The importance of the regulatory state in everyday American life is indisputable. The legal regime governing much significant economic and social policy in the United States comes not from statutes enacted by Congress, but from regulations promulgated by administrative agencies. Over the past forty years, for each statute enacted by Congress, federal agencies have promulgated an average of nineteen final rules.Footnote 1
How have bureaucratic agencies acquired so much lawmaking responsibility? How much control does Congress have over the federal bureaucracy? How easy or difficult is it to dismantle regulation? These questions have taken on heightened importance with the transfer of power from a Democratic to Republican presidential administration in 2017. Deregulation has been a top priority for the Trump administration, and the preservation of environmental and social regulations from the Obama years has been a top priority for many Democrats.Footnote 2
We show that it is difficult to roll back regulatory law—more difficult than many appointees of the Trump administration initially believed—and we argue that federal courts play a more significant role in determining change and stability in regulatory law than does Congress. Congress’s oversight capacity is weaker now than it was in 1970, even though the Code of Federal Regulations has roughly tripled in size since then.Footnote 3 In contrast, the federal courts have experienced a greatly expanding workload in regulatory law, as the number of administrative cases terminated on the merits in the circuit courts has more than doubled since 1992.Footnote 4 The capacity of courts to resolve the increasing number of administrative disputes has been facilitated by two important Supreme Court precedents. We argue that these precedents have taken a substantial body of bureaucratic policy making outside the scope of congressional control, and we show that even when agencies are acting consistently with congressional and presidential preferences, courts regularly decide against the agencies and block policy change.
Major strands of research in American politics do not and cannot account for the extent to which regulatory policy is increasingly shaped by the federal courts, not Congress. According to most scholars, the bureaucracy implements policies that are responsive to the president’s preferences, while Congress is believed to be the preeminent institution for controlling the bureaucracy, with courts playing a secondary enforcement role. Moreover, courts are generally assumed to act purely on judges’ personal policy preferences, not on precedent. These paradigms discourage questions about how legal precedents have affected congressional-bureaucratic relations.
To advance out argument, we focus on two landmark Supreme Court cases in administrative law: Motor Vehicles Manufacturers Association of the U.S. v. State Farm Mutual Automobile Insurance Co., 463 U.S. 29 (1983; hereafter State Farm) and Chevron U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 842 (1984; hereafter Chevron). State Farm resulted from rulemaking by the National Highway Safety Administration (NHTSA) on passive restraints in automobiles. Under the NHTSA rule, passive restraints—either airbags or detachable automatic belts—were required under the Carter administration in 1977 for all new cars by September 1983. However, the rule never took effect because the incoming Reagan administration thought it was economically harmful to the U.S. auto industry and rescinded it. Litigation eventually reached the U.S. Supreme Court, which in 1983 held that the rescission of the rule was arbitrary and capricious under the Administrative Procedure Act of 1946 (APA). The Court instructed NHTSA to present more reasoned analysis for jettisoning the rule. State Farm is important for several reasons, but perhaps most significant is that it established a precedent whereby executive agencies cannot rescind rules simply because administrators dislike them. The binding nature of State Farm was highlighted, once again, in the June 2020 Supreme Court ruling in Department of Homeland Security v. Regents of University of California, 591 U.S. __ (2020), when the Court struck down the Trump administration’s rescission of the Obama administration’s DACA protections, ruling that the administration had provided insufficient justification to warrant such actions.
Chevron dealt with conflicting interpretations of the Clean Air Act, also across the Carter and Reagan administrations. The Environmental Protection Agency (EPA) under President Carter implemented an interpretation favoring environmental interests, but the Reagan EPA chose another favoring manufacturing interests. The Supreme Court decided that both interpretations were admissible because the Clean Air Act was ambiguous. The Court then established a general precedent to deal with ambiguous statutes, holding that when statutes are ambiguous, an agency’s interpretation—providing it is reasonable—should receive deference from the courts. For this reason, Chevron is widely thought to transfer power away from Congress and to the executive branch when statutes are ambiguous (e.g., Eskridge and Ferejohn 1991–1992). Chevron is now the most frequently cited administrative law case of all time (Shane and Walker Reference Shane and Walker2014), and it has been described as a “kind of Marbury, or counter-Marbury, for the administrative state” (Sunstein Reference Sunstein1990).
State Farm and Chevron are well appreciated by administrative law scholars, but notably less so by political scientists. One of our purposes here is to begin to address this disconnect between administrative law and political science research. State Farm and Chevron also have important implications for President Trump’s deregulatory agenda, and we review the relevance of these seminal cases for a handful of important bureaucratic actions under the Trump presidency. We begin with an overview of the legal background and implications of the Chevron and State Farm decisions. We then review the dominant research perspectives on bureaucratic policy making, congressional control, and judicial decision making to assess the implications of Chevron and State Farm for existing scholarship. Finally, we present five brief case studies of how the federal courts have reacted to recent efforts by the Trump administration to roll back regulations, and we demonstrate that these legal precedents serve as binding constraints on agency policy making.
Chevron, State Farm, and Administrative Policy Change
Decided one year apart, State Farm and Chevron established important precedents for the judicial review of agency policy, particularly when policy changes across presidential administrations. We review the basic facts and opinions of each case.
Chevron
The key question in Chevron was what exactly was meant by a "stationary source" of air pollution in the Clean Air Act. The Act required permits for new or modified “stationary sources” of air pollution, and to obtain a permit, industries had to submit to a public hearing and demonstrate that the proposed stationary source would meet EPA standards. Firms also had to implement newer, “best available” technology and monitor the effect of emissions on local air quality. The permitting process was therefore both time-consuming and expensive, and many large manufacturing firms objected.
Prior to 1977, the EPA defined stationary source on a plant-wide basis. This definition treated all pollution-emitting devices within an entire plant as if they were encased in a single bubble. Under the bubble approach, as it was commonly called, plants could increase emissions from one device within the plant as long as they reduced emissions from another. Each plant under the bubble required a permit, but each separate device within the plant did not.
The bubble approach greatly simplified bureaucratic life for factories, for it allowed them to avoid complying with the tighter standards required of new sources of pollution. Environmental interests quickly recognized this loophole and objected. They favored instead a definition of stationary source that treated each pollution-emitting device within a factory as a separate source.
To try and balance producer and environmental interests, the EPA under the Carter administration allowed the bubble for both modified and new sources of pollution, but only in areas that had already attained national ambient air quality standards. This compromise abruptly ended with the change to the Reagan administration. Under Reagan, the EPA reversed course and promulgated a new rule that embraced the bubble concept entirely in both attainment and nonattainment areas. The Natural Resources Defense Council sued, and in August 1982, the D.C. Circuit Court struck down the Reagan EPA’s rule. Chevron U.S.A. then appealed to the Supreme Court, leading to the now-seminal case of Chevron U.S.A., Inc. v. Natural Resources Defense Council (1984).
Writing for the majority, Justice John Paul Stevens set forth the now-famous two-step procedure for reviewing agency rules. First, reviewing courts should determine whether the intent of Congress was clear, or whether the statute instead was “silent or ambiguous with respect to the specific issue.” If the statute was found to be ambiguous in step one, then the court would determine in step two whether the agency’s interpretation was “based on a permissible construction of the statute.” Regarding the Clean Air Act, the Court concluded under step one that the term “stationary source” was ambiguous, and then under step two held that the Reagan EPA’s bubble definition for nonattainment areas was a permissible construction.
Chevron is widely thought to have expanded the lawmaking authority of federal agencies at the expense of Congress. Silverstein (Reference Silverstein1994, 477) claims that Chevron has produced a “steady ratcheting of power away from Congress and toward the executive branch.” Robinson (Reference Robinson2013, 565) argues that ambiguity “effectively shifts lawmaking away from the legislature into less democratically accountable branches.” Farina (Reference Farina1989, 456) warned that the “danger” of Chevron “lies in its apparent obliviousness to the fundamental alterations it makes in our constitutional conception of the administrative state,” and Eskridge and Ferejohn (Reference Eskridge and Ferejohn1992, 533) argue that Chevron has contributed to an “overall shift of lawmaking authority from Congress to the President.”Footnote 5
State Farm
Operating on the basis of rulemaking authority established in the National Traffic and Motor Vehicles Safety Act of 1966, NHTSA proposed to amend Motor Vehicle Safety Standard 208 in July 1969 to require manufacturers to provide passive restraint technology in vehicles.Footnote 6 NHTSA believed that airbags, first patented in 1953, were a technically feasible passive-restraint technology and would save ten to twelve thousand lives each year (Mashaw and Harfst Reference Mashaw and Harfst1990, 85). Accordingly, NHTSA published a final rule in March 1971 that mandated air bags or non-detachable automatic belts by August 1975. Chrysler Corporation and other major automobile manufacturers sued, claiming that NHTSA was not empowered to force new technology upon the auto makers and could only issue standards for existing equipment. Upon review, the Sixth Circuit Court of Appeals sided with NHTSA on the issue of new technology but remanded the rule because NHTSA did not employ “objective” standards.
Department of Transportation Secretary William T. Coleman, appointed by President Ford, then struck a deal with auto manufacturers to produce a demonstration fleet of a half-million automobiles with various types of passive restraints in order to gauge public reaction and gain support. In return, Coleman promised not to issue a passive restraint rule. Coleman’s tenure as secretary was short-lived, however, and with the election of President Jimmy Carter, the new Secretary of Transportation, Brock Adams, promptly reversed Coleman’s policy and issued a new final rule in June 1977 requiring passive restraints in all passenger vehicles beginning in 1981.
Adam’s tenure as secretary was also short-lived, however, and with the election of Ronald Reagan, NHTSA abruptly reversed course once again. NHTSA reopened the rulemaking process for Standard 208 in February 1981, and two months later announced a one-year delay in the application of the standard. Finally, after more than seven months of delay, NHTSA rescinded the passive restraint standard altogether.
State Farm Insurance Company promptly sought judicial review of the agency’s rescission, and the D.C. Circuit Court of Appeals ruled in its favor, holding that the rescission of the rule was arbitrary and capricious. Upon appeal, the U.S. Supreme Court largely agreed with the circuit court in State Farm in a 6–3 decision. Since NHTSA had previously decided that airbags comported with the mandate of the Motor Vehicles Safety Act, the Court concluded that the agency could not simply ignore that technology and rescind the rule without more reasoned analysis. The majority opinion held that an agency “must examine the relevant data and articulate a satisfactory explanation for its action,” and that “an agency changing its course by rescinding a rule is obligated to supply a reasoned analysis for the change beyond that which may be required when an agency does not act in the first instance” (State Farm, 2867).
More broadly considered, State Farm established a precedent that the rescission of a rule must be judged by the same arbitrary and capricious standards governing rule promulgation, and that agencies cannot change policies simply in response to changes in political preferences. To change policy, agencies must begin the process of notice-and-comment rulemaking de novo. In State Farm, therefore, the U.S. Supreme Court imposed a check on policy swings in rulemaking across presidential administrations.
Beyond State Farm and Chevron
The Supreme Court has extended and refined State Farm and Chevron in several instances. In United States v. Mead Corp., 533 U.S. 218 (2001), the Court held that agencies’ interpretive rulings—advice or guidance about what an agency thinks an existing regulation means—do not qualify for judicial deference under Chevron. In Auer v. Robbins, 519 U.S. 452 (1997), the Supreme Court held that agencies should be accorded Chevron deference when interpreting ambiguities in their own regulations.Footnote 7 And in FCC v. Fox Television Stations, Inc., 556 U.S. 502 (2009), the Supreme Court addressed an unresolved issue from State Farm as to whether administrative policy reversals should be subjected to a higher level of scrutiny, or to more reasoned analysis, than that required for the initial adoption of a policy. A narrow 5–4 majority held that administrative policy reversals did not require heightened scrutiny, concluding that “State Farm neither held nor implied that every agency action representing a policy change must be justified by reasons more substantial than those required to adopt a policy in the first instance.”
Together, these precedents establish a legal regime for bureaucratic decision making. The Supreme Court has established broad parameters for how federal agencies create, interpret, and modify regulations and guidance. Agencies may not change policy for purely partisan reasons, and they must provide reasoned analysis for any change. Agencies are free to create new policy of their choosing when Congress has failed to be explicit, but only when they act in accordance with existing law.
The development of this judicially constructed legal regime stands in sharp contrast to congressional efforts to supervise bureaucratic behavior. Congress has incorporated only three significant substantive amendments to the Administrative Procedure Act since its original passage in 1946,Footnote 8 and although Congress often strategically includes provisions for judicial review in individual statutes (e.g., Shipan Reference Shipan2000), Congress has not created a general and comprehensive framework for regulating agency action similar to what the courts have established.
Research on Congress, Bureaucracy, and Courts
Our argument in this article is that contemporary legal precedents ensure that a substantial body of bureaucratic policy making happens outside of the scope of congressional control. While Congress engages in a wide range of oversight activities, including hearings (Aberbach Reference Aberbach1991), sending letters to agencies (e.g., Ritchie and You Reference Ritchie2019), and enacting limitation riders and other budgetary tweaks (McDonald Reference McDonald2010), these tools cannot easily induce an agency to promulgate particular policies, especially when the agency has been delegated broad and ambiguous authority.Footnote 9 Clearly, Congress can exert direct oversight through the Congressional Review Act (CRA), but the practical difficulties of invoking the CRA—both the House and Senate must pass a disapproval motion subject to presidential veto—is seldom feasible and is identical to what would obtain if Congress passed a new law that overrode an existing regulation altogether.Footnote 10 Likewise, even though the president wields influence over the bureaucracy through the appointment power and other staffing decisions (Moe Reference Moe, Chubb and Peterson1985, Lewis Reference Lewis2008), together with the ability to set the regulatory agenda and manage it with OIRA (Farber and O’Connell Reference Farber and O’Connell2014, Wiseman Reference Wiseman2009), once a rule has been promulgated, it cannot be easily undone under State Farm.
Another factor limiting congressional oversight is the political opportunity cost of oversight. Congressional oversight of the bureaucracy requires an expenditure of resources that could be allocated to reelection considerations or to other legislative initiatives, and oversight is not usually a high priority of incumbent legislators. As Richard Fenno (U.S. Congress, 1973, 15) once put it: “When the incentive isn’t there, you are simply not going to get oversight.” Only in the most salient and politically fraught issues is Congress willing to go on record to defend or chastise an agency action, and as a result, Congress is silent on most regulatory matters.
The relevance of State Farm and Chevron to bureaucratic behavior finds little support in American political science research on Congress, the bureaucracy, and courts. Research has evolved along two general themes, or paradigms, neither of which accounts for the role that these two important precedents might play. One theme is that Congress effectively controls the bureaucracy, and the other is that judges’ decisions are adequately explained, not by legal precedent, but by their personal policy preferences.
Congress and the Bureaucracy
Despite concerns that Congress has lost control of public policy because of excessive delegation of lawmaking authority to the bureaucracy (e.g., Lowi, Reference Lowi1969), congressional scholars have concluded that delegation is appropriate and efficient, and that the bureaucracy is held in check through a variety of administrative procedures and institutional arrangements. Fiorina (Reference Fiorina1977, Reference Fiorina, Dodd and Oppenheimer1981), for example, argues that Congress benefits politically from a large and inefficient bureaucracy because of the opportunities it provides for constituency casework, and that agencies are effectively controlled by congressional committees whose members are keenly interested in the pork-barrel benefits agencies provide to their districts. Weingast and Moran (Reference Weingast and Moran1983) advanced a notion of “congressional dominance,” wherein agencies operate in alliance with congressional committees with similar policy interests, and where agency policy changes in response to changes in the ideological preferences of legislators on the agency’s oversight committee.Footnote 11 McCubbins and Schwartz (Reference McCubbins and Schwartz1984) have argued that Congress controls agencies through “fire alarms” sounded by citizens and organized interests when agency policy deviates from congressional intent.
The most prominent theory of congressional influence over the bureaucracy is that proposed by McCubbins, Noll, and Weingast (Reference McCubbins, Noll and Weingast1987, Reference McCubbins, Noll and Weingast1989; a.k.a, McNollGast). They argue that administrative procedures established by Congress—most notably in the Administrative Procedure Act—allow Congress to control the bureaucracy without engaging in direct oversight activities. Administrative procedures such as notice-and-comment require agencies to publish proposed rules and allow citizens and interested groups opportunity to comment. Notice and comment theoretically mitigates much of the information asymmetry between Congress and the bureaucracy, and it ensures that agencies are subjected to the same political forces faced by Congress. The procedure “create[s] a decisionmaking environment [in the agency] that mirrors the political circumstances that gave rise to the establishment of the policy” (McCubbins, Noll, and Weingast Reference McCubbins, Noll and Weingast1987, 255), and it helps ensure that interests that successfully organize a winning coalition in Congress are also likely to dominate agency proceedings—a process McNollGast refer to as political “deck-stacking.” Through deck-stacking, political officials “cause the political environment in which an agency operates to mirror the political forces that gave rise to the agency’s legislative mandate” (McCubbins, Noll, and Weingast Reference McCubbins, Noll and Weingast1987, 262).
Once the deck is appropriately stacked, congressional control will operate on “autopilot,” a process where “policy decisions made by the agency evolve as the composition of participating groups changes … [so] that agencies respond to changes in their environment even if the politicians have not first spotted these changes” (McCubbins, Noll, and Weingast Reference McCubbins, Noll and Weingast1987, 263-264). A good-faith interpretation of McNollGast’s concept of autopilot is that agency policy making occurs as follows: if empowered interest A expresses its preferences for policy a in time t, we would expect the agency to promulgate policy a. If empowered interest B emerges in time (t+1), however, an agency making policy by autopilot would be expected to promulgate a new policy b. Under State Farm, however, agencies cannot reverse course so quickly and easily. They must generate substantively compelling evidence to justify policy change.Footnote 12
Theories of congressional control effectively black-box judicial decision making in that courts are assumed to be mere enforcers of congressional policy preferences.Footnote 13 Chevron and State Farm, however, changed the procedural game between Congress and agencies. Viewed in the context of McNollGast, State Farm imposes a substantial barrier to the realization of agency policy making by autopilot. Autopilot rulemaking implies that agencies will change policies in response to changes in the preferences of their political principals, even without new statutory authority, and even without reasoned analysis. Chevron undercuts the congressional control paradigm because it fundamentally shifts lawmaking authority away from Congress. In the context of McNollGast, Chevron undermines the purpose of deck-stacking. Given ambiguous statutes and Chevron deference, agencies do not need to comply with a politically stacked deck.
Judicial Decision Making
A limited but important body of scholarship engages with the relationship between the judiciary and administrative agencies. Foundational work by Crowley (Reference Crowley1987) and Sheehan (Reference Sheehan1990, Reference Sheehan1992) analyzes the incidence and determinants of win rates of different agencies before the Supreme Court, while more recent work by Miller and Curry (Reference Miller and Curry2013) explores the conditions under which judges will rely on their own expertise in making their decisions at the appellate level. Thrower (Reference Thrower2017) explores how the Supreme Court’s ideological preferences appear to influence its decisions on executive policy making, while Epstein and Posner (Reference Lee and Posner2016) point to a “loyalty” effect among individual justices who appear to favor those presidents who appointed them. Johnson (Reference Johnson2014) explores these topics at the state level, demonstrating that state courts are more likely to uphold agency decisions when governors exert significant control over agency policy making.
While these and other works provide insights regarding the relationship between the judicial branch and administrative agencies, they are not well positioned to inform us about the impact of judicial doctrine on contemporary agency rulemaking. Only some of these studies directly address agency rulemaking, and most consider only Supreme Court decision making, even though most judicial decisions over agency policy making occur in the district and circuit courts. Additionally, those works that do speak directly to agency policy making and the role of judicial doctrine (e.g., Sheehan Reference Sheehan1990, Reference Sheehan1992; Crowley Reference Crowley1987; Humphries and Songer Reference Humphries and Songer1999) draw on data generated largely before Chevron and State Farm.
Most research on judicial decision making focuses on judicial decisions generally rather than on executive agencies, and most focus heavily on judges’ personal policy preferences, or ideology (Segal and Spaeth Reference Segal and Spaeth1996, Reference Segal and Spaeth2002) as explanations of behavior. Legal reasoning through precedent and analogy is seldom considered a viable explanation of judicial behavior (however, see Brenner and Stier Reference Brenner and Stier1996). One prominent judicial scholar has written, “I can think of no political scientists who would take … precedent[s] as good explanations of what the justices do in making decisions” (Caldeira Reference Caldeira1994, 485).
There is some empirical evidence of the importance of precedent, however. Barnett, Boyd, and Walker (Reference Barnett, Boyd and Walker2018) found that ideology and partisanship were significantly less important explanations of judicial decisions in cases where circuit courts applied Chevron than in cases they did not. In an analysis of 1,381 unique opinions across all thirteen federal circuit courts between 2003 and 2013, the authors found that the most conservative three-judge panels agreed with liberal agencies’ interpretations 66% of the time with Chevron deference, but only 18% without.
While few such studies directly test for both attitudinal and precedential effects on judicial outcomes, research by Black and Owens (Reference Black and Owens2009, 1072) and Yates, Cann, and Boyea (Reference Yates, Cann and Boyea2013), for example, suggest that factors other than partisan preference can influence judicial decision making. Black and Owens (Reference Black and Owens2009, 1072) found that in the presence of certain legal factors, justices “sacrifice their policy goals,” and that even though justices’ policy preferences are strong predictors of their decisions, “law constrains them from acting on policy goals alone.” Yates and colleagues found that attitudinal factors emerge as less important predictors of judicial outcomes after controlling for litigants’ strategic decisions to sue or settle in Supreme Court decisions on economic issues in the environment from 1953–2000.
Formal Models
A large formal literature on bureaucratic policy making builds upon a principal-agent framework to understand the conditions under which a legislature might choose to delegate policy-making authority to a bureaucratic agency. Influential models by by Bawn (Reference Bawn1995), Bendor and Meirowiz (Reference Bendor and Meirowitz2004), Epstein and O’Halloran (Reference Epstein and O’Halloran1994, Reference Epstein and O’Halloran1999), Gailmard (Reference Gailmard2002), and Volden (Reference Volden2002), for example, demonstrate that delegation is both rational and efficient for legislatures under a broad range of circumstances.
Principal agent models have yielded valuable insights into the relationship between legislators and bureaucrats, but they have not addressed the implications of Chevron and State Farm for legislative or bureaucratic behavior. It is unclear, for example, whether it is rational for a legislature to delegate policy-making authority to an agency under an ambiguous statute (however, see Wright Reference Wright2010). It is also unclear how a time-lag between legislative delegation and bureaucratic action influences policy making (and the initial decision to delegate authority). An implicit assumption of these models is that there is no substantial time-lapse between the incidence of legislative-delegation and agency-policy making. As illustrated by the passive restraints case, however, long time-lags are not hypothetical situations (see also Potter Reference Potter2017, Reference Potter2019). One wonders whether the Congress that delegated the Department of Transportation rulemaking authority in 1966 would be content to have the NHTSA of the late 1970s and early 1980s implementing that authority.Footnote 14
Only a couple of formal models have explicitly addressed State Farm and Chevron. One is Cohen and Spitzer (Reference Cohen and Spitzer1994), who consider why the Supreme Court adopted the Chevron doctrine, and another is Gely and Spiller (Reference Gely and Spiller1990), who seek to explain State Farm in a rational choice framework. Cohen and Spitzer’s explanation for why the Supreme Court increased deference to agency rulemaking in Chevron is that administrative agencies became more conservative between the Carter and Reagan administrations, which in turn brought Reagan-era agencies more ideologically into line with the Supreme Court. Thus, Cohen and Spitzer propose that a conservative court was willing to defer to a conservative agency, but not to a liberal one. The empirical work by Barnett, Boyd, and Walker (Reference Barnett, Boyd and Walker2018), discussed earlier, does not support this hypothesis.
Gely and Spiller (Reference Gely and Spiller1990) propose an ideological explanation for the Supreme Court’s State Farm opinion. They argue that the Court’s reason for rescinding NHTSA’s rule in 1983 was to move policy on automobile safety slightly closer to the Supreme Court’s ideal point. Their assumption of an ideologically motivated Supreme Court in this case is quite strong, however. Internal memoranda circulated among the justices during the opinion-writing phase of State Farm suggest otherwise. Justice Byron White, who wrote the majority opinion, was quite explicit that the Court’s interest was not in the policy outcome per se. In a note to Justice Lewis F. Powell, White explained that “the reason for remanding to the agency is to insist on a better explanation of its decision rather than any disagreement with the result it has reached.”Footnote 15
Within the formal literature, a foundational model of the strategic interaction of an agency, a legislature, and a court is that of Ferejohn and Shipan (Reference Ferejohn and Shipan1990), who assume that courts are motivated by ideology, and seek optimal policy outcomes. A key takeaway from Ferejohn and Shipan is that judicial review can prevent regulatory policy from drifting away from the legislative median and toward the agency’s ideal point (Ferejohn and Shipan Reference Ferejohn and Shipan1990, 11 and figure 2). This result occurs because courts are expected to overturn agency actions that are suboptimal for both the court and the legislature. But if courts followed Chevron and deferred to agencies, then this result would no longer hold. The model, however, has no provision for courts to review agency actions on a basis other than policy satisfaction.
The Ferejohn–Shipan model also cannot account for State Farm. For some important configurations of preferences, the model does not explain why courts would ever overturn agencies. For example, in situations where Congress and an agency have identical or nearly identical preferences, say on the conservative end of the ideological spectrum, and where the status quo policy is at the other end of the spectrum, and where the court’s preferred policy is somewhere in the middle, Ferejohn and Shipan predict that the court simply accepts the agency’s policy. The legislature has the agency’s back, so to speak, and because (by Ferejohn and Shipan’s assumptions) the legislature is the final mover in the game, it is futile for the court to move policy away from the agency.
The configuration of institutional preferences in this example is non-trivial, for it reflects the situation where Obama-era rulings are pitted against agencies of the Trump administration and a Republican-controlled 115th Congress. We present several case studies of policy making in which courts have overturned agencies under the logic of State Farm, contrary to the prediction of Ferejohn and Shipan.
Cases in Bureaucratic and Judicial Policy Making
In its haste to undo Obama-era regulations, the Trump administration has run headlong into State Farm in the federal courts, which have vacated numerous efforts by Trump-appointed administrators to dismantle social and environmental regulations. We summarize four of these cases that were decided early in the Trump administration, and we review a fifth case—net neutrality—that was decided by the federal court of appeals in October 2019 and is likewise profoundly related to Chevron and State Farm. In each of these cases, we demonstrate that judicial doctrine—Chevron and State Farm in particular—channels agency policy making in ways that make it difficult, if not impossible, for agencies to change policy simply for political expediency.
On its face, one might argue that our consideration of this collection of cases constitutes a poor test of our main argument. After all, if Chevron and State Farm are so influential, why would agencies ever promulgate rules that are inconsistent with these precedents, only to lose in court? Indeed, Walker (Reference Walker2014, 716) demonstrates that 94% of surveyed agency rule drafters reported that they were aware of Chevron, and 90% reported that Chevron influenced how they drafted their agency’s rules.Footnote 16 Hence, if these precedents have force, one might expect them to exhibit the “second face of power” (e.g., Bachrach and Baratz Reference Bachrach and Baratz1962), so that courts would never invoke them because agencies would have already accounted for them in their rulemaking process.
Our set of cases, however, provide a unique opportunity to test the impact of Chevron and State Farm, as they originated in the unusual context of the Trump presidency. President Trump has been anxious to roll back many agency actions promulgated by the Obama administrations, and he also presides over an administration that has seen the departure of many seasoned civil servants, substantial delay in filling many influential positions across agencies, and the appointment of political ideologues instead of pragmatists. In such an environment, it is not surprising that some agencies have taken actions that would not have occurred under more typical administrations. In a sense, then, our set of cases are policy “mistakes” that allow us to observe judicial decisions that might otherwise be hidden.
Our rulemaking cases originated during the Republican-controlled 115th Congress, which was closely aligned ideologically with agencies of the Trump administration and presumably supportive of their actions. Hence, if Congress was fully in control of agency policy making, and if judicial precedent was irrelevant, then the agencies’ policies in the following cases should have prevailed, at least when reviewed by ideologically supportive judges. Yet Congress and the agencies were typically losers to judicial precedent. Moreover, in none of these cases is there any evidence of autopilot policy change.
Atlantic Coast Pipeline
The Atlantic Coast Pipeline is a proposed 600-mile natural gas pipeline extending from West Virginia to North Carolina, traversing twenty-one miles of national forest land (the George Washington and Monongahela National Forests) and crossing the Appalachian National Scenic Trail. Its construction requires an environmental impact statement from the Federal Energy Regulatory Commission (FERC) that must be approved by the Forest Service and that must comply with the Forest Service’s guidelines for forest management. Under authority of the National Forest Management Act, the Forest Service promulgates a Forest Planning Rule to manage impacts to groundwater, soils, threatened and endangered species, landslides, and slope failures. The Atlantic Coast Pipeline needed to conform to the forest management plans for the George Washington and Monongahela National Forests
In September 2015, Atlantic Coast Pipeline, LLC (Atlantic) filed a formal application with FERC to build and operate the pipeline, and in November applied for permits from the Forest Service to construct and operate the pipeline across the national forests, as well as across the Appalachian Trail. The Forest Service and the National Park Service manage separate parts of the trail, and Atlantic could have routed the pipeline across the trail in either agency’s jurisdiction. While crossing the Appalachian Trail on lands managed by the National Park Service could have circumvented the national forests, it would have required congressional approval as the Park Service is not authorized to approve pipeline construction. Thus, to avoid the legislative process, Atlantic’s proposed route was through the national forests and across the trail on lands managed by the Forest Service.
FERC completed its draft environmental impact statement in December 2016, at which time the Forest Service began reviewing Atlantic’s construction plans and environmental impacts. One concern of the Forest Service was the effectiveness of Atlantic’s construction techniques for maintaining stability on steep slopes. The Forest Service requested ten site-specific stabilization designs with full specifications and data on their effectiveness. Additionally, the Forest Service registered concerns about erosion control, the re-introduction of native plants, and the biological impact on the habitat of the little brown bat. These concerns were communicated to Atlantic through letters, meetings, and formal comments between December 2016 and April 2017.
Throughout this time, the chief of the Forest Service was Thomas Tidwell, a holdover from the Obama administration. The chief typically serves at the pleasure of the Secretary of Agriculture, but the Department of Agriculture, which has jurisdiction over the Forest Service, was without a secretary until April 24, 2017, when the U.S. Senate finally confirmed George “Sonny” Perdue. Perdue’s confirmation marked a turning point in the Forest Service’s policies toward the Atlantic Coast Pipeline. On May 14, 2017, the Forest Service informed Atlantic that it would require only two of the ten site-specific stabilization designs, and in July, the Forest Service exempted Atlantic from thirteen standards in its forest plan. FERC issued a permit for construction of the pipeline in October 2017, and the Forest Service granted a special-use permit to begin construction across the national forests and the Appalachian Trail in January 2018.
The Forest Service’s dramatic policy reversal prompted Cowpasture River Preservation Association, together with six other environmental groups, to file suit against the Forest Service in February 2018. Eleven months later, the Fourth Circuit Court of Appeals vacated and remanded the case to the agency. Leaning on State Farm, the circuit court held that the agency’s decision was arbitrary and capricious. The circuit panel, consisting of two Obama-appointed judges and the circuit’s chief judge appointed by George W. Bush, concluded unanimously that the Forest Service, by exempting Atlantic from the forest plan standards, had “entirely failed to consider an important aspect of the problem.” The court observed that
The lengths to which the Forest Service apparently went to avoid applying the substantive protections of the 2012 Planning Rule—its own regulation intended to protect national forests—in order to accommodate the ACP project through national forest land on Atlantic’s timeline are striking, and inexplicable. Accordingly, we conclude that the Forest Service’s determination that the GWNF and MNF Plan amendments would not have substantial adverse effects on the forests was arbitrary and capricious.
This case illustrates how a federal agency, in its haste to overturn rules promulgated under a previous administration of a different political party, failed to consider important aspects of the problem and acted arbitrarily and capriciously. The Republican-controlled 115th Congress, which likely supported the construction of the pipeline, was unable or unwilling to preserve the Forest Service’s ruling on the pipeline. The circuit court determined the policy outcome under State Farm, which proved to be outside the scope of congressional control.
Following its defeat in the circuit, Atlantic Coast Pipeline petitioned to have its case heard by the Supreme Court, which held oral argument in February 2020. The question presented to the Supreme Court was much narrower than that before the circuit court, however. The Supreme Court was asked only to decide whether the Forest Service has authority under the Mineral Leasing Act to grant rights-of-way through national forest lands traversed by the Appalachian Trail. The Fourth Circuit concluded that it did not, but that question constituted only a small portion of the circuit’s opinion. The bulk of the opinion—fifty-one of sixty pages—dealt with the Forest Service’s reversal of environmental standards, which were not contested in ACP’s certiorari petition.
Teen Pregnancy Prevention Program
Congress created the Teen Pregnancy Prevention Program (TPP) in 2010 to encourage and fund scientifically valid research on teen pregnancy. The program was originally established in the Consolidated Appropriations Act of 2010 (Public Law 111-117), and since 2010 Congress has appropriated roughly $100 million annually to public and private entities that work with teen pregnancy programs. The 2010 Act specified that the money was “to fund medically accurate and age appropriate programs that reduce teen pregnancy” and for “replicating programs that have been proven effective through rigorous evaluation to reduce teenage pregnancy.” Although the Trump administration threatened to eliminate the program entirely during budget negotiations in 2017, it instead pursued a strategy of re-allocating money from existing programs to abstinence-only programs.
Under the Obama administration in 2015, the Office of Adolescent Health, the agency within the Department of Health and Human Services (HHS) charged with administering grants, made eighty-one new TPP program awards. The “notice of award” for each grant specified a five-year project period from July 1, 2015 until June 30, 2020. But in July 2017, three years into the project period, HHS under the new Trump administration informed all grantees that their project periods would end on June 30, 2018, two years earlier than originally specified.
This action triggered four separate lawsuits alleging that HHS had acted arbitrarily and contrary to statute in terminating the grants. Four separate district courts ruled in favor of the plaintiffs, finding that HHS acted arbitrarily and capriciously as delineated under State Farm. In the first of the four cases, the district court granted a permanent injunction blocking HHS’s termination of grants to Planned Parenthood affiliates. The court concluded that HHS had failed to satisfy State Farm’s dictum that agencies must “examine the relevant data and articulate a satisfactory explanation for its action.” The court explained that “defendants [HHS] do not offer a rational connection between the facts and the choice made, but merely articulate policy concerns and their own discretion to terminate the program for whatever reason. This reasoning or lack thereof is arbitrary and capricious.”
In a second case with Healthy Teen Network and the City of Baltimore as plaintiffs, the court declared that “the essential question of the plaintiffs’ arbitrary and capricious challenge may be posed as such: Notwithstanding HHS’s broad discretion to determine which organizations will receive funding, when it decided to end the plaintiffs’ funding did it, among other things, consider teenage pregnancy prevention at all?” The court found that it had not. State Farm holds that a court must consider whether an agency’s decision was based on “consideration of the relevant factors,” and because teenage pregnancy was clearly a relevant factor, the court concluded that HHS’s decision was arbitrary and capricious.
The court again relied upon State Farm in the third of the four cases, observing that “under the most elementary precepts of administrative law, an agency has no choice but to provide a reasoned explanation for its actions.” In vacating HHS’s decision to terminate the project early, the court asserted: “The most striking thing about the agency action that Plaintiffs challenge in this case is the fact that HHS provided no explanation whatsoever for its decision.”
The district judges in these three cases were appointed by presidents Clinton or Obama.
The district judge in the fourth case, Judge John Coughenour of the Western District of Washington, was appointed by Ronald Reagan, but he too vacated HHS’s action on the grounds that the agency did not supply a sufficient explanation for its actions. Judge Coughenour concluded that “HHS’s failure to articulate any explanation for its action, much less a reasoned one based on relevant factors, exemplifies arbitrary and capricious agency action meriting reversal.”
Like the case of the Atlantic Coast Pipeline, this case illustrates the roadblock that State Farm presents to agencies wishing to change regulatory policy from a previous administration. The changes to TPP funding were overseen by HHS’s assistant secretary for health, Valerie Huber, an ideological abstinence-only advocate who previously headed the National Abstinence Education Association, and who had no previous agency policy-making expertise. The district courts made it quite clear, however, that ideological preferences were insufficient for changing previous agency policy, and that reasoned explanations had to be provided. The Republican Congress could have included an appropriations rider to direct money exclusively to abstinence-only programs, or axed the program entirely, but it did not.
Royalties from Leasing Federal Lands
The Mineral Leasing Act (MLA) of 1920 allows the federal government to lease public and Indian lands to private companies for fossil-fuel exploration, development, and production. The MLA delegates authority to the Department of Interior to establish and collect royalties, and within DOI, the Office of Natural Resources Revenue (ONRR) manages the setting and collection of royalties. ONRR’s regulations governing royalties on oil, gas, and coal are known collectively as the “valuation rule.”
In 2007, following concerns that ONRR’s valuation rule, which had been in place since 1988, was out-of-sync with the domestic market in that some commodities, particularly coal, were undervalued, the Department of Interior began a review process that carried over to the Obama administration. In 2011, ONRR published two advanced notices of proposed rulemaking inviting suggestions for new valuation methodologies. These notices commenced a five-year rulemaking process to update the agency’s valuation rule. Finally, in January 2015, ONRR proposed a new valuation rule, and following an extended commenting and review period, issued a final rule in July 2016. The new rule was to take effect on January 1, 2017 and was estimated to increase royalty collections by between $71.9 million and $84.9 million.
With the change to the Trump administration, industry groups—especially coal interests—began petitioning ONRR to postpone and ultimately repeal the new valuation rule. Bowing to industry pressure, ONRR published a Postponement Notice in the Federal Register in February 2017, and in April 2017 published a proposal to repeal the valuation rule in its entirety. Four months later, in August 2017, ONRR published the final repeal, which reinstated pre-1988 royalties and leasing provisions. Thus, within a matter of a few months, the Trump ONRR rescinded a rule that had taken over five years to promulgate.
The states of California and New Mexico sued ONRR for injunctive relief in October 2017, contending that ONRR failed to provide a “reasoned explanation” for the policy change. The district court reviewed Interior’s action under the arbitrary and capricious standard of the Administrative Procedure Act, citing State Farm as the standard for judging whether an agency’s action is indeed arbitrary and capricious.
Following precedents established in State Farm and related cases, the district court concluded that “it was incumbent upon it [ONRR] to provide a reasoned explanation as to why the industry concerns it previously rejected [during the five-year rulemaking period preceding issuance of the final rule in 2016]—as well as its prior findings in support of adopting the Valuation Rule—now justified returning to the pre-Valuation Rule regulatory framework.” The court went on to assert that “neither Federal Defendants nor Industry Intervenors identify where in the Final Repeal or elsewhere in the record the ONRR provided such an explanation.” The district judge, appointed by George H.W. Bush, wrote that “in repealing the Valuation Rule, the ONRR completely contradicts its prior findings.” The judge went on to cite Fox, noting that “Fox makes clear that when an agency seeks to disregard facts underlying the original rule, it must provide “a more detailed justification than what would suffice for new policy created on a blank slate.” The court vacated the repeal.
This case closely resembles the circumstances of State Farm, where the agency quickly rescinds a rule on policy grounds without a good explanation. Experienced hands at ONRR should have known better, but it must be remembered that the Department of Interior at the time was headed by Ryan Zinke, a two-term member from the House who had spent most of his life as a Navy SEAL. ONRR’s repeal of the valuation rule may have had the 115th Congress’s blessing, but there was little Congress could do to ensure the repeal. The rescission occurred so quickly that there was no opportunity for new legislation, an appropriations rider, or a hearing. Even though the Congress may have been fully supportive of the agency, the district court had the final say.
Chemical Disasters
Motivated by the chemical disaster in Bhopal, India, in 1984 that killed and injured thousands, Congress charged the EPA to design regulations to prevent accidental releases in amendments to the Clean Air Act of 1990. An “accidental release,” according to the Act, is “an unanticipated emission of a regulated substance or other extremely hazardous substance into the ambient air from a stationary source.” In 1996, following a six-year rule-making process, the EPA promulgated “accidental release” prevention regulations.
Two high-profile chemical disasters in 2013 revealed the inadequacy of EPA’s accidental release regulations. In April, the explosion of a fertilizer plant in Texas killed fourteen people, including twelve first responders, and caused $230 million of damage. Two months later, a chemical plant in Louisiana exploded, killing two workers and injuring others. These disasters provided momentum to environmental groups that had been pushing for stronger regulations. In March 2016, the EPA proposed amendments to its accidental release prevention regulations, and then in a classic case of midnight rulemaking, promulgated a final rule in January 2017 just before President Trump’s inauguration. The Chemical Disaster Rule, as it was called, had an effective date of March 2017.
Following a change in presidential administration, the EPA under Scott Pruitt delayed the effective date of the new Chemical Disaster Rule on three separate occasions. The EPA initially delayed the regulation for one week under the general “crack of dawn” suspensions typical of an incoming administration (O’Connell Reference O’Connell2008). It then imposed another delay for ninety days, and following that, EPA promulgated a rule delaying the effective date for another twenty months, until February 2019. The EPA’s rationale for the twenty-month suspension was “to conduct a reconsideration proceeding and to consider other issues that may benefit from additional comment.”
In enacting the twenty-month delay, the EPA conveniently ignored a critical statutory requirement. When Congress amended the Clean Air Act in 1990, it explicitly required the EPA to implement new regulations within a “period not to exceed three months.” This limitation was imposed as a reaction to EPA’s historically slow pace in regulating environmental hazards. EPA’s disregard for this statutory provision mobilized a coalition of environmental and community groups, including Air Alliance Houston, together with various states and labor unions, to petition the courts for review.
The DC Circuit Court of Appeals found that EPA’s delay action made a “mockery of the statute.” According to the court, EPA’s justification for delay—to reconsider the rule given concerns raised by industry groups—was an insufficient reason. Judges on the DC Circuit acknowledged that the EPA had authority to conduct new rulemaking to amend the chemical disaster rule of the Obama EPA, but they concluded that the agency did not have authority to delay a final rule “merely because EPA is considering revising it.”
For this reason, the court found that EPA’s delay was arbitrary and capricious. The court found that the agency did not “examine the relevant data and articulate a satisfactory explanation for its action,” as required by State Farm, and that the agency had not provided a reasoned explanation “for disregarding facts and circumstances that underlay or were engendered by the prior policy,” as required by FCC v. Fox Television Stations (2009). In vacating the EPA’s delay of the Chemical Disaster Rule, the court concluded that “EPA’s explanations for its changed position on the appropriate effective and compliance dates are inadequate under Fox and State Farm and therefore arbitrary and capricious.”
Courts often allow agencies to delay rules, but as this case illustrates, agencies must present good reasons. Just as agencies cannot rescind a rule simply for political reasons, they also cannot arbitrarily delay the implementation of a rule solely for political reasons, especially when the statute says otherwise. Congress did not publicly register a policy position on the delay of this rule through hearings or an appropriations rider that would have precluded funding for implementation of the chemical disaster rule. However, if congressional Republicans articulated a preference to the agency through letters or ex parte communications, they failed to prevent the court from imposing judicial precedent.
Net Neutrality
In January 2018, the Federal Communications Commission finalized an order to roll back net neutrality rules established in 2015 during the Obama administration. Published just 367 days after President Trump’s inauguration, Restoring Internet Freedom (Federal Communications Commission 2018) called for a major shift in the FCC’s policies on the Internet. The FCC’s about-face on net neutrality is a classic case of policy change across presidential administrations, and the decision of the circuit court in this case illustrates how Chevron can engender policy instability, but also how State Farm promotes policy stability.
The controversy over net neutrality is rooted in a decision by the FCC in 1970 to distinguish basic voice transmission over telephone lines from data transmission. Congress followed this distinction when it overhauled the Communications Act in 1996 and defined both “telecommunications services” and “information services.” Telecommunications services were subject to Title II regulation, but information services were not.
This hands-off approach to regulating information services was motivated in part by the emergence of the World Wide Web in the early 1990s. At that time, the primary means of access to the Internet was through the telephone network. Companies such as CompuServe, America Online, and EarthLink purchased access to the telephone network from local telephone companies, and they provided dial-up Internet service to their customers through these leased lines. Neither Congress nor the FCC wished to regulate access to the Internet.
The demand for dial-up Internet Service Providers (ISPs) exploded as the Internet grew in popularity. However, by the late 1990s, the arrival of broadband service through coaxial cable, a much faster means of connection and access, threatened to drive the dial-up ISP industry out of business. As the demand for cable service expanded, the question arose as to whether cable companies providing Internet service should be regulated as common carriers like telephone companies.
The FCC answered that question in 2002 by ruling that cable modem service was an information rather than a telecommunication service as defined by the Telecommunications Act of 1996. This meant that cable companies were not required to lease their lines to competing ISPs. As a result, Brand X Internet, a small dial-up ISP that aspired to enhance its offerings by leasing facilities from cable companies, sued the FCC, arguing that cable Internet providers should be classified as telecommunications services under Title II.
The case was ultimately decided by the Supreme Court in National Cable & Telecommunications Association et al. v. Brand X Internet Services et al., 545 U.S. 967 (2005). The Court ruled in favor of the FCC after concluding that the term “telecommunications services” as defined in the Telecommunications Act was ambiguous. Following Chevron, the Court deferred to the FCC’s interpretation and upheld it as reasonable.
Following Brand X, the FCC began to encounter problems of network discrimination by cable modem providers, including efforts by ISPs to block companies from offering telephone service over the Internet, as well as Comcast’s attempt to block subscribers from using BitTorrent. In response, the FCC attempted to implement net neutrality policies through a variety of regulatory tools, including cease and desist orders, policy guidance, and notice-and-comment rulemaking. However, the FCC had tied its hands with its 2002 ruling that cable companies should be unregulated, and thus the agency encountered strong resistance from the federal courts in enforcing any type of net neutrality policy.
Finally, in 2015 under the Obama administration, the FCC reversed its 2002 decision and issued new rules regulating ISPs as common carriers under Title II of the Telecommunications Act. The D.C. Circuit Court of Appeals agreed with the FCC in its June 2016 opinion, holding under Chevron deference that ambiguity of the statutory term “telecommunications services” gave the FCC discretion to regulate ISPs so long as it provided a reasoned explanation for its policy change. The court’s 2-1 majority decided that the FCC’s reclassification was reasonable.
This decision did not settle the matter, however, as following the election of Donald Trump five months later, a politically reconstituted FCC reversed course once again and reclassified ISPs as “information services,” effectively repealing the 2015 net neutrality rules. In August 2018 the Mozilla Corporation filed a legal challenge to the 2018 Order, and in October 2019, the D.C. Circuit of the Court of Appeals ruled in Mozilla Corp. v. FCC, 940 F.3d 1 D.C. Cir. (2019) that the agency could indeed reclassify ISPs as an “information service” because of the authority established in Brand X due to Chevron deference.
Yet in recognizing that the FCC’s reclassification was “permissible” under Chevron Step Two, the circuit court in Mozilla also noted that the permissible standard under Chevron Step Two review is much weaker than the arbitrary and capricious standard required by State Farm. Even though the FCC could reclassify ISPs under Chevron, other aspects of its rule would have to satisfy the higher standard of State Farm. The court then vacated or remanded several other provisions in the FCC’s 2018 rule. The court remanded to the agency portions of the rule involving public safety, pole attachments, and the Lifeline Program and required that these be addressed in a more “reasoned manner.” The court vacated entirely the FCC’s broad preemption of any state and local net neutrality requirements, as it stated that Chevron deference did not afford the FCC the authority to establish such a rule.
The three-judge panel for the Mozilla court consisted of two judges appointed by President Obama and one by President Reagan. Ideologically, then, one would have expected the court to endorse the existing net neutrality rules established under the Obama administration. But given the historical prominence of Chevron on this issue, the constraint of precedent was strong, and the decisions of the Democratic-appointed judges coincided with precedent rather than ideology. The case is therefore a clear illustration of how even when changes in presidential preferences map into substantial regulatory changes, agencies are still meaningfully directed and constrained by the precedents of Chevron and State Farm.
Conclusion
Administrative agencies play a crucial role in American policy making. On everyday issues ranging from fuel economy standards for automobiles, to the prosecution of undocumented immigrants, to Internet access, bureaucratic agencies regularly initiate and implement policy change. Competition for control of the bureaucracy is therefore a regular feature of American national politics. We have argued here that institutional control of the federal bureaucracy has changed significantly over the past thirty-five plus years, largely as the result of two landmark Supreme Court decisions. State Farm prevented agencies from shifting policies simply in response to shifting political winds, and Chevron allowed agencies to act unilaterally more easily and frequently by leveraging their policy expertise.
These two cases reflect a basic dilemma facing the courts. On the one hand is a view that agencies’ policies should reflect changing political preferences of electoral institutions. Agency policy initiatives, after all, are the work of officials nominated by the president and confirmed by the U.S. Senate. On the other is the view that administrative policy should exhibit consistency and coherence across time, and that change should not be whimsical or arbitrary. The economic stakes of regulatory policy are often so substantial that economic planning can become difficult when regulations are whipsawed by electoral volatility.
How the federal courts will resolve this dilemma in coming years is unclear. More apparent is the fact that control over administrative policy change has increasingly moved into the purview of the courts and away from Congress. This is not to say that Congress yields no influence over administrative policy, but when it comes to establishing general guidelines for administrative policy-making authority, bureaucratic supervision is currently better understood in terms of the federal courts than Congress, and this point has often been neglected by scholars in consequential ways.
From a normative perspective, some might find this development troubling. After all, if the locus of bureaucratic control has shifted towards the judicial arena, it follows that major aspects of American public policy are being set by an unelected branch of government. In fact, under Chevron, one unelected branch—the federal judiciary—defers to another unelected branch—the federal bureaucracy. The upside is that the binding nature of judicial precedent effectively guarantees that regulatory policy changes only when agencies can provide a compelling rationale. State Farm takes the politics out of rulemaking change across administrations, and Chevron acknowledges the superior policy expertise of the bureaucracy. Chevron also takes federal judges out of the policy equation, as it should. Letting the bureaucracy call the shots when statutes are ambiguous may be the best option. However, the idea of sending ambiguous statutes back to the institution that failed to be explicit in the first place is troubling also, for it could lead to inaction and failure to resolve economic and societal conflicts.Footnote 17
The extant literature has collectively taught us a great deal about the political dynamics inherent in various delegation relationships and institutional arrangements. But existing theory could be improved substantially by accounting for the role of judicial doctrine in a substantively meaningful way. Such efforts will greatly enhance our understanding of the broader lawmaking system in the United States.
