the much smaller universe of decisions affecting their own land or business. Four of them did not choose to litigate at all; they were defendants in criminal prosecutions brought by the government.[208] Another six were plaintiffs in tort actions claiming physical or economic injuries as a result of the government’s management of a particular wilderness area.[209] Most of the rest were owners of land or proprietors of businesses in or near a particular wilderness area challenging a management decision that adversely affected the use of their land or the profitability of their business.[210] These plaintiffs, unlike the organizational plaintiffs in the “more protection” cases, did not have the ability to pick and choose, based on likelihood of success, which among hundreds or thousands of wilderness management decisions to challenge.[211] Kathy Stupak-Thrall owned property on one lake in one wilderness area;[212] Ned Fixel had a mining claim in one wilderness area;[213] the Kerr-McGee Corporation held prospecting permits and had allegedly discovered valuable phosphate deposits in one wilderness area in Florida.[214] These plaintiffs generally had no reason and no standing to challenge the management of any other wilderness area or even any other locale within the same wilderness area.[215]

This is not to say that these parties (other than the criminal defendants) did not have a choice whether to litigate. Like plaintiffs in tort or contract actions, they had the choice between suing and walking away. Further, once having sued, they may have been offered, and declined, opportunities to settle their cases in lieu of litigating to final judgment. But they generally did not have the option of pursuing another case, in another wilderness area, where they had no property or business, simply because it offered a better chance of success.

Like the “more protection” cases, the “less protection” cases in Professor Appel’s sample are likely the small tip of a large iceberg. It is fair to assume that there are many more than forty-four individuals and businesses who are somehow aggrieved by the limitations that the Wilderness Act imposes on uses of public lands adjacent to their property or on which they conduct business.[216] But, in the “less protection” cases we do not have a few organizations able to choose the strongest cases among the universe of potential cases. Instead, we have myriad individuals and businesses, each choosing whether or not to pursue his own case.[217] Doubtless, the perceived likelihood of success is one factor affecting that choice, but there are many other factors unrelated to the strength of the case. These include:

— the magnitude of the plaintiff’s interest. All other things being equal, a landowner or businessperson who suffers a major economic hit as a result of wilderness designation or management is more likely to sue than one who merely suffers a minor inconvenience. However, the magnitude of the grievance does not necessarily correlate with the strength of the legal case.

— the resources of the potential plaintiff. For example, a landowner who finds access to her property inhibited by an adjacent wilderness area will be more likely to sue if she has sufficient funds to hire an attorney or sufficient spare time to devote to the endeavor. But there is no reason to believe that the leisure class has stronger legal claims than do their working class counterparts.

— the litigiousness of the potential plaintiff. Some like to sue; some do not. But there is no reason to believe that the litigation-prone have stronger cases than their more reticent brethren.

In short, the “more protection” and the “less protection” cases studied by Professor Appel are both highly selective samples of much larger universes of potential cases. But the selection processes are markedly different. The former group is, for the most part, selected by organizations that have many potential claims and can be expected to litigate the stronger ones and walk away from the weaker ones.[218] The latter group is the result of many separate decisions by numerous individuals and businesses, each deciding, for its own reasons—many of which are unrelated to the strength of the case—whether or not to pursue its claim.[219] It is therefore no surprise that the plaintiffs in the former group enjoy a much higher winning percentage in the courtroom than do those in the latter group.

VI. Which Model?

The qualitative discussion in the preceding Part suggests that a single model cannot explain the differing success rates of the “more protection” and “less protection” litigants opposing the federal government in wilderness-related litigation.[220] Because the two groups of litigants are different types of entities, pursuing different objectives, and faced with different options, different models are needed to understand their choices and explain their success rates.

A. “More Protection” Cases: Modified Priest-Klein or Criminal Model

Since the overwhelming majority of the “more protection” cases are brought by public-interest environmental organizations, the modified form of the Priest-Klein model suggested in Part II.B above, which was developed specifically to model public-interest challenges to administrative agency decisions, would seem appropriate.[221] Recall that, in this model, a public interest organization will sue to overturn an agency decision only if:


P> (C + O) / (Dp – D)


where P is the organization’s estimate of the likelihood of success in the litigation, C is the out-of-pocket cost of the lawsuit, O is the marginal opportunity cost of litigation or other activities foregone as a result of the choice to invest resources in this litigation, and (Dp – D) is the incremental benefit to the organization of a successful lawsuit.

As noted in Part II.B, this is a crude model, and the key parameters O, Dp,and D are dependent on a complex and difficult to quantify galaxy of the organizations’ values, objectives, and incentives.[222] It would therefore be unrealistic to attempt to use this model to precisely match the 52% success rate that such organizations achieved in the wilderness cases studied by Professor Appel.[223] Nonetheless, the model can provide some understanding of why the success rate is so high.

The marginal opportunity cost, O, can be seen as the foregone value to the organization of the next best case that the organization could bringif it does not invest in this case instead. In other words, it is the value of the case that is “next in line” in the organization’s “wish list” of cases it would have liked to bring.[224] In the limit where that wish list is very long, and the organization’s resources are very limited, the lawsuits actually brought will represent just the very top of the wish list, and therefore many valuable wishes will go unfulfilled—i.e., many meritorious and valuable, to the organization, lawsuits will not be brought. In that limit, O, the unrealized value to the organization of the best lawsuit foregone, may be much greater than the out-of-pocket cost, C, of any of the organization’s actual lawsuits, and may be nearly equal to—Dp – D, the benefit to the organization of the lawsuit in question. In that situation, the value of P in the above equation would approach one—i.e., the organization would only bring lawsuits that it thinks it is virtually certain to win. Thus, a success rate approaching 100% could be possible.

Of course, no organization can perfectly estimate its likelihood of success in any given lawsuit, and it is fair to assume that many environmental litigants have lost cases that they were sure they would win. Furthermore, although it is true that the number of lawsuits brought by wilderness advocates is substantially smaller than the number of wilderness management decisions potentially subject to litigation, the limit of an infinitely long unfulfilled wish list assumed in the discussion above is not realistic. Nonetheless, given the possibility of a near 100% success rate in the extreme scenario, the actual success rate of just over 50% in Professor Appel’s “more protection” cases is not surprising.[225] And the model shows that such a success rate, or an even higher one, can be the result of the limited resources and the strategic choices of wilderness advocates, revealing little or nothing about the tendency of federal wilderness management agencies to follow, or not to follow, the law or about the tendency of federal judges to defer, or not to defer, to the agencies.[226]

The criminal prosecution model, discussed above in Part IV, also suggests that the 52% success rate achieved by environmental organizations is not at all extraordinary.[227] As noted in that Part, criminal prosecutors in the United States routinely achieve conviction rates in excess of 80% despite the presumption of innocence and the strict beyond a reasonable doubt standard of proof that judges and juries are supposed to apply in criminal cases.[228] The 3R model suggests that prosecutors’ choices, and the resultant conviction rates, depend on the prosecutors’ resources and their objectives.[229] Limited resources, combined with objectives that include a high conviction rate—as opposed to a high number of convictions—can cause a prosecutor to forego all but the strongest cases, resulting in a high conviction rate.[230] In Japan, where very limited prosecutorial resources force prosecutors to forgo all but the strongest cases, the conviction rate exceeds 95%.[231]

As previously discussed, the prosecutorial model applies imperfectly to environmental organizations’ choices about which administrative agency decisions to challenge and which to let stand.[232] These organizations are likely to be more interested in maximizing their total number of court victories than in maintaining a high percentage of successes.[233] Moreover, even the most cynical observer could likely agree that the number of final, reviewable agency actions that are blatantly unlawful is far less than the number of crimes committed across the United States each year, so that environmental organizations are working in a much less target-rich environment than are criminal prosecutors.[234] For these reasons, they should not be expected to match the 80–90% success rate achieved in criminal prosecutions. However, given the advantage that they share with prosecutors, namely, the opportunity to select from many potential cases a substantially smaller number to pursue, the 52% rate that they have achieved in wilderness related cases is not surprising, and Professor Appel’s suggestion that it is evidence of judicial bias is unconvincing.[235]

B. “Less Protection” Cases: Random Selection?

While the application of the Priest-Klein and prosecutorial models to environmental organizations’ litigation choices is problematic, the application of these models to Professor Appel’s “less protection” cases is even more fraught with difficulty. As noted in Part II.C, these cases are of several very different types, including not only suits for judicial review of administrative agencies’ decisions but also tort and takings claims and even criminal prosecutions.[236]

For the criminal prosecutions, of course, the criminal prosecutorial model applies, but it applies in the opposite way than it does to the “more protection” cases. That is, in the criminal prosecutions for violations of the Wilderness Act, the anti-government litigant seeking “less protection” of wilderness areas is the defendant, and the ability to selectively prosecute only the strongest cases lies with the government. It is therefore entirely expected that, among these cases, the government win the vast majority.

In the rest of the “less protection” cases—tort and takings claims and suits for judicial review of agency decisions—the anti-government litigator does make the choice of whether or not to sue. However, we saw in Part V that the typical plaintiffs in these cases have little or no opportunity to pick the strongest cases out of a multitude of potential lawsuits.[237] Therefore, the prosecutorial model would not apply to these cases.

One could, however, apply the modified Priest-Klein model that was applied above to the “more protection” cases.[238] To reiterate, in this model:


P > (C + O) / (Dp – D),


where P is the plaintiff’s estimate of the likelihood of success in the litigation, C is the out-of-pocket cost of the lawsuit, O is the marginal opportunity cost of litigation or other activities foregone as a result of the choice to invest resources in this litigation, and (Dp – D) is the incremental benefit to the plaintiff of a successful lawsuit.[239] If the plaintiff has few or no alternative opportunities to litigate, O may be small compared to C and to (Dp – D), in which case the plaintiff will sue whenever her estimate of the likelihood of success is greater than the ratio of the out-of-pocket cost to the perceived benefit of a successful suit. If the out-of-pocket cost is sometimes small, and the importance of the issues to some plaintiffs are large, then one could expect, in sharp contrast to the “more protection” cases, to see some “less protection” lawsuits filed even where plaintiffs believe the likelihood of success is small. Add to this effect the possibility that the actual likelihood of success may be substantially less than a plaintiff believes it to be, and once again one is not surprised to see a relatively low rate of success in the “less protection” cases.

As with the “more protection” cases, it may be informative to consider an extreme scenario to compare with the observable data. In Part V, I argued that, in the “less protection” cases, plaintiffs’ selection of which suits to bring are likely to be significantly influenced by several factors unrelated to the likelihood of success.[240] In the limit where the selection is entirely driven by such factors, the cases brought would be, from the standpoint of success rate, tantamount to a random sample. That is, if plaintiffs were selecting cases entirely according to factors unrelated to the likelihood of success, then their success rate would be the same as if they were challenging a random sample of agency decisions. In that scenario, the success rate might simply reflect the tendency of agencies to follow the law and the tendency of judges to defer to the agencies. If both of those tendencies were very high, as they would be in an ideal world, then one would expect challenges to a random sample of agency decisions to have a very low success rate, perhaps even substantially lower than the 14% that litigants in the “less protection” cases have achieved.[241] Thus, the observed 14% success rate could be seen as the result of a modest degree of selectivity based on likelihood of success, which selectivity is enhancing a success rate that would otherwise be even lower.

About Author

Late Professor of Law, Sandra Day O’Connor School of Law, Arizona State University.

Alex W. Smith Professor, University of Georgia School of Law.

Professor of Law and Director of the Natural Resources Law Center, University of Colorado School of Law.


Footnotes    (↵ returns to text)

  1. Peter A. Appel, Wilderness and the Courts, 29 Stan. Envtl. L.J. 62 (2010).
  2. Federal wilderness areas are managed by four different federal agencies: the U.S. Forest Service, the Bureau of Land Management, the National Park Service, and the U.S. Fish & Wildlife Service. Id. at 70.
  3. 16 U.S.C. § 1131 (2006).
  4. Appel, supra note 1.
  5. Id. at 112.
  6. Id. at 112–13.
  7. Id. at 113.
  8. Id. at 119.
  9. Id.
  10. See, e.g., William S. Jordan, III, Judges, Ideology, and Policy in the Administrative State: Lessons from a Decade of Hard Look Remands of EPA Rules, 53 Admin. L. Rev. 45 (2001); Paul G. Kent & John A. Pendergrass, Has NEPA Become a Dead Issue? Preliminary Results of a Comprehensive Study of NEPA Litigation, 5 Temp. Envtl. L. & Tech. J. 11 (1986); Orin S. Kerr, Shedding Light on Chevron: An Empirical Study of the Chevron Doctrine in the U.S. Courts of Appeals, 15 Yale J. on Reg. 1 (1998); Thomas J. Miles & Cass R. Sunstein, Do Judges Make Regulatory Policy? An Empirical Investigation of Chevron, 73 U. Chi. L. Rev. 823 (2006); Peter H. Schuck & E. Donald Elliott, To the Chevron Station: An Empirical Study of Federal Administrative Law, 1990 Duke L.J. 984 (1990).
  11. See Schuck & Elliott, supra note 10, at 1007–09; Kent & Pendergrass, supra note 10, at 13.
  12. See Kerr, supra note 10, at 1.
  13. See Miles & Sunstein, supra note 10, at 825–26; Jordan, supra note 10, at 47–48.
  14. As was stated regarding a similar flaw in studies of jury verdicts, “For the rate of plaintiff verdicts to be an accurate measure of the influence of a legal standard, of judicial or jury attitudes, or of the substantive fairness of any adjudicatory process, litigated disputes must be representative of the entire class of underlying disputes.” George L. Priest & Benjamin Klein, The Selection of Disputes for Litigation, 13 J. Legal Stud. 1, 4 (1984).
  15. See, e.g., Schuck & Elliott, supra note 10, at 1009 (stating “[o]utcomes in the reviewing courts are a function of (at least) two variables—how much courts are demanding, and how well agencies are conforming to the dictates of the law.”).
  16. See, e.g., Priest & Klein, supra note 14, at 4.
  17. See Eric Rasmusen et al., Convictions versus Conviction Rates: The Prosecutor’s Choice, 11 Am. L. & Econ. Rev. 47, 50 (2009).
  18. The only study to have explicitly noted plaintiffs’ case selection as a major explanatory factor was, ironically, one of the earliest published studies. See Kent & Pendergrass, supra note 10, at 14 (noting that “[o]ne possible explanation [of the high success rate of environmental organizations in litigation under the National Environmental Policy Act] is that, partly because of these economic limits, environmental groups [choose] the cases they litigate very carefully”).
  19. Social Security Act, 42 U.S.C. § 1010(d) (2006) (providing that regulations promulgated by the SSA “may provide for the suspension and termination of entitlement to benefits . . . as the Commissioner determines is appropriate”).
  20. See, e.g., Appel, supra note 1, at 116 (explaining that the Ninth Circuit disallows National Environmental Policy Act challenges for “those who wish to use NEPA . . . for other concerns such as economic effects of a proposed action”).
  21. Id.
  22. Priest & Klein, supra note 14, at 1.
  23. See, e.g., Mark Motivans, U.S. Dep’t of Justice, Bureau of Justice Statistics, Federal Justice Statistics, 2010, at 2 (2013), available at http://www.bjs.gov/content/pub/pdf/fjs10.pdf (finding approximately nine out of ten defendants in cases adjudicated in U.S. district courts were convicted in 2010).
  24. See generally Rasmusen et al., supra note 17 (using two different models and empirical evidence to analyze the relationship between prosecution rates, conviction rates, and budgets).
  25. See infra notes 186­–87 and accompanying text.
  26. Warner W. Gardner, Federal Courts and Agencies: An Audit of the Partnership Books, 75 Colum. L. Rev. 800, 803 (1975).
  27. Id. at 805, 819.
  28. Id. at 804.
  29. Id. at 805.
  30. Id. at 820.
  31. 42 U.S.C. §§ 4321–4347 (2006).
  32. Kent & Pendergrass, supra note 10.
  33. Id. at 12.
  34. Id. at 13–14.
  35. Id. at 14.
  36. Id.
  37. 467 U.S. 837 (1984).
  38. Id. at 842–44.
  39. Schuck & Elliott, supra note 10, at 984.
  40. Id. at 989 n.13.
  41. Id. at 1007.
  42. Id. at 1007–08.
  43. Id. at 1009–10.
  44. Id. at 1011–13.
  45. Id. at 1021.
  46. Id.
  47. Id. at 1022–23.
  48. Id. at 1023.
  49. See discussion infra Part V.
  50. See Richard L. Revesz, Environmental Regulation, Ideology, and the D.C. Circuit, 83 Va. L. Rev. 1717, 1718­–19 (1997).
  51. Id. at 1738–39.
  52. See Kerr, supra note 10, at 39.
  53. Id. at 38–39.
  54. Id. at 39–40.
  55. Id. at 31–35.
  56. Id. at 43.
  57. See Jordan, supra note 10, at 47–48.
  58. Id. at 48.
  59. Id. at 50.
  60. Id. at 98–100.
  61. See Miles & Sunstein, supra note 10.
  62. Id. at 825.
  63. Id. at 830–31.
  64. Id.
  65. Id. at 870–71.
  66. Id.
  67. 16 U.S.C. §§ 1131–1136 (2006 & Supp. II 2008), amended by Omnibus Public Land Management Act of 2009, Pub. L. No. 111-11, 123 Stat. 991.
  68. See generally Appel, supra note 1.
  69. See Wilderness Act of 1964, 16 U.S.C. §§ 1131–1136 (2006).
  70. Id. § 1133(b).
  71. Id. § 1133(c).
  72. Id. § 1131(a).
  73. Wilderness Institute, The Beginnings of the National Wilderness Preservation System, http://www.wilderness.net/NWPS/fastfacts (last visited Apr. 12, 2014).
  74. 16 U.S.C. § 1132(b) (2006).
  75. Id. § 1132(c).
  76. 43 U.S.C. §§ 1701–1782 (2006).
  77. Id. § 1782(a).
  78. See, e.g., Alaska National Interest Lands Conservation Act (ANILCA) of 1980, Pub. L. No. 96-487, §§ 701–703, 94 Stat. 2371, 2417–18 (designating wilderness areas in National Parks, National Wildlife Refuges, and National Forests in Alaska); National Forest System Lands, Designations, Pub. L. No. 96-560, § 102, 94 Stat. 3265, 3265–68 (1980) (designating wilderness areas in National Forests in Colorado); California Desert Protection Act of 1994, Pub. L. No. 103-433, § 102, 108 Stat. 4471, 4472–81 (1994) (designating wilderness areas on BLM lands in the California Desert).
  79. See, e.g., Arizona Desert Wilderness Act of 1990, Pub. L. No. 101-628, § 101(a), 104 Stat. 4469, 4469–72 (1990) (designating almost 1.1 million acres of wilderness areas on BLM lands in Arizona); Ross W. Gorte & Pamela Baldwin, Cong. Research Serv., RS21917, Bureau of Land Management (BLM): Wilderness Review Issues 2 n.4 (2004) (noting that the Arizona Wilderness Act was passed before the BLM could make its recommendations).
  80. Wilderness Institute, supra note 73. These 109 million acres comprise approximately 43.9 million acres (40% of the total) in National Parks, 36.2 million acres (33%) in National Forests, 20.7 million acres (19%) in National Wildlife Refuges and other areas managed by the U.S. Fish & Wildlife Service, and 8.8 million acres (8%) of BLM lands. Wilderness Institute, Wilderness Statistics Reports, http://www.wilderness.net/NWPS/chart (select “Acreage by Agency”) (last visited Apr. 12, 2014).
  81. See, e.g., ANILCA, 16 U.S.C. § 3170(a) (2006) (authorizing, under certain conditions, use of snowmobiles, aircraft, and motorboats in “conservation system units” in Alaska); id. § 3102(4) (defining “conservation system units” to include wilderness areas); Central Idaho Wilderness Act of 1980, Pub. L. No. 96-312, § 5(d)(1), 94 Stat 948, 949 (allowing prospecting for and mining of cobalt and associated minerals within Idaho’s River of No Return Wilderness).
  82. Appel, supra note 1, at 112 n.212.
  83. Id. at 112 and text accompanying supra note 1.
  84. See id.
  85. See, e.g., Friends of Boundary Waters Wilderness v. Bosworth, 437 F.3d 815, 819–20 (8th Cir. 2006) (challenging Forest Service calculation of quotas for use of motorboats in the Boundary Waters Canoe Area Wilderness); Johnson v. U.S. Forest Serv., 93 Fed. App’x. 133 (9th Cir. 2004) (challenging Forest Service decision to disallow motorized travel across Absaroka-Beartooth Wilderness to access private land inholding).
  86. See, e.g., United States v. Gotchnik, 222 F.3d 506 (8th Cir. 2000) (affirming the conviction of unauthorized use of motor vehicles and motor boats in the Boundary Waters Canoe Area Wilderness); United States v. Gregg, 290 F. Supp. 706, 706 (W.D. Wash. 1968) (providing an example of criminal prosecution that dealt with the violation of an unauthorized landing of an aircraft in a National Forest wilderness area).
  87. See, e.g., Zumwalt v. United States, 928 F.2d 951, 951–952, 955 (10th Cir. 1991) (affirming the government’s exemption in a tort claim for injuries sustained in fall allegedly caused by government’s failure to adequately maintain, and to post warning sign along, trail in wilderness area in Pinnacles National Monument); Wright v. United States, 868 F. Supp. 930 (E.D. Tenn. 1994) (dismissing tort claim brought against the government for injuries caused by a falling tree in a wilderness area).
  88. See, e.g., Pete v. United States, 531 F.2d 1018, 1021–22 (Ct. Cl. 1976) (challenging regulation banning the use of barges within a wilderness area as a taking); Barnes v. Babbitt, 329 F. Supp. 2d 1141, 1150 (D. Ariz. 2004) (alleging that restrictions to land access amount to a taking without just compensation).
  89. Appel, supra note 1, at 112–13.
  90. Id.
  91. Id.
  92. Id. at 113.
  93. Id. at 113 tbl.1.
  94. Id.
  95. Id. at 117 tbl.2.
  96. Id. at 119.
  97. Id.
  98. Id.
  99. Id. at 119–20.
  100. Id. at 120–21, 124–25.
  101. Id. at 121–22.
  102. Id. at 122–23.
  103. Id. at 123­–24.
  104. Id. at 119–24.
  105. See, e.g., Theodore Eisenberg & Henry S. Farber, The Litigious Plaintiff Hypothesis: Case Selection and Resolution, 28 RAND J. Econ. (Special Issue) S92, S92–93, S111 (1997).
  106. See generally Priest & Klein, supra note 14, at 4–5, 43–44 (discussing considerations weighed by a plaintiff in selecting whether to pursue litigation); Eisenberg & Farber, supra note 105, at S93, S111 (concluding that potential claims are selected for litigation based upon the litigiousness of the potential plaintiff).
  107. See generally Priest & Klein, supra note 14, at 53–54 (discussing Posner’s model, and finding that budget constraints influence the prosecutorial discretion of administrative agencies).
  108. See id. at 4.
  109. Id.
  110. See id. at 53–54 (noting “[t]here is little reason to regard a government agency as a dollar maximizer”).
  111. See id. at 4, 12–13, 26.
  112. See id. at 4, 12–13, 26–27.
  113. See id.
  114. See id. at 12.
  115. Id. at 13.
  116. Id. at 24.
  117. Id. at 25.
  118. Id.
  119. See, e.g., Jan Chatten-Brown & Douglas Carstens, Practicing Public Interest Environmental Law in the Private Sector, 38 A.B.A. Trends 8 (2007) (noting that “[i]n many environmental legal challenges, irreplaceable resources are protected, but no damages are sought or awarded”).
  120. See Barton H. Thompson Jr., The Continuing Innovation of Citizen Enforcement, 2000 U. Ill. L. Rev. 185, 187, 192–93 (2000) (explaining that nonprofit “citizen monitoring organizations” typically initiate lawsuits under the federal citizen suit provision for an injunction, but that a growing number of provisions also authorize courts to impose monetary penalties, payable to the United States).
  121. See Joel S. Jacobs, Compromising NEPA? The Interplay Between Settlement Agreements and the National Environmental Policy Act, 19 Harv. Envtl. L. Rev. 113, 114, 151 (1995).
  122. Priest & Klein, supra note 14, at 4.
  123. See generally id. at 52-54 (discussing theory’s application when the government is a party in antitrust actions with reasoning that also applies to other government suits).
  124. See supra text accompanying notes 116-17.
  125. See Priest & Klein, supra note 14, at 4 (explaining the relationship between the parties’ perceived economic costs and the likelihood of either litigation or settlement).
  126. See id. at 12 (noting correlation between parties’ expectations and willingness to settle).
  127. See id. at 15–16 (arguing that the decision not to settle a dispute relates to inadequate assumptions about the strength of each party’s case, rather than assumptions of the likelihood of plaintiff’s success in litigation).
  128. See id. at 26 (discussing “where defendants stand to lose more from adverse verdicts than plaintiffs stand to gain, the relative calculus of the parties with respect to litigation and settlement”).
  129. See id. at 26–27 (“[W]here plaintiffs stand to gain more from verdicts than defendants stand to lose, plaintiffs in general will increase their settlement demands, but will tend to increase them more in disputes in which plaintiffs have a relatively greater chance of winning than in disputes in which defendants have a relatively greater chance of winning. . . . Where the stakes to plaintiffs are relatively greater, the level of plaintiff’s demands will increase generally (relative to defendants’ bids or offers), and the rate of litigation will increase.”).
  130. See id. at 27.
  131. See id. at 26–27 (noting that the rate of litigation will increase when plaintiff has heightened stakes and a high likelihood of success).
  132. Id. at 6.
  133. See, e.g., Administrative Procedure Act, 5 U.S.C. § 702 (2006) (allowing judicial review under the Administrative Procedure Act for “[a] person suffering legal wrong because of agency action, or adversely affected or aggrieved by agency action within the meaning of a relevant statute”).
  134. See Robert L. Glicksman & Richard E. Levy, Administrative Law: Agency Action in Legal Context 1131–32 (2010) (discussing exhaustion requirements, which mandate that aggrieved parties present arguments to the agency prior to initiating a proceeding in court).
  135. See 5 U.S.C. § 553 (2006) (requiring that agencies give interested persons the opportunity to comment on proposed rulemaking); Office of the Federal Register, A Guide to the Rule Making Process 5–6 (2011), available at http://www.federalregister.gov/uploads/2011/01/the_rulemaking_process.pdf (discussing agencies’ rulemaking processes and the importance of comments in the context of the doctrine of exhaustion).
  136. See generally McCarthy v. Madigan, 503 U.S. 140, 144–49 (1992) (providing useful background on the exhaustion doctrine, its purposes, and its exceptions).
  137. See B.J. Alan Comp., Inc. v. Interstate Commerce Comm’n, 897 F.2d 561, 562–63 (D.C. Cir. 1990) (“The Commission has discretion to reconsider, so long as its resumption does not conflict with proceedings in court.”).
  138. See generally Office of Federal Register, supra note 135, at 8 (discussing how public comments affect final rules and when courts get involved in the rulemaking process).
  139. See id.
  140. Id.; see also Administrative Procedure Act of 1946, 5 U.S.C. §§ 551, 553 (2006) (definition of “rule making” includes amending rules and requires notice).
  141. Office of Federal Register, supra note 135 (discussing how agencies consider and respond to public comment).
  142. Id.
  143. See National Environmental Policy Act of 1969, 42 U.S.C. §§ 4321–4347 (2006).
  144. See Exec. Order No. 12,866, 58 Fed. Reg. 51,735, 51,735–36 (Sept. 30, 1993); 1 Charles H. Koch, Administrative Law and Practice § 4:50 (2d ed. 1997).
  145. See Paperwork Reduction Act, 44 U.S.C. § 3501 (2006).
  146. See Motor Vehicle Mfrs. Ass’n of the U.S., Inc., v. State Farm Mut. Auto. Ins. Co. (State Farm), 463 U.S. 29, 34 (1983).
  147. Id. at 43, 52; Koch, supra note 144, § 10.5.
  148. State Farm, 463 U.S. at 43, 52.
  149. See Office of the Federal Register, supra note 135 (noting that “the agency must base its reasoning and conclusions on the rulemaking record, consisting of the comments, scientific data, expert opinions, and facts accumulated during the pre-rule and proposed rule stages”). Stanton Wheeler et al., Do the “Haves” Come Out Ahead? Winning and Losing in State Supreme Courts, 1870–1970, 21 Law & Soc’y Rev. 403, 407 n.8 (1987).
  150. See Appel, supra note 1, at 123.
  151. See infra Part IV.
  152. Bureau of Justice Statistics, Compendium of Federal Justice Statistics, 2004, at 59 (2006), available at http://www.bjs.gov/content/pub/pdf/cfjs04.pdf.
  153. Bureau of Justice Statistics, Felony Statistics in State Courts, 2004, at 1 (2007), available at http://www.bjs.gov/content/pub/pdf/fssc04.pdf.
  154. Jessica N. Cohen, The Reasonable Doubt Jury Instruction: Giving Meaning to a Critical Concept, 22 Am. J. Crim. L. 677, 693 (1995).
  155. See Kent & Pendergrass, supra note 10, at 13; Schuck & Elliott, supra note 10, at 1007–09.
  156. Rasmusen et al., supra note 17, at 53.
  157. Id. at 53.
  158. Id. at 55.
  159. Id. at 55.
  160. Id. at 54–55.
  161. See id. at 54.
  162. Id.
  163. Id. at 58.
  164. See id. at 54.
  165. See id. at 59–60.
  166. Id. at 60.
  167. Id.
  168. See id. at 53–55.
  169. Id. at 61–62.
  170. See, e.g., Environmental Health Coalition, Who We Are, http://www.environmentalhealth.org/index.php/en/who-we-are (last visited Apr. 12, 2014) (providing an example of an environmental organization with a human health focus); Wilderness Soc’y, About Us, http://wilderness.org/about-us (last visited Apr. 12, 2014) (providing an example of an environmental organization with a preservation of natural landscape focus); Center for Biological Diversity, About the Center, http://www.biologicaldiversity.org/about/index.html (last visited Apr. 12, 2014) (providing an example of an environmental organization with a wildlife protection focus); 350.org, What We Do, http://350.org/about/what-we-do/ (last visited Apr. 12, 2014) (providing an example of an environmental organization with a climate change focus); The Access Fund, Mission and Vision, http://www.accessfund.org/site/c.tmL5KhNWLrH/b.4992345/k.BC76/Mission_and_Vision.htm (last visited Apr. 12, 2014) (providing an example of an environmental organization with a recreational focus).
  171. See Priest & Klein, supra note 14, at 28–29 (explaining that stakes of different parties to a case may differ when one party seeks an injunction or to establish precedent).
  172. Following the lead of previous authors, I will assume, as a rough approximation, that each outcome can be characterized simply as a win or a loss for the organization.
  173. See, e.g., Press Release, Nat’l Res. Def. Council, Environmental Victory in New York Harbor Dredging Court Battle (Oct. 23, 2007), available at http://www.nrdc.org/media/2007/071023a.asp (providing an example of environmental organizations utilizing a legal victory to advertise, gain publicity, and display effectiveness).
  174. 5 U.S.C. § 504(a)(1) (2006).
  175. Endangered Species Act of 1973, 16 U.S.C. § 1540(g)(4) (2006).
  176. Rasmusen et al., supra note 17, at 47, 55.
  177. Pinchot Inst. for Conservation, Ensuring the Stewardship of the National Wilderness Preservation System i n.1 (2001), available at http://www.wilderness.net/nwps/documents/brown_report_full.pdf.
  178. See id. at i–iii (explaining that many management actions are necessary to handle the growing size and complexity of the Wilderness System and the four wilderness agencies are required to organize their efforts).
  179. Administrative Procedure Act, 5 U.S.C. § 551(13) (2006) (defining “agency action”).
  180. Appel, supra note 1, at 111–12.
  181. Id. at 112–13.
  182. Id.
  183. Although Professor Appel characterized the “more protection” cases as being brought by an “environmental organization,” as they usually were, the classification itself was based on the issues in the cases, not the identity of the parties. Id.
  184. Professor Appel graciously and generously provided the author of this Article with a list and summary of the cases in his study. Peter A. Appel, Wilderness, the Courts and the Effect of Judicial Decisionmaking, 35 Harv. Env. L. Rev. 275, App. (2011) [hereinafter Appel App.].
  185. Id.
  186. Id.
  187. Id.
  188. Id.
  189. Id.
  190. Id.
  191. Id.
  192. Id.
  193. Id.
  194. Id.
  195. See Stupak-Thrall v. Glickman, 346 F.3d 579 (6th Cir. 2003) (seeking declaration that Crooked Lake was not part of the Wilderness Area); Stupak-Thrall v. United States, 89 F.3d 1269 (6th Cir. 1996), cert. denied, 519 U.S. 1090 (1997) (challenging prohibition of use of sailboats and houseboats on Crooked Lake); Stupak-Thrall v. Glickman, 988 F. Supp. 1055 (W.D. Mich. 1997) (challenging prohibition of use of gas-powered motor boats on Crooked Lake).
  196. See Fixel v. United States, 26 Cl. Ct. 353 (1992).
  197. Fixel v. United States, No. 87-4158, 1989 WL 260198, at *2 (D. Idaho 1989).
  198. Appel App., supra note 184.
  199. Id.
  200. Id.
  201. Id.
  202. Id.
  203. Wilderness Soc’y, Partner With Us, http://www.wilderness.org/partner-us (last visited Apr. 12, 2014); Wilderness Soc’y, Regional Offices, http://www.wilderness.org/article/regional-offices (last visited Apr. 12, 2014).
  204. Wilderness Soc’y, About Us, http://www.wilderness.net/NWPS/fastfacts (last visited Apr. 12, 2014) (stating that “the Wilderness Society has led the effort to permanently protect nearly 110 million acres of wilderness”).
  205. Friends of the Boundary Waters Wilderness, Issues, http://www.friends-bwca.org/issues/ (last visited Apr. 12, 2014); Friends of the Boundary Waters Wilderness, Visit, http://www.friends-bwca.org/visit/bwcaw (last visited Apr. 12, 2014).
  206. Appel App., supra note 184.
  207. Id.
  208. Id.
  209. Id.
  210. Id.
  211. Id.
  212. Stupak-Thrall v. Glickman, 988 F. Supp. 1055, 1058–59 (W.D. Mich. 1997).
  213. Fixel v. United States, 26 Cl. Ct. 353, 354 (1992).
  214. Kerr-McGee Corp. v. Hodel, 840 F.2d 68, 70 (D.C. Cir. 1988).
  215. Appel App., supra note 184.
  216. See Wilderness Act of 1964, 16 U.S.C. §§ 1133(c), (d)(3), (4), (5), & 1134 (2006) (regulating commercial enterprises, mining, and access to private interests within wilderness areas).
  217. Appel App., supra note 184.
  218. Id.
  219. Id.
  220. See supra Part V.
  221. See supra Part II.B.
  222. See supra Part II.B.
  223. Appel, supra note 1, at 112–13.
  224. See supra Part II.A.
  225. Appel, supra note 1, at 113 tbl.1.
  226. Id. at 122.
  227. See supra Part IV; see also Appel, supra note 1, at 113 tbl.1.
  228. See discussion supra notes 152–54 and accompanying text.
  229. Rasmusen et al., supra note 17, at 47, 75.
  230. See id. at 47, 72.
  231. Id. at 49.
  232. See supra Part IV.
  233. See supra text accompanying notes 172–75.
  234. Compare Federal Bureau of Investigation, Latest Crime Stats Released, http://www.fbi.gov/news/stories/2013/september/latest-crime-stats-released (last visited Apr. 12, 2014) (indicating that over one million violent crimes, and almost nine million property crimes were committed in the U.S. in 2012), with EPA, Compliance and Enforcement Annual Results 2010 Fiscal Year, http://epa.gov/compliance/resources/reports/endofyear/eoy2010/criminal/index.html (last visited Apr. 12, 2014) (indicating that the EPA opened 346 environmental criminal cases in 2010).
  235. See supra notes 203–07 and accompanying text (describing the number of decisions in which environmental organizations challenged the government’s management of wilderness areas and potential wilderness areas as “just the tip of the iceberg of the government decisions with which they are concerned”); supra text accompanying notes 227–29; Appel, supra note 1, at 123–25.
  236. See supra notes 84–87 and accompanying text.
  237. See supra notes 208–16 and accompanying text.
  238. See supra Part VI.A.
  239. See id.
  240. See supra Part V.
  241. Appel App., supra note 184.
Print this pageEmail this to someoneTweet about this on TwitterShare on Facebook

Comments are closed

Sorry, but you cannot leave a comment for this post.