Emory Law Journal

Agency Legislative History
Jarrod Shobe *Associate Professor of Law, Brigham Young University. For helpful discussions and comments on earlier drafts of this Article, I am grateful to Paul Stancil, Aaron Nielson, Chris Walker, Bijal Shah, Seth Davis, and participants in the Rocky Mountain Jr. Scholars Forum and the BYU law faculty workshop. For excellent research assistance I am grateful to Jessica Payton, Stephanie Nielson, Jeff Cottle, Ryan Ricks, and Jake Barney.

No tool of statutory construction has drawn as much scholarly and judicial attention and controversy as legislative history. This Article shows that the standard account of legislative history often fails to account for legislative history generated through agencyCongress legislative communications, which are often among the most relevant legislative history. These communications, which this Article terms “agency legislative history,” have important implications for theories and the practice of statutory interpretation and agency delegation.

The account of agency legislative history provided here offers a new perspective on the legislative history debate and questions of how empirical realities of the legislative process should influence statutory interpretation. Agency legislative history also sheds new light on the ongoing debate over Chevron’s domain. Agency legislative history reinforces arguments in favor of deference to agencies by raising novel questions about courts’ institutional capacity to effectively uncover congressional deals, and by providing new reasons to believe that agencies may be better statutory interpreters than courts. At the same time, for the many judges skeptical of broad deference but unsure how to limit it, agency legislative history can allow for more narrowly tailored and empirically supported deference decisions that reflect the variety of ways legislation is made.

 

Introduction

Whether, and how, to use legislative history continues to be the most hotly contested issue in statutory interpretation. 1See, e.g., Abbe R. Gluck & Lisa Schultz Bressman, Statutory Interpretation from the Inside—An Empirical Study of Congressional Drafting, Delegation, and the Canons: Part I, 65 Stan. L. Rev. 901, 964 (2013) [hereinafter Gluck & Bressman, Part I] (“The other primary interpretive source that courts consider—and the one whose use is most hotly contested—is legislative history.”); Stephen Breyer, On the Uses of Legislative History in Interpreting Statutes, 65 S. Cal. L. Rev. 845, 845–46 (1992); George A. Costello, Average Voting Members and Other “Benign Fictions”: The Relative Reliability of Committee Reports, Floor Debates, and Other Sources of Legislative History, 1990 Duke L.J. 39, 39–42 (1990); William N. Eskridge, Jr., The New Textualism, 37 UCLA L. Rev. 621, 623 (1990) (“The new textualism posits that once the Court has ascertained a statute’s plain meaning, consideration of legislative history becomes irrelevant.”); William N. Eskridge, Jr., Textualism, the Unknown Ideal?, 96 Mich. L. Rev. 1509, 1512–13 (1998); Abbe R. Gluck, Intersystemic Statutory Interpretation: Methodology as “Law” and the Erie Doctrine, 120 Yale L.J. 1898, 1909–10 n.22 (2011); John F. Manning, Textualism as a Nondelegation Doctrine, 97 Colum. L. Rev. 673, 687–88 (1997); Jane S. Schacter, The Confounding Common Law Originalism in Recent Supreme Court Statutory Interpretation: Implications for the Legislative History Debate and Beyond, 51 Stan. L. Rev. 1, 3, 6 (1998). This controversy is understandable given how important legislative history has been in many of the most noteworthy judicial decisions of the last forty years. 2Manning & Stephenson, infra note 3, at 127; see Gluck, supra note 1. For a few recent Supreme Court examples where legislative history was contested, see Mohamad v. Palestinian Auth., 566 U.S. 449, 459 (2012) (arguing that the statutory text is clear and legislative history therefore unnecessary); id. at 461 (Breyer, J., concurring) (disagreeing that the text is clear and arguing in favor of considering the legislative history); Coleman v. Court of Appeals of Md., 566 U.S. 30, 44–45 (2012) (Scalia, J., concurring in the judgment) (arguing against the plurality’s use of legislative history and in favor of considering the text alone); Reynolds v. United States, 565 U.S. 432, 448 n.* (2012) (Scalia, J., dissenting) (arguing that the majority’s use of legislative history is “superfluous”); Gonzalez v. Thaler, 565 U.S. 134, 165–66 (2012) (Scalia, J., dissenting) (providing a critique of the use of legislative history); DePierre v. United States, 564 U.S. 70, 89 (2011) (Scalia, J., concurring in part and concurring in the judgment) (arguing that the use of legislative history is “not harmless”); Bruesewitz v. Wyeth LLC, 562 U.S. 223, 240 (2011) (acknowledging that the use of legislative history is not considered legitimate by all judges). In this debate, legislative history’s domain is traditionally thought to begin and end with congressional documents, actions, or inactions. 3 These sources of legislative history are covered exhaustively in various leading textbooks on legislation. See, e.g., William N. Eskridge, Jr. et al., Cases and Materials on Legislation and Regulation: Statutes and the Creation of Public Policy (5th ed. 2014); John F. Manning & Matthew C. Stephenson, Legislation and Regulation: Cases and Materials 142 (2d ed. 2013). For interesting cases on the use of legislative inaction as a source of legislative history, see FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120, 146–47 (2000); Bob Jones Univ. v. United States, 461 U.S. 574, 600–01, 607 (1983). Absent from the discussion of legislative history is almost any mention of agency communications with Congress throughout the legislative process, 4 For a comprehensive discussion of the different types of legislative history, see Manning & Stephenson, supra note 3, at 136–45. The current hierarchy of legislative history sources does not always exclude noncongressional sources, but “nonlegislator statements” are rarely referenced and generally thought of as being one of the least authoritative forms of legislative history. William N. Eskridge, Jr., Dynamic Statutory Interpretation 222 (1994). even though agencies are often intimately involved in drafting, revising, and negotiating legislation because of their on-the-ground expertise. 5 Various recent empirical studies have discussed the role of agencies in the legislative process generally. See Lisa Schultz Bressman & Abbe R. Gluck, Statutory Interpretation from the Inside—An Empirical Study of Congressional Drafting, Delegation, and the Canons: Part II, 66 Stan. L. Rev. 725, 758 (2014) [hereinafter Gluck & Bressman, Part II]; Jarrod Shobe, Agencies as Legislators: An Empirical Study of the Role of Agencies in the Legislative Process, 85 Geo. Wash. L. Rev. 451, 454 (2017) [hereinafter Shobe, Agencies as Legislators]; Jarrod Shobe, Intertemporal Statutory Interpretation and the Evolution of Legislative Drafting, 114 Colum. L. Rev. 807, 863–64 (2014) [hereinafter Shobe, Intertemporal Statutory Interpretation]; Ganesh Sitaraman, The Origins of Legislation, 91 Notre Dame L. Rev. 79, 132 (2015); Christopher J. Walker, Legislating in the Shadows, 165 U. Penn. L. Rev. 1377, 1378–79 (2017).

Although scholars and courts traditionally imagine a sharp divide between legislation passed by Congress and implementation by agencies, this Article shows that in reality, there is often a blurred line—with Congress communicating with agencies to understand how they intend to implement statutes and agencies communicating with Congress in a variety of ways and at various stages of the legislative process to influence drafting. 6 This Article focuses on formal agency communications, which generally occur in written form or in formal settings, because these are most likely to be accessible to courts and the public and are more likely to reflect the position of an agency as a whole rather than an individual agency staffer. See infra Section II.A. These communications are often an integral part of forming Congress’s intentions and expectations with respect to legislation. It is not surprising that the specifics of these communications have gone mostly unrecognized and untheorized given how little scholars have studied agencies’ role in the legislative process. 7 Professors Farber & O’Connell made similar arguments for modern regulation generally. See Daniel A. Farber & Anne Joseph O’Connell, The Lost World of Administrative Law, 92 Tex. L. Rev. 1137, 1140 (2014). The lack of understanding about agencyCongress interactions is problematic because it has at times created a mismatch between current statutory interpretation and agency delegation debates, and the realities of the legislative process. This Article aims to remedy this by providing a typology and analysis of these agencyCongress legislative interactions, which this Article collectively terms “agency legislative history.” In doing so, it provides a new perspective on what legislative history is, which is relevant to both textualists who eschew congressional legislative history and purposivists who embrace it.

To understand the implications agency legislative history could have for interpretation, consider the following hypothetical scenarios. Suppose Congress has passed legislation requiring the EPA to implement new environmental restrictions on coal-burning power plants. Throughout the process leading up to enactment, Congress worked closely with the EPA, and the EPA provided hundreds of pages of background material, dozens of hours of testimony, several written letters detailing the Agency’s concerns with the legislation, and a number of written explanations of how the legislation would work on the ground. Much of these materials ended up in congressional committee reports. The owner of a coal-burning power plant has challenged the EPA’s implementation of the legislation. Should the EPA’s voluminous communications with Congress influence a judge’s approach to statutory interpretation and deference?

Now suppose Congress passed a different bill, which was drafted almost entirely by the Department of Defense (DoD). The relevant congressional committee also adopted, as part of its own committee report, language from a section-by-section analysis of the bill provided by the DoD. A number of years later, after the election of a new President, the Department interprets the statute in a way that goes against the description it provided to Congress in the section-by-section analysis, although arguably within the scope of the somewhat vague legislative language. Should a judge interpreting the statute consider the fact that the bill came directly from the DoD when deciding whether to defer to the agency’s interpretation? Should a purposivist judge give more or less weight to the agency’s analysis included in the committee report than other legislative history? And should a textualist judge treat the agency’s analysis included in the committee report like any other legislative history and exclude it from consideration?

Alternatively, suppose an agency sends a letter to Congress opposing a bill and proposing modifications. Congress, despite the agency’s protestations, votes to pass the bill unchanged. A group challenges the agency’s interpretation of the statute. Should a court consider the interactions between the agency and Congress when deciding whether to defer to the agency’s interpretation?

This Article begins to examine how courts could approach questions like these in light of the existence of agency legislative history. One way it does this is by looking at how courts have used agency legislative history in situations similar to those described above. Based on the author’s extensive search of references to agency legislative history in judicial decisions, some courts have used it to help determine whether to uphold agency statutory interpretations, albeit infrequently, inconsistently, and predominantly in the pre-Chevron era. 8See infra Part II. This earlier practice appears to have been mostly lost to modern developments in statutory interpretation and agency delegation, which perhaps helps explain why it has received scant attention from scholars. These pre-Chevron cases provide useful examples of how modern courts could, or perhaps should not, apply agency legislative history in the future. 9See infra Part II.

Agency legislative history is also relevant to ongoing debates about Chevron’s domain. A number of scholars and judges have advocated for a broad application of Chevron that allocates interpretive authority to agencies instead of the judicial branch. 10Vermeule, infra note 12, at 206; William N. Eskridge, Jr., Expanding Chevron’s Domain: A Comparative Institutional Analysis of the Relative Competence of Courts and Agencies to Interpret Statutes, 2013 Wis. L. Rev. 411, 427 (2013); Jerry L. Mashaw, Prodelegation: Why Administrators Should Make Political Decisions, 1 J.L. Econ. & Org. 81, 91–99 (1985); David B. Spence & Frank Cross, A Public Choice Case for the Administrative State, 89 Geo. L.J. 97, 134–41 (2000). These arguments are often based on agencies’ relative expertise and institutional competence. 11Supra note 10. Agency legislative history lends support to these arguments by showing that the creation of legislation is a multi-layered and multi-actor process that often turns on bargains necessary to achieve enactment that may be impossible, or at least incredibly costly, for a court to uncover. It also shows that agencies may be better statutory interpreters than courts because they have rich legislative repositories that record and explain statutory deals and purposes. 12 In this way this Article builds on scholarship by Professors Mashaw and Strauss that argues that agencies should approach interpretation different from courts because of their unique position and relationship with Congress. See, e.g., Adrian Vermeule, Judging Under Uncertainty: An Institutional Theory of Legal Interpretation 205–08 (2006); Jerry L. Mashaw, Norms, Practices, and the Paradox of Deference: A Preliminary Inquiry into Agency Statutory Interpretation, 57 Admin. L. Rev. 501, 504 (2005) (“There are persuasive grounds for believing that legitimate techniques and standards for agency statutory interpretation diverge sharply from the legitimate techniques and standards for judicial statutory interpretation.”); Cass R. Sunstein & Adrian Vermeule, Interpretation and Institutions, 101 Mich. L. Rev. 885, 928 (2003). These records are often inaccessible to courts, and even if they were available they would be difficult and time-consuming for generalist judges, entirely absent from the legislative process, to make sense of. Perhaps, in light of agency legislative history, we should be even more skeptical of courts’ ability to enter into the legislative “black box” 13 Richard H. Pildes, Institutional Formalism and Realism in Constitutional and Public Law, 2013 Sup. Ct. Rev. 1, 2 (2013). and should instead fall back to a more formalist approach to deference that, while “fictional,” provides the best background presumption against which Congress can legislate.

Another concern in light of agency involvement in the legislative process, raised by Professors Walker and Sitaraman, is that Chevron allows agencies to potentially be both drafters and interpreters of legislation, creating a potential for agency self-dealing without judicial oversight. 14 Sitaraman, supra note 5, at 86; Walker, supra note 5, at 1411 (“If the agency is indeed a partner with Congress in the legislative drafting, Justice Scalia’s concern about an agency legislating and executing the law should apply with some force to legislative drafting. The executive and legislative functions are, in essence, combined via agency legislating in the shadows.”). They analogize this to Auer deference, whereby courts defer to agency interpretations of their own regulations, and argue that because of this concern perhaps courts should move away from Chevron toward the less deferential Skidmore standard. 15 Auer v. Robbins, 519 U.S. 452, 461–63 (1997). But see Talk Am., Inc. v. Mich. Bell Tel. Co., 564 U.S. 50, 69 (2011) (Scalia, J., concurring) (criticizing Auer because “deferring to an agency’s interpretation of its own rule encourages the agency to enact vague rules which give it the power, in future adjudications, to do what it pleases”). This analogy certainly sounds alarming, however no research to date, including this Article’s findings, has shown that Congress has ceded all or even a significant portion of its legislative authority to agencies. And courts have not appeared concerned about this risk. In fact, this Article shows that courts appear to take agency involvement as a signal from Congress of an intent to defer to the agency rather than as raising an issue of potential self-dealing.

While a blanket presumption against delegation would be an overly broad response to agency involvement in the legislative process, judges concerned more generally about deference to agencies could use agency legislative history to defer in more contextual ways. 16 A number of Supreme Court Justices have recently expressed concern with the expansion of the administrative state. City of Arlington v. FCC, 569 U.S. 290, 315 (2013) (Roberts, C. J., dissenting) (“[T]he danger posed by the growing power of the administrative state cannot be dismissed.”); see also Dep’t of Transp. v. Ass’n of Am. R.Rs., 135 S. Ct. 1225, 1254 (2015) (Thomas, J., concurring in the judgment) (describing the administrative state as a system that “concentrates the power to make laws and the power to enforce them in the hands of a vast and unaccountable administrative apparatus”). For an excellent discussion of the rise of the administrative state, see Gary Lawson, The Rise and Rise of the Administrative State, 107 Harv. L. Rev. 1231, 1233 (1994). Despite the fact that scholars and judges have written volumes about when and how Chevron should apply, they have been unable to articulate a coherent and predictable set of rules to determine when to defer. If judges wish to approach Chevron in a more contextual way, this Article provides new avenues that would allow them to do so in ways that better reflect the realities of agency–Congress relationships and Congress’s own chosen legislative process. Agency legislative history may often provide the best evidence of whether Congress intended to delegate with respect to a particular ambiguity. Indeed, as discussed above, pre-Chevron courts often looked to agency legislative history as a means of discerning whether to defer to an agency interpretation, so in many ways this would simply be a return to prior judicial practice. This Article provides a path forward for judges looking to trade universalist theories of agency delegation for modes of deference based on the variety of ways legislation is made. 17 This is something the Supreme Court has explicitly pursued. See United States v. Mead Corp., 533 U.S. 218, 236 (2001) (“Justice Scalia’s first priority over the years has been to limit and simplify. The Court’s choice has been to tailor deference to variety.”).

Agency legislative history also has implications for the substantive interpretation of statutes. If Congress uses its legislative history to memorialize an understanding reached with an agency and to hold the agency to the bargain it made during the legislative process, it would make sense for a court to give that a privileged place in its statutory analysis. It may be that recent skepticism towards legislative history has caused courts to deemphasize these types of communications. But courts that ignore legislative history need to be aware that they might also be ignoring important evidence of agency–Congress legislative deals. It would seem that even a textualist judge would view agency legislative history as distinct from other legislative history. Textualists are concerned about individual members or staffers using Congress’s legislative history to do an “end run” around the constitutionally prescribed process, 18 Manning, infra note 34, at 1534. but by ignoring agency legislative history and deferring to agency interpretations, courts are potentially allowing agencies to do an “end run” around Congress. So even if textualists choose to ignore other legislative history, they should still consider agency legislative history, especially when the alternative is simply to defer to an agency.

Agency legislative history has implications for current debates surrounding how empirical realities of the legislative process should influence statutory interpretation. 19See infra Section III.B. These debates have focused on the role of congressional staff as the creators of legislative policy and legislative history, and legislative counsel as technical statutory drafters. 20See infra Section III.B. Agency legislative history indicates that these accounts are oversimplified. These debates have not accounted for the variety of roles agencies play throughout the legislative process. For example, they have not accounted for the role of agencies as statutory drafters and revisers, nor have they accounted for the fact that congressional legislative history is sometimes a reflection of statements made by an agency to Congress rather than an internally generated congressional understanding. Agency legislative history shows that the process of generating both statutory text and legislative history is even more variable and messy than is commonly thought, which calls into question existing arguments about how empirical realities of the legislative process should influence interpretation and delegation.

This Article proceeds in three Parts. Part I provides background on the use of legislative history generally and recent empirical scholarship on the legislative process. Part II describes the various types of agency legislative history, provides examples of each type, and explains how courts have used it, predominantly in the pre-Chevron era. Part III discusses the implications of agency legislative history for theories of agency deference and statutory interpretation.

I. Legislative History’s Domain

Scholars and judges generally think of legislative history as the entire set of circumstances surrounding the conception, deliberation, drafting, and amending of a piece of legislation. 21William N. Eskridge, Jr., Interpreting Law: A Primer on How to Read Statutes and the Constitution 191 (2016) [hereinafter Eskridge, Jr., Interpreting Law]. The standard textbook account of legislative history includes a wide variety of sources, including the general circumstances surrounding the introduction and consideration of legislation, committee reports, conference reports, statements by sponsors or drafters of legislation, the record of changes to the legislation over the course of the drafting process, hearings, floor debates, post-enactment legislative history, and legislative inaction. 22 These sources of legislative history are covered exhaustively in various leading textbooks on legislation. See Eskridge, Jr., supra note 4; supra note 3 and accompanying text. This broad account of legislative history generally focuses on those within Congress who are thought to be most closely involved in creating legislation, including congressional committees and those individual members of Congress who have worked with those committees to shepherd the legislation through the legislative process. 23Eskridge, Jr. et al., supra note 3, at 632. Rarely in discussions of legislative history is any serious consideration given to agency communications with Congress throughout the legislative process. 24 “The court does not rely on statements by nonlegislative officials as the most probative evidence of the meaning of statutory language. There are very few state cases and those that exist reject the use of such evidence of statutory meaning.” Id. at 828; e.g., Hayes v. Cont’l Ins. Co., 872 P.2d 668, 673–74 (Ariz. 1994). This is true even though agencies are often intimately involved in drafting, revising, and negotiating legislation, and the legislative process often turns on agreements between an agency and Congress over what certain provisions mean and how they will be implemented. 25See supra note 24. This Article focuses on this overlooked type of legislative history, which it terms “agency

legislative history,” and argues that it may be, in many instances, the most illuminating and accurate legislative history. This Article contributes a novel typology of agency legislative history that sets a base for discussions of how courts and scholars should account for agency legislative history as a unique form of legislative history.

The fact that agency legislative history is so rarely discussed is unsurprising, because, although observers have long suspected that agencies play some role in creating legislation, 26 Peter L. Strauss, Essay, “Deference” Is Too Confusing—Let’s Call Them “Chevron Space” and “Skidmore Weight”, 112 Colum. L. Rev. 1143, 1146 (2012) [hereinafter Strauss, Chevron Space” and “Skidmore Weight”] (“The agency may have helped to draft the statutory language, and was likely present and attentive throughout its legislative consideration. Its views about statutory meaning may have been shaped in the immediate wake of enactment, under the enacting Congress’s watchful eye.”); see Matthew R. Christiansen & William N. Eskridge, Jr., Congressional Overrides of Supreme Court Statutory Interpretation Decisions, 1967–2011, 92 Tex. L. Rev. 1317, 1323, 1377 (2014) (providing evidence that agencies work with congressional staff to draft override legislation, and lobby for such legislation, based on an examination of committee hearings and reports). what that role is and how it works has only recently come into focus. Recent articles by Professor Walker 27 Walker, supra note 5. Professor Walker’s article was part of a larger project commissioned by the Administrative Conference of the United States. Christopher J. Walker, Federal Agencies in the Legislative Process: Technical Assistance in Statutory Drafting (2015). and Professor Sitaraman, 28 Sitaraman, supra note 5, at 124–32. as well as the author’s own research, 29See Shobe, Agencies as Legislators, supra note 5. have provided a window into the ways agencies are involved in creating the legislation they are charged with implementing. Other recent empirical studies of the legislative process have hinted more generally at agency involvement in the legislative process. 30See Brigham Daniels, Agency as Principal, 48 Ga. L. Rev. 335, 404 (2014) (“Sometimes Congress asks agencies to draft language, and sometimes agencies do so without being asked. It is just the way the game is played, and those with much experience in Washington openly acknowledge this.”); Gluck & Bressman, Part I, supra 1, at 1021 (“There are likely external networks of these noncongressional drafters of federal legislation, with deep resources of institutional and legal knowledge, that may influence statutory drafting in ways that have been underappreciated and merit their own separate study.”); Gluck & Bressman, Part II, supra note 5, at 758 (“[O]ur respondents told us that first drafts are typically written by, respectively, the White House and agencies, or policy experts and outside groups, like lobbyists. Empirical work is lacking for the details of this account . . . .”); Nicholas R. Parrillo, Leviathan and Interpretive Revolution: The Administrative State, the Judiciary, and the Rise of Legislative History, 1890–1950, 123 Yale L.J. 266, 338–42 (2013) (documenting the key role of agencies in the legislative process during the New Deal era as part of a larger study of the rise of the use of legislative history); Christopher J. Walker, Inside Agency Statutory Interpretation, 67 Stan. L. Rev. 999, 1004 (2015) [hereinafter Walker, Inside Agency]. And some prominent scholars have implied that courts (and scholars) should consider agency–Congress legislative communications, but only in a very general and speculative manner. 31Eskridge, Jr. et al., supra note 3, at 829 (“Indeed, since proposed legislation is frequently drafted by the executive department or by private interest groups, their statements and explanations at hearings might be the only truly informed explanation of the structure and operation of the statute.”); Eskridge, Jr., Interpreting Law, supra note 21, at 256. Even the few scholars who have directly addressed the issue of agency involvement in the legislative process have focused on what this involvement means generally, without fully examining the many ways in which agencies communicate legislatively with Congress and how and why this should matter for statutory interpretation. 32See Shobe, Agencies as Legistalators, supra note 5; Shobe, Intertemporal Statutory Interpretation, supra note 5; Walker, Inside Agency, supra note 30; Walker, Legislating in the Shadows, supra note 5. The next Part begins to unpack these agency–Congress legislative interactions by considering the various types of legislative communications between agencies and Congress.

II. Types of Agency Legislative History

Much like congressional legislative history, agency legislative history comes in many different forms. This Part discusses various types of agency legislative history and gives real-world examples of them. The goal of this Part is to show how common agency legislative history is, and that it comes in a variety of forms. This Part also documents, based on an extensive search of references to agency legislative history in judicial decisions, how the various types of agency legislative history have influenced statutory interpretation. While scholars have rarely discussed the existence of agency legislative history, courts have, albeit infrequently and quietly, used agency legislative history as part of interpretation for many decades. 33See cases cited infra note 54. This Part discusses a number of cases, covering each only briefly, with a focus on how agency legislative history influenced the decision. These cases mostly come from the pre-Chevron era when courts used a case-by-case approach to determine whether to defer to agency interpretations instead of Chevron’s broad presumption of deference. Importantly, in this era courts also relied more heavily on legislative history generally, and the recent deemphasis of legislative history may have also reduced judicial reliance on agency legislative history. 34See John F. Manning, Chevron and Legislative History, 82 Geo. Wash. L. Rev. 1517, 1525 n.46 (2014). This discussion provides background for discussions in Part III of what agency legislative history could mean going forward for theories and the practice of statutory interpretation.

A. Formal vs. Informal Agency Legislative History

This section distinguishes two broad types of agency–Congress communications that are part of the history of a piece of legislation. First are informal agency–Congress legislative communications, like phone calls and emails between staffers. Second are formal written or spoken communications with Congress, which are generally documented and often publicly available. This section briefly discusses informal agency–Congress communications, while the rest of this Part focuses primarily on the many types of formal agency–Congress communications.

Agency staff frequently engage congressional staff on an informal basis, including by email, by phone, and in person. 35See Shobe, Agencies as Legislatures, supra note 5, at 489. This type of interaction is perhaps the most frequent contact between agencies and Congress and occurs at the staff level with no oversight from the Office of Management and Budget (OMB). 36Id. These interactions could often provide helpful illumination into the purpose and meaning of statutory language. In a perfect world, judges would have access to, and be able to make sense of, all relevant information about the legislative process, including informal and formal communications between agencies and Congress. Of course, it is difficult or impossible to gain access to informal communications between agencies and Congress, and even if judges could, it would be difficult for them to know how to account for different types of informal communications. The same is true of internal congressional deliberations. For example, although conversations among congressional staff or between congressional staff and Congress’s professional drafters in the offices of legislative counsel might be the most valuable legislative history, no one has seriously argued that every internal congressional email and conversation be made available to a judge interpreting a statute.

Although a good deal of the legislative communications between agencies and Congress are informal, agencies often engage Congress in various formal ways. 37See infra Sections II.B–II.D. These formal communications generally must be cleared through OMB and are then submitted to Congress in written form. 38The Mission and Structure of the Office of Management and Budget, Office of Mgmt. & Budget, https://obamawhitehouse.archives.gov/omb/organization_mission/ (last visited Oct. 20, 2018). These documents are often publicly available through agency websites or can be made available through FOIA requests. A number of websites request agency documents through FOIA and post them online. For example, the website governmentattic.org has posted certain Department of Justice views letters acquired through FOIA requests on their website. See, e.g., Copies of Certain Department of Justice (DOJ) Views Letters from the 107th and the 108th Congresses, 2001-2005, Governmentattic.org (Feb. 1, 2016) [hereinafter Copies of Certain Views Letters, 2001-2005], http://www.governmentattic.org/19docs/DOJviewsLetters_2001-2005.pdf. For this reason, the focus of this Article is on formal agency legislative history that is more likely to be accessible to courts and the public, more salient to all of Congress (or at least congressional committees), and more likely to reflect the position of an agency as a whole rather than an individual agency staffer. The remainder of this Part describes the most common forms of formal agency legislative history and provides real-world examples of them.

B. Agency-Proposed Legislation

The most prominent kind of formal agency legislative history is agency-drafted legislative proposals. Agencies often draft their own statutory language, reflecting their policy preferences, which they then submit to Congress in hopes that Congress will use it as a starting (or even ending) point to the legislative process. 39 Shobe, Agencies as Legislators, supra note 5, at 468 (“[A]gencies commonly originate their own legislative proposals and also draft legislation at Congress’s request”). That the Executive Branch would propose legislation is anticipated by the Constitution. The Recommendation Clause, found in Article II, Section 3, states that “[The President] shall from time to time . . . recommend to [Congress’s] Consideration such Measures as he shall judge necessary and expedient . . . .” U.S. Const. art. II, § 3. This provision underwent multiple drafts and the changes provide helpful context of what the Framers expected out of the President from this Clause. In an earlier draft the Clause allowed the President to recommend legislation but did not require him to do so. See J. Gregory Sidak, The Recommendation Clause, 77 Geo. L.J. 2079, 2081 (1989) (“James Madison’s notes on the Constitutional Convention for August 24, 1787, reveal that the Framers explicitly elevated the President’s recommendation of measures from a political prerogative to a constitutional duty . . . .”). An earlier version also contained the word “matters” rather than “measures.” Id. at 2084. This change “reinforces the inference that the Framers intended the President’s recommendations to be more than precatory statements . . . .” Id. This indicates that the Framers intended the President to make specific legislative proposals in the form of bill language. See Vasan Kesavan & J. Gregory Sidak, The Legislator-in-Chief, 44 Wm. & Mary L. Rev. 1, 48–49 (2002); Shobe, Agencies as Legisltaors, supra note 5, at 469 (“It is no secret that the President proposes bills to Congress. But where does that language come from? It doesn’t appear by magic. Someone in an agency is the one who wrote it.”). Agency proposals are sometimes made public by the agency as part of the President’s budget or as part of the agency’s process of promoting the proposal. Some agencies post their legislative proposals on their website. For example, the Department of Transportation recently drafted an expansive proposal to fund improvements to transportation infrastructure. Although the Department of Transportation is the lead agency, the 350-page draft bill includes roles for various agencies, including the EPA, Department of Interior, and Department of Labor. See, e.g., GROW AMERICA Act, H.R. 2410, 114th Cong. (2015), https://www.transportation.gov/sites/dot.gov/files/docs/GROW_AMERICA_Act_1.pdf (describing roles for various departments). This form of agency legislative history generally shows a high level of agency sophistication and involvement in the legislative process. Agency drafting can happen for a one-time bill that an agency wants enacted or as part of an ongoing agency–Congress relationship. For example, Congress considers certain bills, like the National Defense Authorization Act and the Farm Bill, on a regular basis, and the agencies affected by those bills frequently draft proposals they hope Congress will include in those bills. 40 Congress has enacted a version of the National Defense Authorization Act every year since at least 1961. See House Armed Servs. Comm., History of the NDAA, https://armedservices.house.gov/ndaa/history-ndaa (last visited Oct. 20, 2018) (providing links to each National Defense Authorization Act since 1961). The DoD has an Office of Legislative Counsel whose job is to oversee the agency’s legislative proposals for the National Defense Authorization Act. See Dep’t of Def., Office of Legislative Counsel, An Overview of DoD’s Legislation Program 4 (2006). Other times an agency sees a particular need that is not being properly addressed by existing law, either because of a change in circumstances, a judicial opinion, or a poorly drafted or underspecified statute, so it will propose a legislative change to Congress. Either way, the fact that an agency was the primary drafter of legislation is a relevant part of the legislative history of a bill from which courts have drawn inferences about congressional intent to delegate, as discussed below. 41See Section III.A.

Courts have occasionally considered the fact that an agency proposed or opposed the language at issue when determining whether and to what degree they should defer to agency interpretations of statutes. 42See cases cited infra note 47. These courts have essentially created a canon of interpretation that affords greater deference to an agency where the agency drafted the relevant language. As the Court said in United States v. American Trucking Associations, “the Commission’s interpretation gains much persuasiveness from the fact that it was the Commission which suggested the provisions’ enactment to Congress.” 43 310 U.S. 534, 549 (1940). In a rare post-Chevron case invoking agency legislative history—Zuni Public School District No. 89 v. Department of Education—Justice Breyer, writing for the majority, took the unusual step of considering the fact that the Department of Education had originally drafted the language at issue before he looked to the language of the statute. 44 550 U.S. 81, 98 (2007). Based partially on the Department of Education’s involvement in the legislative process, which was noted in the Congressional Record, 45 139 Cong. Rec. 2,599 (1993) (House sponsor of the bill referring to the bill as “the administration’s proposal.”); Id. at 23,416 (Senate sponsor of the bill stating the bill was introduced by the agency. E.g., “I am pleased to introduce on behalf of the administration . . . .”; “The administration is proposing”; and “The administration’s proposal calls for . . . .”). the Court determined that Congress must have intended to defer to agency interpretations of the statute. 46Zuni Pub. Sch. Dist. No. 89, 550 U.S. at 98. Several Justices went out of their way to disavow Justice Breyer’s approach and to encourage the use of the traditional Chevron method. Id. at 107 (Kennedy, J., concurring). Judges have relied on similar reasoning in a number of earlier Supreme Court and lower court opinions. 47 Id. at 90 (examining the legislative history and finding that the “present statutory language” had come from draft legislation that the Secretary of the Department of Education had submitted to Congress in 1994); Smith v. City of Jackson, 544 U.S. 228, 239 (2005) (“[T]he Department of Labor . . . initially drafted the legislation [being examined in the case.]”); Howe v. Smith, 452 U.S. 473, 485 (1981) (arguing that “the Bureau’s interpretation of the statute merits greater than normal weight because it was the Bureau that drafted the legislation and steered it through Congress . . . .”); United States v. One Bell Jet Ranger II Helicopter, 943 F.2d 1121, 1126 (9th Cir. 1991) (“The interpretation of the agency charged with administering the statute is entitled. . . to ‘greater than normal weight’ when that agency drafts the legislation and steers it through Congress with little debate.”); Watkins v. Blinzinger, 789 F.2d 474, 478 (7th Cir. 1986) (arguing that when an agency drafted language courts should defer unless the interpretation is “beyond the pale of reasonableness”); W. Nuclear, Inc. v. Andrus, 664 F.2d 234, 240 (10th Cir. 1981); Beshaw v. Fenton, 635 F.2d 239, 245 (3d Cir. 1980) (stating that the agency interpretation was given more weight because it drafted the section at issue and got it through Congress); Saldivar v. Rodela, 894 F. Supp. 2d 916, 930 (W.D. Tex. 2012); Creaton v. Bowen, No. CV 85-3306-R, 1986 U.S. Dist. LEXIS 25485, at *12 (C.D. Cal. 1986).

The Supreme Court has also considered Congress’s rejection of an agency legislative proposal as a useful way of inferring congressional intent. In Nacirema Operating Co. v. Johnson48 396 U.S. 212, 217–18 (1969). the Court considered that Congress had passed the legislation at issue despite Department of Labor opposition. The Department of Labor wanted broader legislation and proposed language to that effect, but Congress rejected the Department’s proposal and instead adopted a bill with narrow language. 49Id. (“In fact, a representative of the Labor Department objected to the bill precisely for that reason, urging the Committee to extend coverage to embrace the contract, ‘and not the man simply when he is on the ship.’ If Congress had intended to adopt that suggestion, it could not have chosen a more inappropriate way of expressing its intent than by substituting the words ‘upon the navigable waters’ for the words ‘within the admiralty jurisdiction.’”). The Court used Congress’s rejection of the Department of Labor’s position to argue that the ambiguous statutory language should be construed in a narrow manner that went against the agency’s preferred interpretation. 50Id. This shows that courts could benefit from looking at both positive and negative interactions between agencies and Congress when determining whether to defer to an agency interpretation.

C. Agency Involvement in the Drafting Process

Another type of agency legislative history is agency involvement in reviewing and drafting legislation for which it was not the primary drafter. Because of their comparative expertise, Congress often allows agencies to be closely involved in reviewing and commenting on proposed legislation drafted within Congress or by other outside parties. 51 Shobe, Agencies as Legislators, supra note 5, at 468 (“[A]gencies provide extensive review of, and revisions to, statutory language drafted by outside agencies.”). Of course, it is not always possible for a court to determine whether an agency was involved in drafting. But, often it is. Although it is more difficult to tell when an agency proposed edits to a bill, it is not uncommon for agencies to submit edits as part of a views letter or through other public means. There are also strong indicators of when an agency did not draft language. For example, if an agency or administration issues a views letter or testifies in opposition of a bill, or certain sections of a bill, then it would seem likely that the agency did not draft, and was not heavily involved in reviewing, the bill or those sections of the bill. This can happen in a variety of ways, either at Congress’s request or through an agency’s own monitoring of legislation. 52Id. As the author has described elsewhere, this involvement can be substantial or minor depending on the issue and the relationship between the agency, Congress, and Congress’s own internal expertise. 53Id. at 482. The level and type of involvement can serve as a useful signal of the agency–Congress relationship and is another relevant part of the legislative history of an enacted statute.

Courts have often considered agency participation in creating legislation as a factor in their decisions. 54See, e.g., Aluminum Co. of Am. v. Cent. Lincoln Peoples’ Util. Dist., 467 U.S. 380, 390 (1984) (noting that the agency “was intimately involved in the drafting and consideration of the [Regional Act] by Congress”); Guardians Ass’n v. Civil Serv. Comm’n of the City of New York, 463 U.S. 582, 592 (1983); Pub. Util. Dist. No. 1 of Douglas County v. Bonneville Power Admin., 947 F.2d 386, 390 (9th Cir. 1991) (arguing that because the agency was involved in drafting the court should defer unless the interpretation is unreasonable); Aluminum Co. of Am. v. Bonneville Power Admin. 903 F.2d 585, 590 (9th Cir. 1990) (same); Cal. Energy Res. Conservation & Dev. Co. v. Johnson, 807 F.2d 1456, 1459 (5th Cir. 1987) (same); Almendarez v. Barrett-Fisher Co., 762 F.2d 1275, 1282 n.3 (5th Cir. 1985); Sweeny v. Murray, 732 F.2d 1022, 1029 (1st Cir. 1984) (arguing that greater deference should be given when agency played a role in drafting statute); McDannell v. U.S. Office of Pers. Mgmt., 716 F.2d 1063, 1066 (5th Cir. 1983) (giving greater deference to the Office of Personnel Management (defendant-appellant) in their interpretation of the statute because their predecessor agency had “an actual hand in its drafting and passage”); Int’l Nutrition, Inc. v. U.S. Dep’t of Health & Human Servs., 676 F.2d 338, 342 (8th Cir. 1982) (arguing that because the FDA participated in drafting the statute, the court defers to its interpretation, as long as that interpretation furthers goals of legislation); Comm. for Auto Responsibility v. Solomon, 603 F.2d 992, 1004 (D.C. Cir. 1979) (noting that the General Services Administration had been actively involved in drafting and adopting the statutory language that was in dispute in the case); Hercules, Inc. v. EPA, 598 F.2d 91, 125 (D.C. Cir. 1978); Morgan v. United States, No. CV 84-4664 (RR), 1991 WL 353371, at *4 (E.D.N.Y. June 30, 1991) (“It is axiomatic that where an agency assists in the drafting of legislation and aids in its passage, it views on that legislation are entitled to great deference.”); Air Courier Conference of Am./Int’l Comm. v. U.S. Postal Serv., 762 F. Supp. 86, 88 (D. Del. 1991) (noting that “due deference” was supported by the fact that Postal Department officials had participated in the Act’s drafting); Faught v. Heckler, 577 F. Supp. 1180, 1186 (S.D. Iowa 1983); Turney v. United States, 525 F. Supp. 675, 677 (D. Md. 1981) (“Administrative interpretations are especially persuasive where, as here, the agency participated in developing the provision.”); Am. Waterways Operators, Inc. v. United States, 386 F. Supp. 799, 804 (D.D.C. 1974) (“Of higher significance, however, is the construction placed on an act by those administrators who participated in its drafting and directly made known their views to Congress.”). The Supreme Court has generally treated agency participation in the legislative process the same as when the agency was the original drafter, as discussed in the previous section, by creating a canon whereby the Court granted greater deference when an agency was involved in creating the statute at issue. 55See supra Section II.B. To quote the Court in Miller v. Youakim, “[a]dministrative interpretations are especially persuasive where, as here, the agency participated in developing the provision.” 56 440 U.S. 125, 144 (1979). An earlier case, Adams v. United States, invoked a similar idea: “These agencies cooperated in developing the Act, and their views are entitled to great weight in its interpretation.” 57 319 U.S. 312, 315 (1943). The Court never made it clear why agency involvement alone was enough to create a presumption of deference, but implicit in these cases seems to be a desire not to interfere with the work of agencies that were much closer to the legislative process and therefore better able to interpret the statute. 58See cases cited supra note 54. Courts that have applied this canon of greater deference have not automatically accepted agency interpretations. Aluminum Co. of Am., 903 F.2d at 594, 598–99 (noting that the commission had a significant role in drafting the legislation); Watkins v. Blinzinger, 789 F.2d 474, 478 (7th Cir. 1986) (stating that the court should defer to agency interpretations where the agency drafted the language, unless the interpretation is “beyond the pale of reasonableness”); Cent. Lincoln Peoples’ Util. Dist. v. Johnson, 735 F.2d 1101, 1106 (9th Cir. 1984) (noting that the commission had a significant role in drafting the legislation); Watts v. Hadden, 686 F.2d 841, 843 (10th Cir. 1981) (same); Patagonia Corp. v. Bd. of Governors, 517 F.2d 803, 811, 813 (9th Cir. 1975) (noting that the agency suggested the language at issue but still relying on an analysis of the meaning of the words to rule against the agency). Much like in Chevron, these courts have still looked to whether the interpretation was “permissible” or “reasonable.” Chevron U.S.A. Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 840, 866 (1984). The Supreme Court in Zuber v. Allen implied that judges should look beyond the mere fact of agency involvement in drafting to other types of agency legislative history to make sure that no other evidence pointed to a disagreement between Congress and the responsible agency. 396 U.S. 168, 181–82, 184, 188 (1969).

Courts have placed caveats on this general presumption of greater deference to agency interpretations when the agency participated in the legislative process. For example, in Barnett v. Weinberger, the D.C. Circuit declined to defer to an agency interpretation of a statute that the agency helped draft because the agency’s interpretation came many years after the statute was enacted. 59 818 F.2d 953, 960–61 (D.C. Cir. 1987) (“It is well established that the prestige of a statutory construction by an agency depends crucially upon whether it was promulgated contemporaneously with enactment of the statute.”). Conversely, in Peters v. City of Shreveport, the Fifth Circuit upheld an agency interpretation made soon after enactment of the legislation, noting that when “an administrative interpretation . . . is made contemporaneously with the enactment of the statute, courts give the construction more deference” because the agency is “in a position to accurately interpret the [statute] in accordance with Congress’s intentions.” 60 818 F.2d 1148, 1155–56 n.4 (5th Cir. 1987); see also Aluminum Co. of Am. v. Cent. Lincoln Peoples’ Util. Dist., 467 U.S. 380, 390 (1984) (noting that the agency was involved in drafting the statute and its interpretation came immediately following the enactment of the statute); Howe v. Smith, 452 U.S. 473, 474 (1981) (noting that the agency’s interpretation was “contemporaneous”). These caveats to the general rule of deferring when an agency is involved in the legislative process make sense because when an agency interprets a statute many years after enactment, it is unlikely that interpretation is based on the agency’s proximity to the legislative process and Congress’s intent.

D. Agency Legislative Analysis

Agencies communicate with Congress during the legislative process in ways beyond drafting and revising legislation, yet these types of legislative communications have gone almost entirely unnoticed in legal literature. Agencies engage in various types of formal communications with Congress to express opinions about legislation, to raise issues with legislation, and to suggest changes to legislation. This section discusses various types of legislative analysis that agencies provide to Congress. The types of agency legislative history discussed here are not mutually exclusive, and often various types will exist for any bill.

1. Section-by-Section Analyses

When agencies propose their own legislation to Congress, they often include a section-by-section analysis that explains the legislation in relatively plain-language terms and provides color and context to the statutory language. 61 For examples of agency section-by-section analyses, see OLC Detailed Guidelines for Preparation of Legislative Proposals for the Fiscal Year 2016 DoD Legislative Program, Dep’t of Def. (2014), http://ogc.osd.mil/olc/docs/DetailedGuidelinesforPreparingProposalsFY16.pdf (discussing section-by-section analysis requirement for DoD proposals); Surface Transportation Reauthorization Bill, U.S. Dep’t of Transp. (Feb. 11, 2015), https://www.transportation.gov/content/surface-transportation-reauthorization-bill-laying-foundation-us-economic-growth-and-job (providing proposed legislative language and accompanying section-by-section analysis to Congress). This is similar to congressional committee reports, which normally contain a similar section-by-section analysis of legislation. 62Eskridge, Jr., Interpreting Law, supra note 21, at 242. An agency-drafted section-by-section analysis can be important to understanding how Congress perceived the agency’s proposed legislative language. It is now known that many members of Congress and their staff are more likely to read committee reports than legislative language. 63 Archer-Daniels-Midland Co. v. United States, 37 F.3d 321, 323–24 (7th Cir. 1994) (Posner, C.J.). That members of Congress are more likely to read a committee report than actual statutory language is confirmed by some of those involved in the legislative process. See Gluck & Bressman, Part I, supra note 1, at 968–69. Judge Katzmann recently explained, in light of “the expanding, competing demands on legislators’ time,” “they cannot read every word of the bills they vote upon, but they, and certainly their staffs, become educated about the bill by reading the materials produced by the committees and conference committees from which the proposed legislation emanates.” Robert A. Katzmann, Statutes, 87 N.Y.U. L. Rev. 637, 653 (2012). It may also be true that these same legislators and staff read an agency’s plain-language section-by-section analysis more closely than an agency’s proposed statutory text, which can be difficult to decipher because it often amends various portions of existing law and cross-references other statutory provisions. These section-by-section analyses therefore may be the best evidence of what Congress believes it is enacting when it adopts agency-drafted legislation. 64See Lisa Schultz Bressman, Chevron’s Mistake, 58 Duke L.J. 549, 582 (2009) (“[A]n agency may be involved in drafting the legislation, sharing its interpretations with legislative staff. If those understandings are sufficiently specific, Congress may rely on them when enacting legislation.”).

Various agencies make their legislative proposals and accompanying section-by-section analyses publicly available. For example, the DoD’s Office of Legislative Counsel posts its legislative proposals and section-by-section analyses on its website. 65See, e.g., DoD Legislative Proposals, Fiscal Year 2017, Dep’t of Def., http://ogc.osd.mil/olc/legispro17.html (last visited Oct. 20, 2018). The DoD has a sophisticated and coordinated legislative drafting process, due partially to the fact that Congress passes a yearly defense reauthorization bill that requires significant input from the DoD. 66See, e.g., National Defense Authorization Act for Fiscal Year 2017, H.R. 4909, 114th Cong. §§ 113, 143 (2d Sess. 2016). The Farm Bill is an example of another must-pass bill that receives significant agency input. See, e.g., Agriculture, Rural Development, Food and Drug Administration, and Related Agencies Appropriations Act, H.R. 5054, 114th Cong. (2d Sess. 2016). The DoD proposes hundreds of pages of legislative language every year, and all of this legislative language is accompanied by a relatively plain-language section-by-section analysis of the purpose and function of the bill. 67DoD Legislative Proposals, Fiscal Year 2018, Dep’t of Def., http://ogc.osd.mil/olc/legispro18.html (last visited Oct. 20, 2018). It is likely that the DoD creates these section-by-section analyses for a reason: it knows that committee staff and members of Congress want a clear explanation of what the proposals do, which is hard to provide through relatively technical and dense legislative text.

2. Views Letters

Views letters are an additional type of agency legislative analysis. Views letters are formal letters sent to Congress that state an agency’s position on proposed legislation. These letters generally include a description of why the agency supports or opposes the legislation and what the agency believes the legislation will do. 68See, e.g., Morgan v. United States, No. CV 84-4664 (RR), 1991 WL 353371, at *4 (E.D.N.Y. June 30, 1991) (citing to a views letter from the Veterans Administration showing the agency’s position and interpretation of what the amendment would do); see also Dep’t of Defense, Office of Legislative Counsel, supra note 40, at 4 (“Periodically, the Chairman of the House or Senate Armed Services Committee requests that DoD review a bill and provide an official position, or ‘views’ of the Department via a formal letter on the merits of the subject bill. Such a letter declaring DoD’s official position, or ‘views,’ is commonly referred to as a ‘Views Letter.’ OLC receives the requests for Departmental views letters and coordinates them with all concerned DoD components.”); Nat’l Oceanic & Atmospheric Admin., Congressional Testimony and Views Letters, Off. of Legis. & Intergovernmental Aff., http://www.legislative.noaa.gov/115testimony.html (last visited Oct. 20, 2018) (“Views letters are statements of the Agency’s, Department’s, and Administration’s position, thoughts, and comments, on specific issues or legislation being considered by Congress.”); Views Letters, U.S. Dep’t of Just., Off. of Legis. Aff., https://www.justice.gov/ola/views-letters (last visited Oct. 20, 2018); Letters to Congress, Dep’t of Com., Off. of the Gen. Couns., https://ogc.commerce.gov/letters (last visited Oct. 20, 2018); Copies of Certain Department of Justice (DOJ) Views Letters from the 109th and 110th Congresses, 2005–2007, Governmentattic.org (Nov. 14, 2014), http://www.governmentattic.org/13docs/DOJviewsLetters_2005-2007.pdf (compiling Department of Justice views letters through FOIA requests). Sometimes an agency will also include technical comments as part of a views letter. 69See Negonsott v. Samuels, 507 U.S. 99, 106 (1993) (noting that the Department of Interior proposed amendments to the original bill in a letter to Congress). For major pieces of legislation, an agency may send multiple views letters throughout the legislative process as the legislation evolves. Views letters sometimes provide direct insight into the meaning of statutory language. 70See supra note 68. Even when they do not provide insight into specific language, they are helpful to establish what an agency has told Congress it believes the purpose and scope of the legislation is and to contextualize the relationship between the agency and Congress. 71See supra note 68.

3. Pre-Drafting Reports and Memos

Agencies also often send reports, letters, and memos to Congress early in the legislative process as Congress is contemplating legislation. Because these types of communications come before legislation is drafted, they can take a variety of forms and are often speculative and preliminary in nature. These communications are often focused on describing an issue that requires a legislative solution, so that Congress is aware of it, rather than attempting to resolve the issue. To the extent these communications propose resolutions, they often describe various potential solutions and explain the pros and cons of each without getting to the level of technical legislative language. 72 For example, the America Invents Act (AIA) was enacted after many years of negotiation between the United States Patent and Trademark Office (USPTO) and various congressional committees. The USPTO prepared an early version of the legislation, then proceeded to send at least six views letters and various reports to congressional committees in the following years. Letter from Gary Locke to the Honorable Patrick J. Leahy, Chairman, Senate Comm. on the Judiciary, and the Honorable Jeff Sessions, Ranking Member, Senate Comm. on the Judiciary (Apr. 20, 2010) (on file with U.S. Dep’t of Commerce) (stating that the new post-grant review procedures “will serve as a faster, lower-cost alternative to litigation.”); Letter from John J. Sullivan to the Honorable Patrick J. Leahy, Chairman, Senate Comm. on the Judiciary, and the Honorable Arlen Specter, Ranking Member, Senate Comm. on the Judiciary (May 18, 2007) (on file with U.S. Dep’t of Commerce); Letter from Nathaniel F. Wienecke to the Honorable Patrick J. Leahy, Chairman, Senate Comm. on the Judiciary 4 (Feb. 4, 2008) (on file with U.S. Dep’t of Commerce) (“The Administration supports establishment of an effective, efficient post-grant patent review process that truly functions as a lower-cost alternative to litigation . . . .”). This bill was subject to much litigation, including a case that reached the Supreme Court very recently, Cuozzo Speed Technologies, LLC v. Lee, 136 S. Ct. 2131 (2016), but the Court made no mention of this agency legislative history.

Because these types of communications tend to be relatively broad and are often sent to Congress long before legislation is drafted, they will usually be less useful than other types of agency legislative history and are very unlikely to be dispositive. To the extent they are useful, it will generally be as background to uncover the agency–Congress relationship and why Congress chose to legislate in the manner it did.

4. Agency Testimony

Another way in which agencies participate in the legislative process is through testimony in congressional hearings. 73 Agencies generally post this testimony on their websites. See, e.g., Congressional Testimony, U.S. Dep’t of Just., Off. of Legis. Aff., https://www.justice.gov/ola/congressional-testimony (last visited Oct. 20, 2018); Congressional Testimony, U.S. Dep’t of Lab., http://www.dol.gov/newsroom/congress (last visited Oct. 20, 2018); Congressional Testimony, U.S. Dep’t of St., Bureau of Legis. Aff., https://www.state.gov/s/h/tst/ (last visited Oct. 20, 2018); Testimony, Dep’t of Homeland Security, https://www.dhs.gov/news-releases/testimony (last visited Oct. 20, 2018); Testimony, U.S. Dep’t of Transp., Off. of Inspector Gen., https://www.oig.dot.gov/testimony (last visited Oct. 20, 2018). Congress regularly invites agencies to testify about particular issues or proposed legislation. This testimony is drafted within an agency the same way a legislative proposal would be: the relevant bureau creates a draft and then that draft goes through an internal agency clearance process and OMB clearance before it is submitted to Congress. 74See Dep’t of Def., Office of Legislative Counsel, supra note 40, at 5 (“Congress frequently invites DoD leaders to testify before various committees and subcommittees. Because the subject of the testimony often crosses jurisdictional boundaries with other DoD components and government agencies, once again extensive coordination is required to ensure that DoD, and ultimately the entire Administration, speak with one voice.”). This testimony is therefore meant to reflect official administration policy.

Scholars and judges already consider congressional hearings to be a type of legislative history, since hearings are almost always publicly available. 75 For example, the Government Publishing Office maintains a record of hearings for each House and Senate Committee. See Browse by Committee, Govinfo, https://www.govinfo.gov/browse/committee (last visited Oct. 20, 2018) (listing each House and Senate committee with links to congressional hearings for each committee). However, scholars have not emphasized the importance of agency testimony. 76 Patricia M. Wald, Some Observations on the Use of Legislative History in the 1981 Supreme Court Term, 68 Iowa L. Rev. 195, 202 (1983) (“The hornbook rule that hearings are relevant only as background to show the purpose of the statute no longer holds. In many cases the best explanation of what the legislation is about comes from the executive department or outside witnesses at the hearings.”). Because agencies are closely involved in drafting and revising legislation, their testimony is likely to be informative and accurate in explaining how legislation is intended to work. It is also very likely that committee members form their opinions on legislation based on how it is described by agencies and rely on representations made by agencies of how legislation will be carried out after enactment.

5. Courts’ Use of Agency Legislative Analysis

Courts, mostly in the pre-Chevron era, have occasionally considered the various types of legislative analyses described above. For example, in United States v. Vogel Fertilizer Co., the Supreme Court looked to a “General Explanation” accompanying a bill proposed by the Department of Treasury. 77 455 U.S. 16, 29 (1982). This document is similar to a section-by-section analysis in that it was drafted by the Department to explain the purpose and operation of specific legislative provisions in the agency’s proposal. Congress enacted the legislative language proposed by the Department of Treasury, and a few years later the Department promulgated implementing regulations. 78Id. at 28. The Court looked to the written explanations provided by the Department to Congress with the proposed legislation and determined that they were “wholly incompatible” with the agency’s interpretation in the regulation. 79Id. at 31–32. This position was bolstered by the fact that the House Committee Report adopted language similar to the Treasury Department’s explanations. 80Id. at 32.

Other courts have also considered these types of agency legislative analyses. In Watkins v. Blinzinger, Judge Easterbrook, a noted textualist, approached the question of whether personal injury awards are income for purposes of determining whether a family qualified for the Aid to Families with Dependent Children program. 81 789 F.2d 474, 475 (7th Cir. 1986). The language at issue in the case was drafted by the Department of Health and Human Services, which was charged with carrying out the program. 82Id. at 475 n.2. As part of his analysis, Judge Easterbrook looked to a section-by-section analysis the agency provided to Congress—which the House committee also appended to its report—and to a summary of the draft bill that the Department provided with the legislative proposal. 83Id. at 479–80. Although these documents were not dispositive, Judge Easterbrook used the section-by-section analysis to support his ruling in favor of the agency by showing that the agency’s interpretation was in line with the stated purpose of the bill. 84Id. at 480.

Courts have occasionally referenced agency views letters, generally where these letters were included in a congressional committee report. In United States v. One Bell Jet Ranger II Helicopter, a case over the potential forfeiture of a helicopter used to hunt big horn sheep, the Ninth Circuit considered whether the forfeiture provision of the Airborne Hunting Act was subject to judicial discretion or agency discretion. 85 943 F.2d 1121, 1126 (9th Cir. 1991). The Department of Interior claimed that under the statute it was up to the agency whether to seize the helicopter. 86Id. The lower court ruled that it was in the court’s discretion and decided that forfeiture was not warranted. 87Id. at 1122. To resolve the question the Ninth Circuit noted that the statutory language was proposed by the Department of Interior and looked to a views letter written by the Assistant Secretary of the Interior, which was included in the Senate report of the bill. 88Id. at 1126 (citing S. Rep. No. 92-1157 (1972), as reprinted in 1972 U.S.C.C.A.N. 3835, 3839). The views letter advised the Senate that the language was intended “to confer upon the courts discretion to determine whether or not forfeiture of animals taken or equipment used in violation of the Act is appropriate in a particular case.” 89Id. The court noted the historic Supreme Court practice of granting deference where Congress enacts legislation proposed by an agency, but ruled against the agency’s interpretation in this case, instead relying on “the statement made to Congress by the agency at the very time it presented its own amendment to the Congress as one it urged for adoption, as the more reliable.” 90Id.

In another case, United States v. Sotelo, the Supreme Court considered whether statutory language should allow liability for taxes withheld on behalf of a third party to be dischargeable in bankruptcy. 91 436 U.S. 268, 270 (1978). The Court looked to two views letters sent from the Assistant Secretary of the Treasury to the House and Senate Judiciary Committees, and included in the House committee report, 92Id. at 276. expressing Treasury’s position that “any discharge of liability for collected withholding taxes was undesirable.” 93Id. The House committee noted that legislation was amended to eliminate Treasury’s opposition. 94Id. at 276–77 (explaining that the amendment was created “to meet the objection of Treasury to the discharge of so-called trust fund taxes”). The legislation, however, was not entirely clear, and the taxpayer argued that his withholding tax liability should be discharged. 95Id. at 272. The Court relied on the views letters to show that Congress had intended to ameliorate Treasury’s concern and to conclude that the liability at issue was not dischargeable in bankruptcy. 96Id. at 277.

Similarly, in Lindahl v. Office of Personnel Management97 470 U.S. 768, 784–85 (1985). the Court considered both agency testimony and views letters in deciding in favor of the agency’s interpretation. Congress had proposed a bill that would grant broad judicial review to determinations made by the Office of Personnel Management. 98Id. at 780. The Court noted that in committee hearings on the proposed bill, OPM testified in opposition to the bill as written, and in response, the committee amended the statute to limit the scope of judicial review. 99Id. at 784 (“Thereafter, the full Committee adopted an amendment in the nature of a substitute to H.R. 2510 that limited full judicial review ‘to cases involving agency-filed applications for disability retirement based on an employee’s mental condition.’”). The Director of OPM then sent both the House and Senate committees a views letter expressing support for the bill as amended, and these letters were included in the House and Senate committee reports. 100Id. at 784–85 (citing Letter from Alan K. Campbell, Dir., OPM, to Representative James M. Hanley, Chairman, House Comm. on Post Office and Civil Service (May 14, 1980), reprinted in H.R. Rep. No. 96-1080, at 2 (1980); Letter from Alan K. Campbell, Dir., OPM, to Senator Abraham A. Ribicoff, Chairman, Senate Comm. on Governmental Affairs (Sept. 25, 1980), reprinted in S. Rep. No. 96-1004, at 4–5 (1980). The Court relied on these statements to determine that Congress intended to restrict the scope of judicial review of OPM’s determinations. 101Id. at 785.

The Court has also considered letters sent from agencies to Congress in the pre-drafting stages of legislation, although less frequently than other types of agency legislative history and only where the pre-drafting documents turned out to be relevant to the legislative process. For example, in Thompson v. Thompson102 484 U.S. 174, 185 (1988). the Justice Department, knowing that Congress was contemplating drafting legislation, sent a letter outlining a variety of legislative options to deal with parental kidnapping. This letter was referred to extensively in the congressional debates over the legislation, which is probably why the Court viewed it as reliable legislative history. The letter focused on two options: either granting jurisdiction to federal courts to enforce state custody decrees or imposing on states the duty to give full faith and credit to custody decrees of other states. 103Id. The agency’s letter discussed the pros and cons of each approach and ultimately argued in favor of leaving states to enforce custody decrees and against the federal approach. 104Id. Although the enacted legislation did not preclude a federal cause of action, the Court relied on this letter to infer that Congress did not intend to allow a federal cause of action. 105Id. at 185–86.

Agency legislative analysis also commonly occurs during agency testimony in congressional hearings. Because this testimony is recorded in the Congressional Record and therefore always publicly available, courts have unsurprisingly cited to it frequently, most commonly when the agency was also involved in drafting the legislation. 106 For examples other than those discussed in the text here, see Wright v. City of Roanoke Redev. & Hous. Auth., 479 U.S. 418, 425 n.7 (1987); Pension Benefit Guar. Corp. v. R.A. Gray & Co., 467 U.S. 717, 722 n.2 (1984); Monroe v. Standard Oil Co., 452 U.S. 549, 558 (1981) (“The legislation was proposed by the Department of Labor. Accordingly, the testimony [of agency staff] . . . is instructive.”); Dawson Chem. Co. v. Rohm & Haas Co., 448 U.S. 176, 204 (1980) (relying on DOJ testimony as evidence that the legislature intended to expand patentee protection under Section 271(d) of the Patent Act); Friedsam v. Nicholson, 19 Vet. App. 222, 225–26 (2005) (reviewing agency testimony and determining that it was relevant to the case). Like other types of agency legislative history, the Court relied on agency testimony most commonly in the pre-Chevron era. For example, in Tennessee Valley Authority v. Hill107 437 U.S. 153, 179–80 (1978). the famous snail darter case, the Supreme Court looked to testimony by various officials from the Department of Interior to support its broad reading of protections for endangered species under the Endangered Species Act. 108Id. (“These provisions were designed, in the words of an administration witness, ‘for the first time [to] prohibit [a] federal agency from taking action which does jeopardize the status of endangered species,’ . . . . [T]he proposed bills would ‘[direct] all . . . Federal agencies to utilize their authorities for carrying out programs for the protection of endangered animals.’”). Similarly, in Zuber v. Allen, the Supreme Court stated that it gives greater deference to an agency interpretation where the agency participated in the drafting and “directly made known their views to Congress in committee hearings.” 109 396 U.S. 168, 192 (1969). In NLRB v. Servette, Inc.110 377 U.S. 46, 48–49 (1964). the Supreme Court considered whether a certain action by union workers was an “unfair labor practice.” The question was whether amendments to labor law that the Department of Labor proposed encompassed the union’s actions. 111Id. at 51. The Court noted that the provisions came from the agency and looked at testimony by the Secretary of Labor to confirm that the agency’s interpretation was within the intended scope of the legislative changes. 112Id. at 52 n.8, 53 n.9.

The Supreme Court has also used agency testimony to overturn agency interpretations. In Piper v. Chris-Craft Industries, Inc.113 430 U.S. 1, 33 (1977). the Court looked to the testimony of the SEC Chairman during the legislative process to demonstrate the purpose of the legislation. In that case, the SEC argued that the legislation was intended to protect both tender offerors and shareholders. 114Id. However, the Court looked to statements made by the SEC Chairman at the time the bill was being considered in Congress, which indicated that the legislation was targeted solely at shareholders and not tender offerors. 115Id. at 34–35. By looking to the agency legislative history, the Court attempted to ensure that later political changes did not upset the legislative bargain that led to enactment a number of years before.

In a case with similar reasoning, United States v. Giordano116 416 U.S. 505, 507–08 (1974). the Court considered a question of who had authority to approve wiretap applications. The Attorney General argued in favor of a broad ability to delegate approval of wiretaps, and argued specifically that authorization of a wiretap by the Attorney General’s executive assistant was not inconsistent with the statute. 117Id. at 512. The Court looked to testimony by the Department of Justice (DOJ) to resolve the issue. 118Id. at 516–17. Early versions of the bill granted the Attorney General broad authority to delegate authorization of wiretaps, and at the time of the earlier bills the DOJ had testified that the ability to delegate should be limited to only high-level officials. 119Id. at 516. Congress amended the bill to match the DOJ’s views. 120Id. at 520. The language was not immediately enacted but instead was included in later legislative proposals until it ultimately passed a number of years later. 121Id. at 517. The Court relied on the earlier DOJ testimony to show that the bill was meant to restrict who could authorize a wiretap to only high-level officials, and that the authorization at issue in the case was therefore unlawful. 122Id. at 520.

E. Accessibility of Agency Legislative History

One of the main contributions of this Article is to point out just how common agency legislative history is, yet how rarely it is referenced by judges and scholars. One explanation for this is that agency legislative history is only sporadically available to the public. A number of agencies post at least some of their legislative communications online. 123See supra Section II.E. Others post none. And an examination of various congressional committee reports reveals that Congress sometimes incorporates many of the types of agency legislative history described above into its own legislative history. 124 This is confirmed by agency staff who are involved in the legislative process. See Shobe, Agencies as Legislators, supra note 5, at 491 n.156 (“[A]gencies offer section-by-section analysis of bills drafted in an agency and supply other documents and studies to Congress, and respondents reported that these types of agency-produced documents can end up in a committee or conference report.”). Interestingly, Congress has, on occasion, incorporated agency legislative history into enacted statutory language. See, e.g., Medical Device User Fee and Modernization Act of 2002, Pub. L. No. 107-250, § 101(3), 116 Stat. 1588, 1589 (2002) (“[T]he fees authorized by this title will be dedicated to meeting the goals identified in the letters from the Secretary of Health and Human Services to the Committee on Energy and Commerce of the House of Representatives and the Committee on Health, Education, Labor, and Pensions of the Senate, as set forth in the Congressional Record.”); Food and Drug Administration Modernization Act of 1997, Pub. L. No. 105-115, § 101(4) , 111 Stat. 2296, 2298 (1997) (“[T]he fees authorized by amendments made in this subtitle will be dedicated toward expediting the drug development process and the review of human drug applications as set forth in the goals identified, for purposes of part 2 of subchapter C of chapter VII of the Federal Food, Drug, and Cosmetic Act, in the letters from the Secretary of Health and Human Services to the chairman of the Committee on Commerce of the House of Representatives and the chairman of the Committee on Labor and Human Resources of the Senate, as set forth in the Congressional Record.”). In fact, in most of the cases discussed in this Part, citations to agency legislative history are actually to agency documents that Congress incorporated into its legislative history. This is unsurprising in light of the relative inaccessibility of agency legislative history.

Most commonly, a congressional committee will include agency-produced documents as part of a committee report. For example, Congress sometimes incorporates an agency-drafted views letter into a committee report either in the body of the report or as an appendix. 125See, e.g., H.R. Rep. No. 112-98, at 85–88 (2011) (incorporating a views letter from the Department of Commerce); S. Rep. No. 110-143, at 17–23 (2007) (including both excerpts from a statement of Under Secretary for Benefits of the Department of Veterans Affairs and a letter from the Secretary of the Department of Veterans Affairs); S. Rep. No. 108-264, at 28–29 (2004) (including a letter from Director of Office of Congressional and Legislative Affairs of Dept. of Interior); S. Rep. No. 106-325, at 5–9 (2000) (including a letter from Henry H. Shelton, Chairman of the Joint Chiefs of Staff, and William S. Cohen, Secretary of Defense); S. Rep. No. 105-207, at 14–17 (1998) (including two letters from Kevin Gover, Assistant Secretary–Indian Affairs); H.R. Rep. No. 99-660, at 63–67 (1986) (including letters from John R. Bolton, Assistant Attorney General, United States Department of Justice). Committee reports also sometimes include reports from an agency or portions of agency testimony that are relevant to the enacted language. 126See, e.g., H.R. Rep. No. 114-396, at 13–14 (2016) (“[T]he Department of Homeland Security Inspector General recently released a report entitled ‘TSA Can Improve Aviation Worker Vetting’ . . . . The report made six recommendations to strengthen the vetting of credentialed aviation workers. This legislation codifies several of those recommendations and ensures that TSA has access to the necessary data to properly vet aviation employees, strengthen its criminal background check capabilities, and better-resolve issues of lawful status for credential applicants.”); S. Rep. No. 105-20, at 8–12 (1997) (including testimony from Gary Niles Kimble, Commissioner, Administration for Native Americans, Administration for Children and Families–U.S. Department of Health and Human Services); NAACP v. Detroit Police Officers Ass’n, 900 F.2d 903, 909 (6th Cir. 1990) (considering a statement from the Justice Department that had been placed in the Congressional Record); Krinsk v. Fund Asset Mgmt., Inc., No. 85 Civ. 8428 (JMW), 1986 WL 205, at *4 n.6 (S.D.N.Y. 1986) (citing to a memo drafted by the SEC and included in the House committee report). In one case, the Supreme Court even noted that both the House and Senate committee reports consisted “almost entirely of a letter and memorandum from Acting Secretary of the Interior,” and relied heavily on the committee reports in its interpretation. 127 Negonsott v. Samuels, 507 U.S. 99, 106–09 (1993). Sometimes the committee report will also describe interactions with the agency and suggestions from the agency that Congress chose to adopt or reject. 128See, e.g., H.R. Rep. No. 114-107, pt. 1, at 245 (2015) (“We even received a letter from the widely respected Secretary of Energy. This may well be the first time in the history of this Committee that a sitting Cabinet member has provided formal opposition to a piece of legislation that we are considering, certainly at this stage of the legislative process. That should be a strong indication of just how bad this bill really is.”); H.R. Rep. No. 111-97, at 4 (2009) (“Following the hearings, the legislation was refined to take into account concerns raised by the Department of Justice and potential defendants, and the False Claims Amendments Act of 1986 was enacted on October 27, 1986.”); S. Rep. No. 105-379, at 10 (1998) (“At the hearing held on July 8, 1998, the Department of the Interior testified that it could not support S. 391 unless section 9(d) (Affirmative Defenses Waived) was removed.”); Am. Tobacco Co. v. Patterson, 456 U.S. 63, 73 (1982) (noting that a Justice Department statement was placed in the Congressional Record by one of the bill’s sponsors); Int’l Bhd. of Teamsters v. United States, 431 U.S. 324, 351 (1977) (same); Franks v. Bowman Transp. Co., 424 U.S. 747, 759–60 (1976) (same).

While it is often clear when Congress has chosen to incorporate agency legislative history, other times Congress adopts agency legislative history as its own without attribution. For example, as discussed above, the DoD makes annual legislative proposals to Congress and posts these proposals and a section-by-section analysis to its website. 129See supra note 65. This makes it possible to compare the section-by-section analyses drafted by the DoD with the committee report produced by Congress. The author undertook a comparison of many of these documents, which revealed that language included in both Senate and House reports is often taken verbatim from the sectional analysis provided by the DoD, without attribution. 130Compare Dep’t of Def., 2008 Section-by-Section Analysis 10, 16 (2008), http://ogc.osd.mil/olc/docs/26April2013NDAASectionalAnalysis.pdf, with S. Rep. No. 110-77, at 398–99 (2008); compare Dep’t of Def., 2013 Section-by-Section Analysis 27 (2013), http://ogc.osd.mil/olc/docs/26April2013NDAASectionalAnalysis.pdf, with S. Rep. No. 112-173, at 103 (2012); compare Dep’t of Def., 2014 Section-by-Section Analysis 14, 68, 136 (2014), http://ogc.osd.mil/olc/docs/26April2013NDAASectionalAnalysis.pdf, with H.R. Rep. No. 113-102, at 173–74, 187, 362 (2014); compare Dep’t of Def., 2016 Section-by-Section Analysis 141 (2016), http://ogc.osd.mil/olc/docs/26April2013NDAASectionalAnalysis.pdf, with S. Rep. No. 114-49, at 178 (2016). Other comparisons of agency comments and congressional legislative history reveal that Congress is likely heavily influenced by agencies when creating legislative history, even if that influence is not clearly acknowledged in Congress’s legislative history. 131 For example, in the America Invents Act, the USPTO was especially influential in drafting and revising various versions of the bill. The USPTO sent a number of views letters to the relevant committees and these views letters used language very similar to language that ended up in the congressional committee reports about the purpose and function of the post-grant review proceedings that were an important part of the AIA. Compare Letter from Gary Locke to the Honorable Patrick J. Leahy, Chairman, Senate Comm. on the Judiciary, and the Honorable Jeff Sessions, Ranking Member, Senate Comm. On the Judiciary, supra note 72, and Letter from Nathaniel F. Wienecke to the Honorable Patrick J. Leahy, Chairman, Senate Comm. on the Judiciary, supra note 72, with Cuozzo Speed Techs., LLC v. Lee, 136 S. Ct. 2131, 2143 (2016) (“Inter partes review is a ‘quick and cost effective alternativ[e] to litigation.”) (citing to the House Committee Report), and id. (“Inter partes review is ‘a quick, inexpensive, and reliable alternative to district court litigation”) (citing S. Rep. No. 110-259, at 20 (2008)). The House Committee Report for the AIA included a views letter from the Department of Commerce. See H.R. Rep. No. 112-98, at 85–88 (2011). Of course, it is impossible to tell exactly how common this is without access to all agency legislative history.

To increase accessibility, litigants could try to request agency legislative documents through FOIA, but they would have to know what to ask for and agencies may not keep good track of documents, especially older ones. 132 The website governmentattic.org has posted certain DOJ views letters acquired through FOIA requests on its website. See, e.g., Copies of Certain Views Letters, 2001-2005, supra note 38. As part of this project the author made a number of FOIA requests for agency legislative history and while some agencies were able to provide the relevant documents, other agencies were unable to find them, especially those created many years ago. Requiring litigants to use FOIA requests to uncover agency legislative history is an imperfect solution to the access problem. Were courts to begin to rely on agency legislative history more generally, access could become an even bigger issue. Because agencies have access to their own documents but the public does not, they could use agency legislative history strategically in litigation to support their positions, and leave it out when it hurts their positions. If courts are going to rely on agency legislative history, then it needs to be available to those who wish to challenge agency interpretations, not just agencies wishing to defend their interpretations. As it stands now, if courts were to rely on all agency legislative history, whether or not Congress adopts it, agencies would be able to use their privileged position to cherry-pick helpful agency legislative history and suppress unhelpful agency legislative history. It could also allow agencies to fill the agency legislative history with language that does not reflect the legislative deal, and that was never approved by Congress, so that they could reference it in the future to support their interpretation. 133 One solution to level the playing field would be to make the types of formal agency legislative history discussed here available in easily searchable online sources like the U.S. Code Congressional and Administrative News, although this would obviously be a costly and burdensome project. Precedent exists for this type of project. The Reagan Administration did this with presidential signing statements when it entered into an agreement with West Publishing Company to publish these statements. Marc N. Garber & Kurt A. Wimmer, Presidential Signing Statements as Interpretations of Legislative Intent: An Executive Aggrandizement of Power, 24 Harv. J. on Legis. 363, 367 (1987) (citing Att’y Gen. Edwin Meese III, Address to the National Press Club, Washington, D.C. (Feb. 25, 1986)) (“[W]e have now arranged with the West Publishing Company that the presidential statement on the signing of a bill will accompany the legislative history from Congress so that all can be available to the court for future construction of what the statute really means.”). However, this would not necessarily address the concerns raised here, because it would still allow agencies to fill the legislative record with their preferred interpretations without congressional oversight.

The best solution to this accessibility problem would be for courts to consider only agency legislative history that Congress incorporates into its own legislative history, either in a congressional document like a committee report or through other recorded communications that end up in the Congressional Record. By consulting only agency legislative history that shows up in Congress’s legislative history, courts could shape Congress’s generation of legislative history by putting it on notice that courts will consider materials incorporated into Congress’s legislative history, but only those materials. 134See Abbe R. Gluck et al., Unorthodox Lawmaking, Unorthodox Rulemaking, 115 Colum. L. Rev. 1789, 1857–58 (2015) (“[M]any of the Court’s interpretive rules aim to improve how Congress drafts. . . . Textualists have long argued that one salutary effect of a text-centric approach is that it teaches Congress to draft better the next time.”). Of course, there is a debate over whether courts should be trying to shape congressional behavior at all, or merely reflect it. See id. at 1835–36. This would put the burden on Congress to ensure that only agency legislative history that accurately reflects the legislative deal is considered by courts, while also reducing the litigation costs associated with finding and parsing all of the agency legislative history. It would also reduce the incentive of agencies to generate agency legislative history to influence judges rather than communicate honestly with Congress. 135 Critics of legislative history have argued that the more legislative history is used, the less reliable it becomes, and this same argument could be leveled against agency legislative history. Blanchard v. Bergeron, 489 U.S. 87, 98–99 (1989) (Scalia, J., concurring in part and concurring in the judgment); see also Int’l Bhd. of Elec. Workers, Local Union No. 474 v. NLRB, 814 F.2d 697, 717 (D.C. Cir. 1987) (Buckley, J., concurring) (arguing that the use of legislative history by courts creates an incentive “to salt the legislative record with unilateral interpretations of statutory provisions they were unable to persuade their colleagues to accept”). For example, Congress could insert an exact copy of a views letter it receives from an agency in committee reports, could note in the Congressional Record when it introduces a bill that was drafted by an agency, could note where a section-by-section analysis in a committee report came from an agency, and could even describe any disagreements with the agency in the Congressional Record. As discussed above, Congress has done this in the past, 136See infra Part III (discussing courts’ use of agency legislative history, including instances where the agency legislative history is included in congressional legislative history). but not as frequently and explicitly as it could.

Including more agency legislative history in congressional legislative history would require more work from Congress, although the benefits to Congress appear to justify the costs. 137 It could be that members of Congress or their staff want to take credit for drafting the legislative history because of some kind of prestige or status benefits associated with being the drafter. It seems that this benefit would be outweighed by the benefits described here to Congress from memorializing agency–Congress legislative deals. It would not require Congress to draft more legislative history, but simply to include the agency legislative history that it considers to be reliable and authoritative in its own legislative history. The benefit to Congress would be twofold: First is that it would memorialize agency–Congress deals in a way that is easily accessible and salient to agencies, which would make congressional legislative history more valuable to future administrators attempting to implement statutory language. Second is that it would allow Congress to make it clear to courts where they have relied on agency–legislative communications so that courts could hold agencies accountable for these communications in future implementation. To the extent Congress is concerned about the expanding power of the administrative state, this would constrain agency interpretations in ways that would more closely align with the enacting Congress’s intent.

III. Implications of Agency Legislative History

Agency legislative history has implications for the ongoing debates about agency deference and modes of statutory interpretation. Arguments about deference are often empirical in nature, and this Part argues that agency legislative history can help uncover the empirical realties of agency–Congress interactions in the legislative process. Similarly, arguments about interpretation often have an empirical component, and agency legislative history provides new empirical background that could be relevant to how judges approach interpretation. This Part considers these implications.

A. Implications for Deference

This section discusses the implications the under-explored world of agency legislative history could have for important debates surrounding agency delegation theory and practice. A number of scholars and judges have advocated for a broad conception of Chevron that allocates interpretive authority to agencies instead of the judicial branch. 138Vermeule, supra note 12, at 206; Eskridge, Jr., supra note 10; Mashaw, supra note 10, at 93–95; Spence & Cross, supra note 10, at 138–41. These arguments are often based on agencies’ expertise and relative institutional competence. 139See supra note 138. Agency legislative history lends support to these arguments by showing that agencies are closely involved in the legislative process and that the pool of materials judges would need to obtain and understand is larger than typically thought. It also appears that concerns surrounding deference expressed by some scholars and judges may be overstated in light of the legislative relationships between Congress and agencies.

It seems unlikely, however, that courts will be willing to entirely defer to agencies. The passing of Justice Scalia, who was the Court’s strongest proponent of broad Chevron deference, 140See infra notes 147–57 and accompanying text. and the recent confirmation of Justice Gorsuch, who often appeared to be a Chevron skeptic as a circuit court judge, makes it more likely that the Court will look for ways to limit Chevron141See, e.g., SAS Inst., Inc. v. Iancu, 138 S. Ct. 1348, 1358 (2018). Judges and scholars have engaged in seemingly endless debates about when Chevron-style deference should apply without reaching any consensus. 142See infra Sections III.A.1.–III.A.3. For judges who are skeptical of broad deference to agencies and are looking to return to a pre-Chevron approach to deference that is more contextual, agency legislative history will often provide useful background that would allow courts to defer on a case-by-case basis in ways that are more likely to reflect congressional intent. 143See supra Section II.D. Indeed, as discussed in above, pre-Chevron courts often looked to agency legislative history as a means of discerning whether to defer to an agency interpretation, so in many ways this would simply be a return to prior judicial practice. This section considers in more detail what agency legislative history might mean for the many debates surrounding agency delegation.

1. Support for Deference

Arguments in support of a broad application of Chevron generally turn on institutional arguments about courts’ ability to determine congressional intent to defer. Those who support broad deference argue that bills are a series of complex deals that go through various committees, internal logrolling, confidential discussions, and negotiations with various lobbyists. 144See, e.g., Frank H. Easterbrook, The Role of Original Intent in Statutory Construction, 11 Harv. J.L. & Pub. Pol’y 59, 64–65 (1998) (discussing complexities of the legislative process); Kenneth A. Shepsle & Barry R. Weingast, The Institutional Foundations of Committee Power, 81 Am. Pol. Sci. Rev. 85, 89 (1987) (same). They argue that it is impossible for a court to reconstruct the legislative process to determine what Congress would have done if it was presented with the issue in any particular case, and that attempts to do so are costly. 145Vermeule, supra note 12, at 190; Frank H. Easterbrook, Statutes’ Domains, 50 U. Chi. L. Rev. 533, 548 (1983) (arguing that the many complexities of the legislative process are “so integral to the legislative process that judicial predictions of how the legislature would have decided issues it did not in fact decide are bound to be little more than wild guesses”). As Professor Vermeule has noted, legislative history “contains an astonishing range of unusual material produced in various ways by various actors.” 146Vermeule, supra note 12, at 113. Relatedly, Justice Scalia argued that deference should not depend on determining any actual legislative intent to defer, which he deemed a “wild-goose chase” 147 Antonin Scalia, Judicial Deference to Administrative Interpretations of Law, 1989 Duke. L.J. 511, 517 (1989). that is merely “an invitation to make an ad hoc judgment regarding congressional intent.” 148 City of Arlington v. FCC, 569 U.S. 290, 306 (2013). Even non-textualists agree that it is difficult or impossible for a court to determine whether Congress would have wanted to defer in any particular instance. 149See Henry M. Hart, Jr. & Albert M. Sacks, The Legal Process: Basic Problems in the Making and Application of Law 1183 (William N. Eskridge, Jr. & Philip P. Frickey eds., 1994); William N. Eskridge, Jr. & Philip P. Frickey, Statutory Interpretation as Practical Reasoning, 42 Stan. L. Rev. 321, 348–49 (1990). Agency legislative history adds weight to these arguments by showing that the creation of legislation is a multi-layered process that often turns on bargains necessary to achieve enactment that would be impossible for a court to uncover. 150 Easterbrook, supra note 145, at 547–48.

This Article’s findings indicate that there are even more documents a court would need to find and understand to be able to accurately reconstruct the many legislative deals that lead to enactment of a bill. For a generalist court to become fully informed of the legislative process and purposes would require a level of resources and expertise that may be beyond their capacity, especially for lower courts with high caseloads. This complex, behind-the-scenes process, especially as it relates to communications between agencies and Congress, is one that courts may be more likely to disrupt than improve upon. Perhaps, in light of agency legislative history, we should be even more skeptical of courts’ ability to enter into the governmental “black box” 151 Pildes, supra note 13. and should instead fall back to a more formalist approach that, while “fictional,” provides the best background presumption against which Congress can legislate. 152 United States v. Mead Corp., 533 U.S. 218, 240–45 (2001) (Scalia, J., dissenting).

Agency legislative history not only calls into question courts’ institutional capacity to uncover congressional intent to delegate, but it also supports arguments, first expressed by Professors Mashaw and Strauss, that agencies may be better statutory interpreters than courts. 153See, e.g., Vermeule, supra note 12; Mashaw, supra note 12 (“There are persuasive grounds for believing that legitimate techniques and standards for agency statutory interpretation diverge sharply from the legitimate techniques and standards for judicial statutory interpretation”); Sunstein & Vermeule, supra note 12. Professors Walker and Sitaraman found support for this idea in their discussions of agencies’ role in the legislative process. 154 Sitaraman, supra note 5, at 128 (“The executive’s role in legislative drafting provides additional support to the Strauss-Mashaw thesis that agency interpretive practice can and should diverge from judicial interpretive process.”); Walker, Legislating in the Shadows, supra note 5, at 1403 (arguing that the role of agencies in the legislative process provides “additional empirical support for a more purposivist approach to agency statutory interpretation”). These arguments are based on the idea that agencies may have internal understandings of what legislation means from their involvement in the legislative process, something judges and scholars have long speculated. 155See infra notes 156–58. For example, Justice Breyer said that “[t]he agency that enforces the statute may . . . possess an internal history in the form of documents or ‘handed-down oral tradition’ that casts light on the meaning of a difficult phrase or provision.” 156 Stephen Breyer, Judicial Review of Questions of Law and Policy, 38 Admin. L. Rev. 363, 368 (1986); see also Peter L. Strauss, When the Judge Is Not the Primary Official with Responsibility to Read: Agency Interpretation and the Problem of Legislative History, 66 Chi.-Kent L. Rev. 321, 329–30 (1990) [hereinafter Strauss, When the Judge Is Not the Primary] (noting that agencies maintain “transcripts of relevant hearings, correspondence, and other informal traces of the continuing interactions that go on between an agency and Capitol Hill as a statute is being shaped in the legislative process . . . .”). Justice Scalia similarly noted that agencies often have “intense familiarity with the history and purposes of the legislation at issue.” 157 Scalia, supra note 147, at 514. Professor Vermeule posited that “[a]gencies will often possess far better information about the legislative process that produced the statute, about the specialized policy context surrounding the statute’s enactment, and about the resulting legislative deal.” 158Vermeule, supra note 12, at 209. These arguments have always been based on speculation about the institutional roles of agencies.

This Article shows that this speculation is indeed correct and that agencies have their own rich legislative repositories of the types discussed here that record and explain legislative deals and purposes. These documents are often unavailable to courts and may be difficult for courts to understand and apply, but they likely serve as a useful interpretive tool for those within agencies. Agencies are able to use the institutional knowledge contained in agency legislative history to recognize when congressional legislative history is speaking to the agency, or even when congressional legislative history incorporates agency–Congress legislative communications, in a way a court could not. Agencies are therefore better able to separate reliable legislative history from “cheap talk.” 159 Daniel B. Rodriguez & Barry R. Weingast, The Positive Political Theory of Legislative History: New Perspectives on the 1964 Civil Rights Act and Its Interpretation, 151 U. Pa. L. Rev. 1417, 1445–46, 1448 (2003); Strauss, When the Judge Is Not the Primary, supra note 156, at 347 (arguing that agencies are better able “to distinguish reliably those considerations that served to shape the legislation, the legislative history wheat, from the more manipulative chaff”). Agency legislative history provides support for the idea that agencies should be better interpreters than courts, and therefore that a court would be better off deferring to agencies than attempting to insert itself into something that it does not have the institutional competence to do well.

Agency legislative history also adds to the argument that Chevron is more likely to reflect congressional intent, which is sometimes used to justify the Chevron doctrine. 160 The clearest articulation of this at the Supreme Court is in Smiley v. Citibank (South Dakota), N.A.: We accord deference to agencies under Chevron . . . because of a presumption that Congress, when it left ambiguity in a statute meant for implementation by an agency, understood that the ambiguity would be resolved . . . by the agency, and desired the agency (rather than the courts) to possess whatever degree of discretion the ambiguity allows.517 U.S. 735, 740–41 (1996). Congressional staff have made it clear that they view “Congress’s primary interpretive relationship as one with agencies, not with courts.” 161 Gluck and Bressman, Part II, supra note 5, at 765. Agency legislative history may provide documentation for why that is: agencies and Congress are engaging in deep and complex legislative communications throughout the legislative process. Agencies have direct lines of communication with Congress in a way that courts do not, and the extent of these communications are evidence of why agencies would be Congress’s preferred interpreter.

2. Concerns with Deference

A number of Supreme Court Justices have expressed concern about the ever-expanding powers of administrative agencies while questioning an unlimited application of the Chevron doctrine. 162See, e.g., Michigan v. EPA, 135 S. Ct. 2699, 2712 (2015) (Thomas, J., concurring in the judgment) (arguing that the EPA’s “request for deference raises serious questions about the constitutionality of our broader practice of deferring to agency interpretations of federal statutes”); King v. Burwell, 135 S. Ct. 2480, 2488–89 (2015); Dep’t of Transp. v. Ass’n of Am. R.Rs., 135 S. Ct. 1225, 1254 (2015) (Thomas, J., concurring) (describing the administrative state as a system that “concentrates the power to make laws and the power to enforce them in the hands of a vast and unaccountable administrative apparatus”). For an excellent discussion of the rise of the administrative state, see Lawson, supra note 16. As Chief Justice Roberts said in his dissent in City of Arlington v. FCC, “the danger posed by the growing power of the administrative state cannot be dismissed.” 163 City of Arlington v. FCC, 569 U.S. 290, 314–15 (2013). Similar concerns have also been raised by scholars and even some within Congress. 164See, e.g., Lisa Schultz Bressman, supra note 64 (arguing against automatic deference). The House recently passed a bill aimed at regulatory agencies that directly repudiated Chevron. Separation of Powers Restoration Act, H.R. 5, 115th Cong. § 202(1)(B) (2017) (“If the reviewing court determines that a statutory or regulatory provision relevant to its decision contains a gap or ambiguity, the court shall not interpret that gap or ambiguity as [1] an implicit delegation to the agency of legislative rulemaking authority . . . or [2] a justification for interpreting agency authority expansively or for deferring to the agency’s interpretation on the question of law.”). It is possible that broad judicial deference to agencies makes Congress less willing to pass legislation, especially in times of divided government, because Congress cannot rely on courts to ensure agencies carry out Congress’s will. If courts began to consider agency legislative history, then legislative gridlock might be, at least somewhat, reduced. These concerns are often constitutional in nature, based on understandings of separation of powers and a fear of the rise of the administrative state in ways that could not have been imagined by the Framers of the Constitution. 165See, e.g., Michigan, 135 S. Ct. at 2712 (Thomas, J., concurring in the judgment) (arguing that the transfer of interpretive authority from courts to agencies “is in tension with Article III’s Vesting Clause, which vests the judicial power exclusively in Article III courts, not administrative agencies”); Gutierrez-Brizuela v. Lynch, 834 F.3d 1142, 1149 (10th Cir. 2016) (Gorsuch, J., concurring) (“[T]he fact is Chevron and Brand X permit executive bureaucracies to swallow huge amounts of core judicial and legislative power and concentrate federal power in a way that seems more than a little difficult to square with the Constitution of the framers’ design. Maybe the time has come to face the behemoth.”); Philip Hamburger, Is Administrative Law Unlawful? 315–17 (2014). Agency legislative history cannot resolve this debate, although it does add context to it by showing that Congress may be more sophisticated than judges and scholars give it credit for. Congress has been able to manage and benefit from the rise of the administrative state better than is typically acknowledged. Congress seeks out input from various experts, agencies among them, and weighs options as it legislates. Agency legislative history shows that Congress is capable of both accepting and rejecting agency legislative requests and proposals based on its policy goals, which provides evidence that Congress may not be as beholden to agencies as some have feared. 166 Walker, supra note 5, at 1419. Agency legislative history also raises questions of whether the Chief Justice’s concerns about the power of the administrative state are well-founded. Judges and scholars who advocate for reduced deference often do so under the guise of protecting Congress from an out-of-control administrative state, but agency legislative history might cut back on that notion by showing that Congress may be able to take care of itself, and that in any case courts may be more likely to disturb an agency–Congress legislative bargain than improve upon it.

Professors Walker and Sitaraman raise another concern about Chevron deference in light of agency involvement in the legislative process. They have argued that granting deference to agencies that are also involved in creating legislation implicates similar concerns to those raised by Justice Scalia and others with respect to Auer deference, which grants deference to agency interpretations of their own regulations. 167 Auer v. Robbins, 519 U.S. 452, 461 (1997). Justice Scalia criticized Auer deference because “deferring to an agency’s interpretation of its own rule encourages the agency to enact vague rules which give it the power, in future adjudications, to do what it pleases.” 168 Talk Am., Inc. v. Mich. Bell Tel. Co., 564 U.S. 50, 69 (2011) (Scalia, J., concurring). Following this line of reasoning, Professor Walker argues that the fact that agencies are both involved in the legislative drafting process and are charged with implementing statutes potentially allows agencies to engage in self-dealing by drafting broad statutes that delegate significant leeway to themselves. 169 Walker, supra note 5, at 1411 (“If the agency is indeed a partner with Congress in legislative drafting, Justice Scalia’s concern about an agency legislating and executing the law should apply with some force to legislative drafting. The executive and legislative functions are, in essence, combined via agency legislating in the shadows.”). He argues that these concerns with agency self-dealing arising from agency involvement in the legislative process may be the “last straw” for Chevron and that the appropriate delegation presumption going forward may be the less deferential Skidmore standard. 170Id. at 1380–81.

Professor Walker’s analogy between Auer and Chevron certainly sounds alarming. If agencies have usurped Congress’s legislative process and are both primary creators and implementers of legislation, then a solution of less judicial deference seems reasonable. However, Professor Walker’s study, and other empirical studies to date, provide no evidence that Congress has ceded all, or even a significant portion of, the creation of statutes to agencies. 171See, e.g., James J. Brudney, Essay, Contextualizing Shadow Conversations, 166 U. Pa. L. Rev. Online 37, 41–42 (2017) (recounting the author’s experience as a congressional staffer and that this experience did not indicate that Congress ceded authority to agencies and instead indicated that “the congressional drafting process is complex, messy, and far from uniform”); Gluck & Bressman, Part I, supra note 1, at 1000 (reporting “that agencies often . . . participate in drafting and can be very useful partners,” but that congressional staff “sometimes avoided the agency, particularly when they were aware of a conflicting position”). The agency legislative history reviewed as part of this Article certainly does not indicate a legislative process that is controlled by agencies. Congress requests technical and substantive assistance from all kinds of interest groups, lobbyists, and outside experts while creating legislation. 172 Gluck & Bressman, Part II, supra note 5 (“[O]ur respondents told us that first drafts are typically written by, respectively, the White House and agencies, or policy experts and outside groups, like lobbyists. Empirical work is lacking for the details of this account . . . .”); see Kenneth W. Starr, Observations About the Use of Legislative History, 1987 Duke L.J. 371, 376 (“Lobbyists maneuver to get their clients’ opinions into the mass of legislative materials . . . .”). Lobbyists and other outside experts understand how legislation will work on the ground in much the same way agencies do, since they represent those who must comply with legislation. 173 Victoria F. Nourse & Jane S. Schacter, The Politics of Legislative Drafting: A Congressional Case Study, 77 N.Y.U. L. Rev. 575, 611 (2002) (“Lobbyists are the closest to the people who will be affected by the bill . . . .”). These outside groups certainly help close the expertise gap between agencies and Congress by reviewing agency drafting and alerting Congress to potential issues or potential overdelegations. 174 Shobe, Intertemporal Statutory Interpretation, supra note 5, at 847–49. Spending on lobbyists has increased significantly in recent years, so there is reason to believe they are taking an active role in policing legislative language. Id. at 847 (“Recent reports show that spending on lobbying has more than doubled in the last fifteen years, from $1.45 billion in 1998 to $3.3 billion in 2012. Long-serving committee staff and legislative counsel anecdotally report that lobbyist involvement in the drafting process has increased significantly over the last twenty years.”). This is not to understate the role of agencies, which is deep and significant, but to point out that this Auer concern understates the complexity of the legislative process. Congress, and the lobbyists who work closely with Congress, are not naive, and it seems unlikely that they have broadly ceded the legislative drafting process to agencies.

While a blanket presumption against delegation would be an overly broad response to agency involvement in the legislative process, judges concerned about agency self-dealing could use agency legislative history to determine deference in more targeted ways. For example, sometimes agency legislative history will reveal that an agency was hardly involved in creating the legislation or that Congress rejected an agency’s legislative suggestions. It would seem odd in those cases for a judge not to defer based on an Auer-type concern. Other times the agency legislative history could reveal that Congress accepted a draft of the legislation from an agency with few or no changes. In that case, perhaps a judge would be justified in raising concerns about potential agency self-dealing. Agency legislative history would allow a judge with these concerns to tailor deference to specific circumstances. This type of context-specific approach to Chevron is discussed in more detail in the next section.

Interestingly, courts have never raised this Auer concern about agency involvement in the legislative process even though, as discussed above, they have considered agency involvement in drafting as a factor in deciding whether to uphold an agency interpretation. 175See supra Section II.B. Courts appear to take agency involvement as a signal from Congress of an intent to defer to the agency rather than as raising an issue of potential self-dealing. For example, the majority in Zuni Public School District No. 89 v. Department of Education relied on the fact that an agency was the original drafter of the language at issue as a sign that Congress must have intended courts to defer to agency interpretations of the statute. 176 550 U.S. 81, 98, 107 (2007). The Court did not explicitly say why, but it seems likely that the Court was less concerned about agency self-dealing and more concerned about disrupting a likely agency–Congress legislative understanding. Even when courts have chosen not to defer when an agency was involved in the legislative process, they have done so only because the interpretation was not close enough in time to the enactment of the statute, while still noting that deference is generally due to agencies that are involved in creating the statute at issue. 177See Barnett v. Weinberger, 818 F.2d 953, 960–61 (D.C. Cir. 1987); cases cited supra note 60 and accompanying text. In fact, the Court has viewed lack of agency involvement in creating legislation as a reason not to defer to an agency interpretation, 178See supra Section III.B. (discussing Nacirema Operating Co. v. Johnson, 396 U.S. 212 (1969)). which again emphasizes the Court’s view that agency involvement should be viewed as favoring agency interpretations. So it may be difficult to sell judges on the idea that agency involvement in creating legislation should cause them to defer less rather than more.

Perhaps a better argument against deference in light of agency legislative history than the Auer concern is that agency legislative history makes it clear that agencies have the opportunity to influence legislative drafting before enactment, so if they want Congress to enact policy they should use their influence to do it legislatively. If an agency cannot get something in the legislative text, the argument would be, then courts should not defer to that agency and should instead conduct their own judicial review of the statutory text. This would put agencies on notice that they need to resolve ambiguities during the legislative process if they want to avoid judicial review and that if a statute is ambiguous, a court will substantively review the agency’s work. This argument seems to be the more appealing grounds for limiting Chevron deference than concerns of agency self-dealing or collusion, although it would impose significant costs on both courts and agencies and is unlikely to accurately reflect congressional preferences.

3. Contextual Deference

Scholars have written volumes about when and how Chevron should apply, and many have proposed other assumptions courts could apply in lieu of, or in addition to, Chevron179See Lisa Schultz Bressman, Essay, Reclaiming the Legal Fiction of Congressional Delegation, 97 Va. L. Rev. 2009, 2015, 2025 (2011) (demonstrating that a consensus exists in the academic literature that Chevron is based on a fiction); Sitaraman, supra note 5, at 129; Strauss, Chevron Space” and “Skidmore Weight”, supra note 26, at 1145 (arguing that courts should use a case-by-case common law approach to determine when to give deference); Walker, supra note 5, at 1409. In City of Arlington, Chief Justice Roberts looked to the wording the Court used in Chevron to propose a case-by-case determination of “whether Congress had ‘delegat[ed] authority to the agency to elucidate a specific provision of the statute . . . .’” 180 City of Arlington v. FCC, 569 U.S. 290, 318 (2013) (Roberts, C.J., dissenting). In the same case, Justice Breyer wrote a concurrence that attempted to provide additional guidance on how this context-specific approach could work. Quoting Barnhart v. Walton, he listed a number of factors the Court had already considered when deciding whether to defer, including “the interstitial nature of the legal question, the related expertise of the Agency, the importance of the question to administration of the statute, the complexity of that administration, and the careful consideration the Agency has given the question over a long period of time.” 181Id. at 309 (Breyer, J., concurring in part and concurring in the judgment) (quoting Barnhart v. Walton, 535 U.S. 212, 222 (2002)). This echoes Justice Breyer’s earlier writings in favor of a flexible approach to deference. 182 Breyer, supra note 156, at 371–73. The Court itself, in Mead, left open the door for considering a wide range of factors by noting that courts can look for “circumstances reasonably suggesting that Congress . . . thought of [the agency interpretation] as deserving . . . deference.” 183 United States v. Mead Corp., 533 U.S. 218, 231 (2001). Mead famously restricted the application of Chevron to cases where Congress explicitly delegated, and agencies followed, relatively formal procedures. See id. at 226–27; Cass R. Sunstein, Beyond Marbury: The Executive’s Power to Say What the Law Is, 115 Yale L.J. 2580, 2602–03 (2006).

Prominent scholars have also argued for a move toward a more context-specific approach to delegation similar to that proposed by Justice Breyer. Professor Strauss has argued for courts to engage in common law reasoning to determine deference where Congress has not explicitly stated what kind of deference it prefers courts to give. 184 Peter L. Strauss, Courts or Tribunals? Federal Courts and the Common Law, 53 Ala. L. Rev. 891, 893 (2002). His preferred approach would be to allow courts to allow for “case-by-case development of an imperfect statutory framework to resolve a difficult issue of federal administrative law.” 185Id. Professors Gluck and Bressman also expressed a desire for judges to maintain an “umpireal” role for courts in deciding when to defer. 186 Gluck & Bressman, Part II, supra note 5, at 791. Professor Walker proposed an inquiry into whether “the ambiguity seems like a deliberate delegation by the collective Congress, or whether it seems more like the result of administrative collusion during the legislative process.” 187 Walker, supra note 5, at 1421. Despite all of these judicial and scholarly attempts to figure out Chevron’s domain, we currently have only a series of unweighted factors that are difficult to predict or apply in any particular case, a series of assumptions that often do not reflect reality, and vague descriptions of how to approach a search for congressional intent. 188 Gluck et al., supra note 134, at 1847; see also David J. Barron & Elena Kagan, Chevron’s Nondelegation Doctrine, 2001 Sup. Ct. Rev. 201, 203 (2001) (“Given the difficulty of determining actual congressional intent, some version of constructive—or perhaps more frankly said, fictional—intent must operate in judicial efforts to delineate the scope of Chevron.”). It is worth noting that many of the factors used in the various cases cited above did receive some support from respondents in Gluck and Bressman’s surveys. See Gluck & Bressman, Part I, supra note 1, at 999–1004. Unfortunately, Gluck and Bressman did not ask about agency legislative history, so we do not know how congressional staff feel about agency legislative communications, although if the author’s interviews with agency staffers are correct, those communications are an integral part of the legislative process through which congressional intent is formed.

If judges wish to approach Chevron in a more contextual way despite the institutional issues discussed earlier in this Part, as many judges undoubtedly will, this Article provides new avenues that would allow them to do so in ways that better reflect the realities of agency–Congress relationships and Congress’s own chosen legislative process. In fact, this Article shows that pre-Chevron courts used agency legislative history exactly this way, which could be helpful to judges looking to return to a pre-Chevron mode of judicial deference. Before Chevron, courts’ decisions of whether to defer to an agency’s interpretation were contextual, and agency legislative history often played a role in determining this context. 189See supra Part II. Instead of looking at gaps or ambiguities as an automatic sign of deference, 190 Chevron U.S.A. Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 865–66 (1984); see also Cuozzo Speed Techs., LLC v. Lee, 136 S. Ct. 2131, 2142 (2016) (“But where a statute leaves a ‘gap’ or is ‘ambigu[ous],’ we typically interpret it as granting the agency leeway to enact rules that are reasonable in light of the text, nature, and purpose of the statute.”). these pre-Chevron courts sometimes used agency legislative history to determine the relationship and interactions between an agency and Congress at the time of enactment, which helped inform the court about whether an ambiguity was a result of Congress’s intent to defer. 191 For example, where there was conflict between the agency and Congress during the legislative process, a court could apply the Skidmore standard, which means it would independently judge the quality of the agency’s interpretation where there is conflict between the agency and Congress. On the other hand, if it appeared that the agency and Congress were on the same page with respect to the legislation, then a court could grant Chevron-type deference. For example, as discussed above, courts looked at the fact that an agency proposed the legislation or was involved in its drafting as a strong reason to defer to agency interpretations. 192See supra Sections II.B.–II.C. Courts also considered the fact that an agency proposal was rejected as a signal not to defer, 193See supra Section II.B. and they also looked at things like how close in time the agency interpretation was to enactment when deciding whether to uphold an agency’s interpretation. 194See cases cited supra note 60. These types of factors could be instructive for the current Supreme Court, which focuses on textual clues to determine congressional intent to delegate, mostly ignoring the clues of congressional intent that often lie in the exchanges between agencies and Congress.

Of course, agency legislative history can only serve as a useful tool of judicial interpretation if it is available. One reason why it has not been used more is that it is often not publicly available, as discussed above. 195See infra Section II.E. If Congress began to include more agency legislative history in its own congressional legislative history, as this Article argues it should, it would become even easier for courts to engage in a more contextual deference analysis. This would give courts a more complete legislative picture, because the reality of the complex modern legislative process is that Congress may have little choice but to rely on agency expertise and capacity when forming its legislative expectations and intentions. For judges looking to determine whether to defer on a case-by-case basis, agency legislative history might be the best way to determine whether Congress intended to delegate “authority to the agency to elucidate a specific provision of the statute,” as the Court stated in Chevron196 Chevron U.S.A. Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 843–44 (1984).

While judges might hesitate to dig into legislative history to decide whether to defer to an agency interpretation because of the types of concerns described earlier in this Part, arguments could be made that agency legislative history is in some ways less problematic than other congressional legislative history. The types of agency legislative history discussed in this Article are communicated in a formal way to the relevant committees within Congress. There are two levels of scrutiny for all agency legislative history, which has to be cleared through an intra-agency process, 197 As discussed in Section II.A., agencies also engage in many informal communications with Congress on the staff level, and it is possible that these types of communications could be more likely to be “cheap talk.” That is one reason why this Article focuses only on formal agency legislative history. These formal communications require a substantial agency drafting process and formal clearance that is likely to weed out empty rhetoric about statutory language. while congressional legislative history is, at times, subject to none. If an agency legislative staffer wanted to create deceptive agency legislative history, it would require an incredible amount of coordination within a large bureaucratic organization. 198 This is different from the concern often raised by textualists about a sneaky congressional staffer being able to modify legislative history all on their own. As Justice Scalia argued, “[a]s anyone familiar with modern-day drafting of congressional committee reports is well aware, the references to the cases were inserted, at best by a committee staff member on his or her own initiative, and at worst by a committee staff member at the suggestion of a lawyer-lobbyist; and the purpose of those references was not primarily to inform the Members of Congress what the bill meant . . . but rather to influence judicial construction.” Blanchard v. Bergeron, 489 U.S. 87, 98–99 (1989) (Scalia, J., concurring in part and concurring in the judgment); see also Frank H. Easterbrook, Text, History, and Structure in Statutory Interpretation, 17 Harv. J.L. & Pub. Pol’y 61, 61 (1994) (“These clues [in legislative history] are slanted, drafted by the staff and perhaps by private interest groups.”); Gluck & Bressman, Part I, supra note 1, at 973 (showing that legislative drafters use legislative history to get “something we couldn’t get in the statute in order to make key stakeholders happy”); Antonin Scalia & John F. Manning, A Dialogue on Statutory and Constitutional Interpretation, 80 Geo. Wash. L. Rev. 1610, 1612 (2012); Starr, supra note 172 (“Lobbyists maneuver to get their clients’ opinions into the mass of legislative materials . . . .”). To generate misleading agency legislative history that ends up in Congress’s legislative history, an agency staffer would not only have to convince others within the agency to go along, but would also have to sneak the misleading statement past the relevant congressional committee to bias the legislative record. 199 Those working on legislation inside an agency and inside Congress would have to coordinate with regulators to generate the false legislative history to ensure that the regulators knew what to do and then they would have to work with litigators to make sure they knew to point out the false agency legislative history to a court if the agency’s interpretation was challenged. All of this would have to happen over a period of many years because of the time it takes to enact legislation and for legal challenges to arise. In the process of their attempted deception, they run the risk of being discovered, which could cause Congress to amend the statutory language in ways that could be even worse for the agency. The agency and its staff would also lose their credibility for future interactions with Congress. 200 Agencies are repeat actors full of career employees who rely on positive relationships with Congress to achieve their agency’s goals. Being uncovered as a dishonest actor would reduce the agency’s influence and potentially subject staffers and the agency to increased scrutiny from Congress. Commentators have made similar arguments about those within Congress who might attempt to use legislative history in devious ways. McNollgast, Legislative Intent: The Use of Positive Political Theory in Statutory Interpretation, 57 L. & Contemp. Probs. 3, 26 (1994) (“In practice, political actors have two routes to enforce truthfulness. First, members who prevaricate can be and occasionally are removed from gatekeeping positions, such as party leadership or committee chairs. Second, because Congress passes a very large number of legislative provisions each year, the same member is likely to be in a position of delegated authority on many occasions. To succeed in accomplishing numerous legislative objectives over a lifetime in politics, a legislator will find it valuable to develop a reputation for not taking strategic advantage when acting as an agent for other members.”); see also Katzmann, supra note 63, at 654 (“The system works because committee members and their staffs will lose influence with their colleagues as to future bills if they do not accurately represent the bills under consideration within their jurisdiction.”). There are good reasons to think that agency legislative history is a relatively accurate record of agency negotiations and deals with Congress.

To illustrate this argument through a hypothetical example not dissimilar from actual cases discussed above, suppose that an agency submitted proposed legislation to Congress along with a section-by-section analysis of the legislation, and Congress enacted the proposed legislation virtually unchanged and incorporated the agency’s section-by-section analysis into a committee report. Suppose a new President from a different political party is elected and the agency then tries to interpret the statute in ways that go against what the agency said in the section-by-section analysis because there is some plausible ambiguity in the statute. Should a court defer to the agency interpretation under Chevron because of the ambiguity? Rather than automatically deferring, a court could check this political creep by looking to the agency legislative history that gives insight into the deal Congress thought it was agreeing to at the time the bill was passed. A textualist might still object to the use of legislative history in this case on constitutional or institutional grounds, but it is clearly a harder question in this circumstance, and the agency legislative history could cause some judges to think twice about granting broad deference.

A recent example where the Court failed to recognize that the agency was the initial drafter of the enacted text helps to illustrate how a court could use this type of legislative history on a contextual basis. In Rumsfeld v. Forum for Academic and Institutional Rights, Inc.201 547 U.S. 47, 51 (2006). the Supreme Court considered whether law schools are allowed to prohibit military representatives from recruiting on campus because of the military’s then-existing “don’t ask, don’t tell” policy. 202Id. at 52. The case turned mostly on a question of constitutional law, but the Court also considered a question of statutory interpretation in response to a number of law professors who argued in an amicus brief that universities are permitted to prohibit military recruiters under the statute. 203Id. at 55–56 (“Certain law professors participating as amici, however, argue that the Government and FAIR misinterpret the statute.”). Neither the Court nor the amici looked at the agency legislative history and, therefore, did not know that the language in question was actually proposed by the DoD specifically to require schools to allow military recruiters access to campus on terms “at least equal in quality and scope to the access provided other potential employers,” and that Congress adopted the DoD’s proposed language almost verbatim. 204Id. at 54. For the agency’s proposal, see Dep’t of Def., National Defense Authorization Act for Fiscal Year 2005 (2004), http://ogc.osd.mil/olc/docs/March12-Text.pdf. The language at issue was introduced in Congress the same day it was sent from DoD to Congress. 205Compare Dep’t of Def., supra note 204 (showing date legislation was sent to Congress), with H.R. Rep. No. 108-443, pt. 1, at 4 (2004) (showing date legislation was proposed in Congress). DoD’s draft bill was accompanied by a letter to Congress, which was ultimately included as part of the House’s committee report, explaining the purpose of the legislation. 206H.R. Rep. No. 108-443, pt. 1, at 7 (2004). It seems clear from looking at the agency legislative history that Congress and the DoD had an understanding of what the bill was intended to do, and with this context it seems impossible that Congress intended the bill to allow schools to ban military recruiters, as the amicus brief argued. The Court’s position in favor of the DoD would have been stronger if it had considered the agency legislative history, which provided important context as to why Congress was enacting the legislation.

Agency legislative history may not always speak directly to whether Congress intended to defer in a particular case, but it can still be informative to courts even when it is not directly on point. It can serve as a strong indicator of whether the relationship between an agency and Congress is collaborative or combative with respect to certain parts of that legislation, which is relevant to determining Congress’s intent. When Congress rejects agency proposals, it probably does not view the agency as a partner in creating the legislation, and courts could therefore be more skeptical of agency interpretations of that legislation. For example, if an agency wanted Congress to create legislation that was drafted in a certain way and Congress instead decided to draft it a different way, a court could view agency actions with respect to that legislation more circumspectly to ensure that the agency is not trying to do something administratively that it could not get legislatively. This is similar to the way some courts already treat rejected proposals in statutory interpretation generally. 207See Hamdan v. Rumsfeld, 548 U.S. 557, 579 (2006); FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120, 125–26 (2000); Patterson v. McLean Credit Union, 491 U.S. 164, 176–78 (1989); INS v. Cardoza-Fonseca, 480 U.S. 421, 442–43 (1987); United States v. Riverside Bayview Homes, 474 U.S. 121, 137 (1985); United States v. Yermian, 468 U.S. 63, 72 (1984) (relying on congressional adoption of broadening language of false swearing statute after earlier bill was vetoed by President); Runyon v. McCrary, 427 U.S. 160, 174–75 (1976) (relying on the fact that the Senate had rejected attempts to override an interpretation as support for that interpretation); Blau v. Lehman, 368 U.S. 403, 411–12 (1962); Mont. Wilderness Ass’n v. U.S. Forest Serv., 314 F.3d 1146, 1150 (9th Cir. 2003). For a deeper discussion of the rejected proposal rule, see Eskridge, Jr. et al., supra note 3, at 831. Not all judges find the rejected proposal canon persuasive because legislative inaction can occur for a number of reasons that are unclear to courts. See, e.g., Rapanos v. United States, 547 U.S. 715, 750 (2006) (“Congress takes no governmental action except by legislation. What the dissent refers to as ‘Congress’[s] deliberate acquiescence’ should more appropriately be called Congress’s failure to express any opinion.”); Solid Waste Agency of N. Cook Cty. v. U.S. Army Corps of Eng’rs, 531 U.S. 159, 170 (2001) (“A bill can be proposed for any number of reasons, and it can be rejected for just as many others. The relationship between the actions and inactions of the 95th Congress and the intent of the 92d Congress in passing Section 404(a) is also considerably attenuated.”); Cent. Bank of Denver, N.A. v. First Interstate Bank of Denver, N.A., 511 U.S. 164, 187 (1994) (“[F]ailed legislative proposals are ‘a particularly dangerous ground on which to rest an interpretation of a prior statute.’”); Pattern Makers’ League of N. Am. v. NLRB, 473 U.S. 95, 110–12 (1985) (finding the legislative history too ambiguous to influence the Court’s decision). Some scholars have also been critical of the rule. See, e.g., William N. Eskridge, Jr., Interpreting Legislative Inaction, 87 Mich. L. Rev. 67, 84–95 (1988) (collecting cases using and declining to use the rejected proposal rule). These challenges could also apply to rejected agency legislative proposals, because agencies make many legislative proposals and it is hard to know why any particular proposal is rejected. Not every rejected proposal is significant, but many are, especially when Congress proceeds to pass legislation on the same topic. Similarly, when Congress disregards an agency’s views and passes a bill despite agency opposition, some judges might find it incongruous to then give strong deference to the agency’s interpretation. An assumption of an intent to defer in this circumstance leaves the agency abundant interpretive space even though the risk of agency subversion is high because the agency gets a second chance to “win” in implementation. 208 On the other hand, recall the Cuozzo case, described above, where the Court relied on Chevron to defer to the USPTO. See supra Section II.B. In that case, Chevron’s presumption made sense in light of the agency legislative history. There was ample evidence of a cooperative relationship between the USPTO and Congress and an intent to delegate to the USPTO. Although the agency legislative history would not have changed the Court’s decision, it would have strengthened it by situating it within the empirical reality of how the bill came to be rather than an assumption that may or may not have been true. See supra Section II.B. Courts may therefore want to be on guard to ensure that the agency is acting in accordance with Congress’s preferences rather than the agency’s conflicting preferences. On the other hand, where an agency supported a bill, or opposed a bill and Congress made changes to appease the agency, a court could view that as a reason to grant greater deference to agency interpretations, as courts have in the past. 209See supra Part III. Agency legislative history may not always allow courts to determine actual congressional intent with certainty, but it may allow courts to craft a more accurate constructive congressional intent. 210 One concrete way in which courts have attempted to approach Chevron on a case-by-case basis is through the application of the major questions doctrine. This doctrine presumes that Congress does not intend to delegate issues of major importance to agencies without explicitly doing so, and instead keeps the interpretation of those provisions to the court. See Kevin O. Leske, Major Questions About the “Major Questions” Doctrine, 5 Mich. J. Envtl. & Admin. L. 479, 480 (2016). This doctrine was an important part of the majority’s decision in King v. Burwell, which stated that it was inconceivable that Congress would have left a question of such “deep ‘economic and political significance’” to an agency without expressly doing so. 135 S. Ct. 2480, 2489 (2015). While the wisdom of allowing courts to determine whether an issue is “major” enough to do away with deference is certainly debatable, if courts are to engage in this type of inquiry then they would be better served by considering agency legislative history to determine whether such an issue had actually been contemplated. Perhaps the agency legislative history contains a discussion of the major issue or perhaps the agency pointed out the major issue and Congress still chose not to amend the legislation. Before a court decides that an issue is too major for Congress to have left it to an agency, it could check the agency legislative history to confirm that Congress in fact had not considered the issue, in much the same way courts check congressional legislative history before applying the absurdity doctrine to make sure that the absurdity was not intentional.

B. Implications for Interpretation

1. Interpreting Statutes in Light of Agency Legislative History

Whether a judge prefers broad deference or greater judicial scrutiny, they still must engage with the text of a statute to determine its possible meanings. 211 The Court has accepted evidence of statutory purpose, “including those revealed in part by legislative and regulatory history,” to determine whether a statute is ambiguous at Chevron step one. See City of Arlington v. FCC, 569 U.S. 290, 310 (2013); see also Brown & Williamson Tobacco Corp., 529 U.S. at 143–47. Agency legislative history often provides useful evidence of the understanding reached between an agency and Congress, especially when it has been incorporated into congressional legislative history. Even a judge who prefers broad deference could check the agency legislative history to make sure an agency is not acting in ways that go against the representations the agency made to Congress. If the agency legislative history contains relevant explanations from agencies to Congress or descriptions of how a provision would be carried out, and Congress included that in a committee report or elsewhere in the Congressional Record, it would make sense for a court to make an analysis of those communications an important part of its interpretation process. Courts could engage these materials at step one of Chevron when deciding whether Congress intended a single meaning for the text, or at step two of Chevron when deciding whether the agency’s interpretation was “reasonable.” 212Cf. Strauss, Chevron Space” and “Skidmore Weight”, supra note 26, at 1163 (arguing that determinations of whether an agency interpretation is “reasonable” might consider whether the agency’s interpretation is consistent with statutory purposes). Courts have disagreed about whether to consult legislative history at step one or step two of Chevron. Compare Halverson v. Slater, 129 F.3d 180, 184 (D.C. Cir. 1997) (holding that “the lower court erred by failing to ‘exhaust the traditional tools of statutory construction,” including legislative history, at Chevron step one), with Cohen v. JP Morgan Chase & Co., 498 F.3d 111, 122 (2d Cir. 2007) (“This court has generally been reluctant to employ legislative history at step one of Chevron analysis, mindful that the ‘interpretive clues’ to be found in such history will rarely speak with sufficient clarity to permit us to conclude ‘beyond reasonable doubt’ that Congress has directly spoken to the precise question at issue.”). The distinction is unlikely to matter in whether a court upholds an agency interpretation. Or if a court determines that Chevron should not apply, it could include agency legislative history as part of the many factors it considers under a Skidmore analysis. 213 Courts would still need to distinguish reliable from unreliable agency legislative history in much the same way they already do for congressional legislative history.

Agency legislative history has ramifications for the ongoing debate about the use of legislative history generally. One previously unknown consequence of judges’ increased skepticism towards legislative history is that they might be missing not only congressional legislative history but also the agency legislative history incorporated into congressional documents. Members of Congress and their staff know that agencies read legislative history, so Congress may use its own legislative history to memorialize an understanding reached with an agency and to hold the agency to the bargain it made during the legislative process, and courts who ignore legislative history need to be aware that they might also be ignoring the best evidence of agency–Congress legislative deals. 214 Congressional staff confirm that legislative history is often written with agencies as the intended audience. See Gluck & Bressman, Part II, supra note 5, at 789–90. It would seem that even a textualist judge would view agency legislative history as distinct from other legislative history. Textualists are concerned about individual members or staffers using Congress’s legislative history to do an “end run” around the constitutionally prescribed process, 215 Manning, supra note 34, at 1534. but by ignoring agency legislative history and deferring to agency interpretations, courts are potentially allowing agencies to do an “end run” around Congress.

Courts have already shown how agency legislative history could be used in interpretation. Recall United States v. Vogel and Piper v. Chris-Craft, two pre-Chevron Supreme Court cases discussed above. 216See supra Sections II.E.1.–II.E.2. In both of these cases, the Court relied on agency legislative history to overturn an agency interpretation of an ambiguous statute because the agency legislative history showed that the agency’s interpretation went against contemporaneous agency–Congress communications, even though the statute itself was ambiguous. 217See supra Sections II.E.1.–II.E.2. And in many other cases discussed above, the agency legislative history was helpful in showing that the agency’s interpretation was consistent with its statements to Congress, and the Court therefore upheld the interpretations. 218See supra Part II. These types of inquiries may require courts to conduct a more searching and purposivist judicial inquiry in many circumstances, which is subject to all of the caveats discussed above about courts’ competency to recreate complicated legislative deals. In many ways this is not that different from what courts often already do, since courts have frequently relied on legislative history in Chevron cases, 219See William N. Eskridge, Jr. & Lauren E. Baer, The Continuum of Deference: Supreme Court Treatment of Agency Statutory Interpretations from Chevron to Hamdan, 96 Geo. L.J. 1083, 1136 (2008) (collecting statistics on the use of legislative history in Chevron cases). but this Article’s proposals provide a different perspective on how to view those materials. It is up to a court to decide how much weight to give to agency legislative history in any particular case, and this Article provides a place from where they can start. 220 The conventional wisdom is that committee and conference reports are by far the most reliable legislative history, followed by on-the-record statements by the members of Congress who sponsored or supported the bill. Eskridge, Jr. et al., supra note 3, at 631; Eskridge, Jr., supra note 4, at 221–22; Robert A. Katzmann, Judging Statutes 35–54 (2014); Manning & Stephenson, supra note 3, at 136 (“The conventional wisdom has been that the most reliable form of legislative history consists of the reports prepared by the House and Senate committees, which accompany bills favorably reported to the chamber, and the conference committee reports which accompany the reconciled version of the House and Senate bills.”); Costello, supra note 1, at 43–50 (discussing the importance of conference committee reports). For example, Justice Jackson objected to the use of legislative history generally but made an exception for committee reports. See Schwegmann Bros. v. Calvert Distillers Corp., 341 U.S. 384, 395 (1951) (Jackson, J., concurring). As Justice Frankfurter remarked, “[w]hatever we may think about the loose use of legislative history, it has never been questioned that reports of committees and utterances of those in charge of legislation constitute authoritative exposition of the meaning of legislation.” Orloff v. Willoughby, 345 U.S. 83, 98 (1953) (Frankfurter, J., dissenting). Quantitative analyses confirm that committee reports are relied on far more than other kinds of legislative history. See, e.g., James J. Brudney & Corey Ditslear, The Warp and Woof of Statutory Interpretation: Comparing Supreme Court Approaches in Tax Law and Workplace Law, 58 Duke L.J. 1231, 1262 (2009) (discussing the prevalence of committee reports used in tax law and workplace law cases); Jorge L. Carro & Andrew R. Brann, The U.S. Supreme Court and the Use of Legislative Histories: A Statistical Analysis, 22 Jurimetrics J. 294, 304 (1982) (reporting that, over the forty year period analyzed by the authors, over 60% of the Supreme Court’s citations to legislative history were references to committee reports).

2. Agency Legislative History and Empirical Realities of the Legislative Process

Agency legislative history also has bearing on many recent debates about the role empirical realities of the legislative process should play in statutory interpretation. It turns out that recent empirical work likely undersells the complexity of the legislative process. For example, both the author’s own research and Professors Gluck and Bressman’s studies revealed the important role Congress’s non-partisan, professional statutory drafters in the House and Senate Offices of Legislative Counsel play in drafting statutory text, but not legislative history. 221See Shobe, Intertemporal Statutory Interpretation, supra note 5, at 861–65; Abbe R. Gluck, Congress, Statutory Interpretation, and the Failure of Formalism: The CBO Canon and Other Ways that Courts Can Improve on What They Are Already Trying to Do, 84 U. Chi. L. Rev. 177, 208–09 (2017) [hereinafter Gluck, Congress] (describing the Gluck and Bressman study and its findings that congressional staffers “make the policy and sketch out statutory contours, often in the form of policy ‘bullet points.’ The nonpartisan Legislative Counsels then take over and turn the ‘asks’ into statute-ese”). Under this account, congressional staff develop a basic legislative plan and then turn it over to legislative counsel to draft the statutory particulars, with the congressional staff drafting accompanying legislative history explaining the bill. While this is certainly an accurate reflection of how the process works in some cases, agency legislative history seems to indicate that the process is often more complicated, with agencies sometimes drafting statutory language and legislative history. The simplified account of the legislative process likely reflects reality only for short, simple, single-subject bills, which have become increasingly rare in recent years. For more complex bills, agencies are often involved in creating the legislative plan, drafting the actual statutory text, and making significant revisions to statutory text. 222See supra Part II. Agency legislative history also seems to indicate that legislative counsel might not be substantively knowledgeable enough to draft well in all circumstances, and they must rely on agency input in many cases to get the legislative language to work. This is unsurprising given the small number of relatively generalist legislative counsel as compared to the large number of agency experts involved in the legislative process.

Not only are agencies closely involved in drafting statutes, it also turns out that agencies often draft legislative history, which calls into question the clean division of labor between congressional staff and legislative counsel. For example, Professor Gluck notes that the section-by-section summary of a statute is generally considered to be one of the most important pieces of legislative history and claims that this section-by-section analysis is drafted by committee staff, not legislative counsel. 223See Gluck, Congress, supra note 221, at 209 (“The section-by-section is drafted by committee staff (the policy staff, not the Legislative Counsels) . . . .”). The findings here complicate this somewhat. It appears that these section-by-section analyses are at least sometimes drafted by agency staff rather than congressional staff, and that congressional staff merely import the agency’s words into the committee report. Committee reports, like statutory text, appear to be the result of a long legislative process that often indirectly and directly incorporates materials from agencies, and probably from other outside groups too. 224 Blanchard v. Bergeron, 489 U.S. 87, 98–99 (1989) (Scalia, J., concurring in part and concurring in the judgment) (“As anyone familiar with modern-day drafting of congressional committee reports is well aware, the references to the cases were inserted, at best by a committee staff member on his or her own initiative, and at worst by a committee staff member at the suggestion of a lawyer-lobbyist; and the purpose of those references was not primarily to inform the Members of Congress what the bill meant . . . but rather to influence judicial construction.”). That Congress incorporates agency legislative history into its own legislative history is likely meant to guide the agency’s implementation, which provides support for the contention that agencies are often Congress’s intended audience for legislative history. 225 Strauss, When the Judge Is Not the Primary, supra note 156, at 329–35. Professors Gluck and Bressman’s respondents almost uniformly agreed with the idea that one purpose of legislative history is to shape the way agencies interpret statutory ambiguities. Gluck & Bressman, Part II, supra note 5, at 768. Congress may adopt agency legislative history as a tool of oversight that is contemporaneous with the enactment of the statute, unlike other means of post-enactment oversight like congressional hearings. For a discussion of tools of congressional oversight, see Mathew D. McCubbins et al., Structure and Process, Politics and Policy: Administrative Arrangements and the Political Control of Agencies, 75 Va. L. Rev. 431, 432–45 (1989). We cannot know why and how often this occurs without further investigation of the process, and this certainly merits further investigation. For purposes of this Article it is sufficient to note that this phenomenon exists, and that it complicates many of the apparently oversimplified understandings we have about the origins of legislative history. Much like with statutory text, the generation of legislative history is the result of a messy and ununiform process. If scholars and judges want to rely on source arguments to support or oppose the use of legislative history, agency legislative history suggests that more work must be done to understand how legislative history is generated.

Conclusion

Scholars and judges need a more contextual understanding of how statutes are created, with a greater emphasis on the role of agencies in creating them. Agency legislative history provides important context and has broad implications for statutory interpretation and agency delegation. The typology of agency legislative history presented here gives courts and scholars a roadmap for how they can more accurately account for legislative realities in their theories and the practice of statutory interpretation and agency delegation.

 

Footnotes

*Associate Professor of Law, Brigham Young University. For helpful discussions and comments on earlier drafts of this Article, I am grateful to Paul Stancil, Aaron Nielson, Chris Walker, Bijal Shah, Seth Davis, and participants in the Rocky Mountain Jr. Scholars Forum and the BYU law faculty workshop. For excellent research assistance I am grateful to Jessica Payton, Stephanie Nielson, Jeff Cottle, Ryan Ricks, and Jake Barney.

1See, e.g., Abbe R. Gluck & Lisa Schultz Bressman, Statutory Interpretation from the Inside—An Empirical Study of Congressional Drafting, Delegation, and the Canons: Part I, 65 Stan. L. Rev. 901, 964 (2013) [hereinafter Gluck & Bressman, Part I] (“The other primary interpretive source that courts consider—and the one whose use is most hotly contested—is legislative history.”); Stephen Breyer, On the Uses of Legislative History in Interpreting Statutes, 65 S. Cal. L. Rev. 845, 845–46 (1992); George A. Costello, Average Voting Members and Other “Benign Fictions”: The Relative Reliability of Committee Reports, Floor Debates, and Other Sources of Legislative History, 1990 Duke L.J. 39, 39–42 (1990); William N. Eskridge, Jr., The New Textualism, 37 UCLA L. Rev. 621, 623 (1990) (“The new textualism posits that once the Court has ascertained a statute’s plain meaning, consideration of legislative history becomes irrelevant.”); William N. Eskridge, Jr., Textualism, the Unknown Ideal?, 96 Mich. L. Rev. 1509, 1512–13 (1998); Abbe R. Gluck, Intersystemic Statutory Interpretation: Methodology as “Law” and the Erie Doctrine, 120 Yale L.J. 1898, 1909–10 n.22 (2011); John F. Manning, Textualism as a Nondelegation Doctrine, 97 Colum. L. Rev. 673, 687–88 (1997); Jane S. Schacter, The Confounding Common Law Originalism in Recent Supreme Court Statutory Interpretation: Implications for the Legislative History Debate and Beyond, 51 Stan. L. Rev. 1, 3, 6 (1998).

2Manning & Stephenson, infra note 3, at 127; see Gluck, supra note 1. For a few recent Supreme Court examples where legislative history was contested, see Mohamad v. Palestinian Auth., 566 U.S. 449, 459 (2012) (arguing that the statutory text is clear and legislative history therefore unnecessary); id. at 461 (Breyer, J., concurring) (disagreeing that the text is clear and arguing in favor of considering the legislative history); Coleman v. Court of Appeals of Md., 566 U.S. 30, 44–45 (2012) (Scalia, J., concurring in the judgment) (arguing against the plurality’s use of legislative history and in favor of considering the text alone); Reynolds v. United States, 565 U.S. 432, 448 n.* (2012) (Scalia, J., dissenting) (arguing that the majority’s use of legislative history is “superfluous”); Gonzalez v. Thaler, 565 U.S. 134, 165–66 (2012) (Scalia, J., dissenting) (providing a critique of the use of legislative history); DePierre v. United States, 564 U.S. 70, 89 (2011) (Scalia, J., concurring in part and concurring in the judgment) (arguing that the use of legislative history is “not harmless”); Bruesewitz v. Wyeth LLC, 562 U.S. 223, 240 (2011) (acknowledging that the use of legislative history is not considered legitimate by all judges).

3 These sources of legislative history are covered exhaustively in various leading textbooks on legislation. See, e.g., William N. Eskridge, Jr. et al., Cases and Materials on Legislation and Regulation: Statutes and the Creation of Public Policy (5th ed. 2014); John F. Manning & Matthew C. Stephenson, Legislation and Regulation: Cases and Materials 142 (2d ed. 2013). For interesting cases on the use of legislative inaction as a source of legislative history, see FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120, 146–47 (2000); Bob Jones Univ. v. United States, 461 U.S. 574, 600–01, 607 (1983).

4 For a comprehensive discussion of the different types of legislative history, see Manning & Stephenson, supra note 3, at 136–45. The current hierarchy of legislative history sources does not always exclude noncongressional sources, but “nonlegislator statements” are rarely referenced and generally thought of as being one of the least authoritative forms of legislative history. William N. Eskridge, Jr., Dynamic Statutory Interpretation 222 (1994).

5 Various recent empirical studies have discussed the role of agencies in the legislative process generally. See Lisa Schultz Bressman & Abbe R. Gluck, Statutory Interpretation from the Inside—An Empirical Study of Congressional Drafting, Delegation, and the Canons: Part II, 66 Stan. L. Rev. 725, 758 (2014) [hereinafter Gluck & Bressman, Part II]; Jarrod Shobe, Agencies as Legislators: An Empirical Study of the Role of Agencies in the Legislative Process, 85 Geo. Wash. L. Rev. 451, 454 (2017) [hereinafter Shobe, Agencies as Legislators]; Jarrod Shobe, Intertemporal Statutory Interpretation and the Evolution of Legislative Drafting, 114 Colum. L. Rev. 807, 863–64 (2014) [hereinafter Shobe, Intertemporal Statutory Interpretation]; Ganesh Sitaraman, The Origins of Legislation, 91 Notre Dame L. Rev. 79, 132 (2015); Christopher J. Walker, Legislating in the Shadows, 165 U. Penn. L. Rev. 1377, 1378–79 (2017).

6 This Article focuses on formal agency communications, which generally occur in written form or in formal settings, because these are most likely to be accessible to courts and the public and are more likely to reflect the position of an agency as a whole rather than an individual agency staffer. See infra Section II.A.

7 Professors Farber & O’Connell made similar arguments for modern regulation generally. See Daniel A. Farber & Anne Joseph O’Connell, The Lost World of Administrative Law, 92 Tex. L. Rev. 1137, 1140 (2014).

8See infra Part II.

9See infra Part II.

10Vermeule, infra note 12, at 206; William N. Eskridge, Jr., Expanding Chevron’s Domain: A Comparative Institutional Analysis of the Relative Competence of Courts and Agencies to Interpret Statutes, 2013 Wis. L. Rev. 411, 427 (2013); Jerry L. Mashaw, Prodelegation: Why Administrators Should Make Political Decisions, 1 J.L. Econ. & Org. 81, 91–99 (1985); David B. Spence & Frank Cross, A Public Choice Case for the Administrative State, 89 Geo. L.J. 97, 134–41 (2000).

11Supra note 10.

12 In this way this Article builds on scholarship by Professors Mashaw and Strauss that argues that agencies should approach interpretation different from courts because of their unique position and relationship with Congress. See, e.g., Adrian Vermeule, Judging Under Uncertainty: An Institutional Theory of Legal Interpretation 205–08 (2006); Jerry L. Mashaw, Norms, Practices, and the Paradox of Deference: A Preliminary Inquiry into Agency Statutory Interpretation, 57 Admin. L. Rev. 501, 504 (2005) (“There are persuasive grounds for believing that legitimate techniques and standards for agency statutory interpretation diverge sharply from the legitimate techniques and standards for judicial statutory interpretation.”); Cass R. Sunstein & Adrian Vermeule, Interpretation and Institutions, 101 Mich. L. Rev. 885, 928 (2003).

13 Richard H. Pildes, Institutional Formalism and Realism in Constitutional and Public Law, 2013 Sup. Ct. Rev. 1, 2 (2013).

14 Sitaraman, supra note 5, at 86; Walker, supra note 5, at 1411 (“If the agency is indeed a partner with Congress in the legislative drafting, Justice Scalia’s concern about an agency legislating and executing the law should apply with some force to legislative drafting. The executive and legislative functions are, in essence, combined via agency legislating in the shadows.”).

15 Auer v. Robbins, 519 U.S. 452, 461–63 (1997). But see Talk Am., Inc. v. Mich. Bell Tel. Co., 564 U.S. 50, 69 (2011) (Scalia, J., concurring) (criticizing Auer because “deferring to an agency’s interpretation of its own rule encourages the agency to enact vague rules which give it the power, in future adjudications, to do what it pleases”).

16 A number of Supreme Court Justices have recently expressed concern with the expansion of the administrative state. City of Arlington v. FCC, 569 U.S. 290, 315 (2013) (Roberts, C. J., dissenting) (“[T]he danger posed by the growing power of the administrative state cannot be dismissed.”); see also Dep’t of Transp. v. Ass’n of Am. R.Rs., 135 S. Ct. 1225, 1254 (2015) (Thomas, J., concurring in the judgment) (describing the administrative state as a system that “concentrates the power to make laws and the power to enforce them in the hands of a vast and unaccountable administrative apparatus”). For an excellent discussion of the rise of the administrative state, see Gary Lawson, The Rise and Rise of the Administrative State, 107 Harv. L. Rev. 1231, 1233 (1994).

17 This is something the Supreme Court has explicitly pursued. See United States v. Mead Corp., 533 U.S. 218, 236 (2001) (“Justice Scalia’s first priority over the years has been to limit and simplify. The Court’s choice has been to tailor deference to variety.”).

18 Manning, infra note 34, at 1534.

19See infra Section III.B.

20See infra Section III.B.

21William N. Eskridge, Jr., Interpreting Law: A Primer on How to Read Statutes and the Constitution 191 (2016) [hereinafter Eskridge, Jr., Interpreting Law].

22 These sources of legislative history are covered exhaustively in various leading textbooks on legislation. See Eskridge, Jr., supra note 4; supra note 3 and accompanying text.

23Eskridge, Jr. et al., supra note 3, at 632.

24 “The court does not rely on statements by nonlegislative officials as the most probative evidence of the meaning of statutory language. There are very few state cases and those that exist reject the use of such evidence of statutory meaning.” Id. at 828; e.g., Hayes v. Cont’l Ins. Co., 872 P.2d 668, 673–74 (Ariz. 1994).

25See supra note 24.

26 Peter L. Strauss, Essay, “Deference” Is Too Confusing—Let’s Call Them “Chevron Space” and “Skidmore Weight”, 112 Colum. L. Rev. 1143, 1146 (2012) [hereinafter Strauss, Chevron Space” and “Skidmore Weight”] (“The agency may have helped to draft the statutory language, and was likely present and attentive throughout its legislative consideration. Its views about statutory meaning may have been shaped in the immediate wake of enactment, under the enacting Congress’s watchful eye.”); see Matthew R. Christiansen & William N. Eskridge, Jr., Congressional Overrides of Supreme Court Statutory Interpretation Decisions, 1967–2011, 92 Tex. L. Rev. 1317, 1323, 1377 (2014) (providing evidence that agencies work with congressional staff to draft override legislation, and lobby for such legislation, based on an examination of committee hearings and reports).

27 Walker, supra note 5. Professor Walker’s article was part of a larger project commissioned by the Administrative Conference of the United States. Christopher J. Walker, Federal Agencies in the Legislative Process: Technical Assistance in Statutory Drafting (2015).

28 Sitaraman, supra note 5, at 124–32.

29See Shobe, Agencies as Legislators, supra note 5.

30See Brigham Daniels, Agency as Principal, 48 Ga. L. Rev. 335, 404 (2014) (“Sometimes Congress asks agencies to draft language, and sometimes agencies do so without being asked. It is just the way the game is played, and those with much experience in Washington openly acknowledge this.”); Gluck & Bressman, Part I, supra 1, at 1021 (“There are likely external networks of these noncongressional drafters of federal legislation, with deep resources of institutional and legal knowledge, that may influence statutory drafting in ways that have been underappreciated and merit their own separate study.”); Gluck & Bressman, Part II, supra note 5, at 758 (“[O]ur respondents told us that first drafts are typically written by, respectively, the White House and agencies, or policy experts and outside groups, like lobbyists. Empirical work is lacking for the details of this account . . . .”); Nicholas R. Parrillo, Leviathan and Interpretive Revolution: The Administrative State, the Judiciary, and the Rise of Legislative History, 1890–1950, 123 Yale L.J. 266, 338–42 (2013) (documenting the key role of agencies in the legislative process during the New Deal era as part of a larger study of the rise of the use of legislative history); Christopher J. Walker, Inside Agency Statutory Interpretation, 67 Stan. L. Rev. 999, 1004 (2015) [hereinafter Walker, Inside Agency].

31Eskridge, Jr. et al., supra note 3, at 829 (“Indeed, since proposed legislation is frequently drafted by the executive department or by private interest groups, their statements and explanations at hearings might be the only truly informed explanation of the structure and operation of the statute.”); Eskridge, Jr., Interpreting Law, supra note 21, at 256.

32See Shobe, Agencies as Legistalators, supra note 5; Shobe, Intertemporal Statutory Interpretation, supra note 5; Walker, Inside Agency, supra note 30; Walker, Legislating in the Shadows, supra note 5.

33See cases cited infra note 54.

34See John F. Manning, Chevron and Legislative History, 82 Geo. Wash. L. Rev. 1517, 1525 n.46 (2014).

35See Shobe, Agencies as Legislatures, supra note 5, at 489.

36Id.

37See infra Sections II.B–II.D.

38The Mission and Structure of the Office of Management and Budget, Office of Mgmt. & Budget, https://obamawhitehouse.archives.gov/omb/organization_mission/ (last visited Oct. 20, 2018). These documents are often publicly available through agency websites or can be made available through FOIA requests. A number of websites request agency documents through FOIA and post them online. For example, the website governmentattic.org has posted certain Department of Justice views letters acquired through FOIA requests on their website. See, e.g., Copies of Certain Department of Justice (DOJ) Views Letters from the 107th and the 108th Congresses, 2001-2005, Governmentattic.org (Feb. 1, 2016) [hereinafter Copies of Certain Views Letters, 2001-2005], http://www.governmentattic.org/19docs/DOJviewsLetters_2001-2005.pdf.

39 Shobe, Agencies as Legislators, supra note 5, at 468 (“[A]gencies commonly originate their own legislative proposals and also draft legislation at Congress’s request”). That the Executive Branch would propose legislation is anticipated by the Constitution. The Recommendation Clause, found in Article II, Section 3, states that “[The President] shall from time to time . . . recommend to [Congress’s] Consideration such Measures as he shall judge necessary and expedient . . . .” U.S. Const. art. II, § 3. This provision underwent multiple drafts and the changes provide helpful context of what the Framers expected out of the President from this Clause. In an earlier draft the Clause allowed the President to recommend legislation but did not require him to do so. See J. Gregory Sidak, The Recommendation Clause, 77 Geo. L.J. 2079, 2081 (1989) (“James Madison’s notes on the Constitutional Convention for August 24, 1787, reveal that the Framers explicitly elevated the President’s recommendation of measures from a political prerogative to a constitutional duty . . . .”). An earlier version also contained the word “matters” rather than “measures.” Id. at 2084. This change “reinforces the inference that the Framers intended the President’s recommendations to be more than precatory statements . . . .” Id. This indicates that the Framers intended the President to make specific legislative proposals in the form of bill language. See Vasan Kesavan & J. Gregory Sidak, The Legislator-in-Chief, 44 Wm. & Mary L. Rev. 1, 48–49 (2002); Shobe, Agencies as Legisltaors, supra note 5, at 469 (“It is no secret that the President proposes bills to Congress. But where does that language come from? It doesn’t appear by magic. Someone in an agency is the one who wrote it.”). Agency proposals are sometimes made public by the agency as part of the President’s budget or as part of the agency’s process of promoting the proposal. Some agencies post their legislative proposals on their website. For example, the Department of Transportation recently drafted an expansive proposal to fund improvements to transportation infrastructure. Although the Department of Transportation is the lead agency, the 350-page draft bill includes roles for various agencies, including the EPA, Department of Interior, and Department of Labor. See, e.g., GROW AMERICA Act, H.R. 2410, 114th Cong. (2015), https://www.transportation.gov/sites/dot.gov/files/docs/GROW_AMERICA_Act_1.pdf (describing roles for various departments).

40 Congress has enacted a version of the National Defense Authorization Act every year since at least 1961. See House Armed Servs. Comm., History of the NDAA, https://armedservices.house.gov/ndaa/history-ndaa (last visited Oct. 20, 2018) (providing links to each National Defense Authorization Act since 1961). The DoD has an Office of Legislative Counsel whose job is to oversee the agency’s legislative proposals for the National Defense Authorization Act. See Dep’t of Def., Office of Legislative Counsel, An Overview of DoD’s Legislation Program 4 (2006).

41See Section III.A.

42See cases cited infra note 47.

43 310 U.S. 534, 549 (1940).

44 550 U.S. 81, 98 (2007).

45 139 Cong. Rec. 2,599 (1993) (House sponsor of the bill referring to the bill as “the administration’s proposal.”); Id. at 23,416 (Senate sponsor of the bill stating the bill was introduced by the agency. E.g., “I am pleased to introduce on behalf of the administration . . . .”; “The administration is proposing”; and “The administration’s proposal calls for . . . .”).

46Zuni Pub. Sch. Dist. No. 89, 550 U.S. at 98. Several Justices went out of their way to disavow Justice Breyer’s approach and to encourage the use of the traditional Chevron method. Id. at 107 (Kennedy, J., concurring).

47 Id. at 90 (examining the legislative history and finding that the “present statutory language” had come from draft legislation that the Secretary of the Department of Education had submitted to Congress in 1994); Smith v. City of Jackson, 544 U.S. 228, 239 (2005) (“[T]he Department of Labor . . . initially drafted the legislation [being examined in the case.]”); Howe v. Smith, 452 U.S. 473, 485 (1981) (arguing that “the Bureau’s interpretation of the statute merits greater than normal weight because it was the Bureau that drafted the legislation and steered it through Congress . . . .”); United States v. One Bell Jet Ranger II Helicopter, 943 F.2d 1121, 1126 (9th Cir. 1991) (“The interpretation of the agency charged with administering the statute is entitled. . . to ‘greater than normal weight’ when that agency drafts the legislation and steers it through Congress with little debate.”); Watkins v. Blinzinger, 789 F.2d 474, 478 (7th Cir. 1986) (arguing that when an agency drafted language courts should defer unless the interpretation is “beyond the pale of reasonableness”); W. Nuclear, Inc. v. Andrus, 664 F.2d 234, 240 (10th Cir. 1981); Beshaw v. Fenton, 635 F.2d 239, 245 (3d Cir. 1980) (stating that the agency interpretation was given more weight because it drafted the section at issue and got it through Congress); Saldivar v. Rodela, 894 F. Supp. 2d 916, 930 (W.D. Tex. 2012); Creaton v. Bowen, No. CV 85-3306-R, 1986 U.S. Dist. LEXIS 25485, at *12 (C.D. Cal. 1986).

48 396 U.S. 212, 217–18 (1969).

49Id. (“In fact, a representative of the Labor Department objected to the bill precisely for that reason, urging the Committee to extend coverage to embrace the contract, ‘and not the man simply when he is on the ship.’ If Congress had intended to adopt that suggestion, it could not have chosen a more inappropriate way of expressing its intent than by substituting the words ‘upon the navigable waters’ for the words ‘within the admiralty jurisdiction.’”).

50Id.

51 Shobe, Agencies as Legislators, supra note 5, at 468 (“[A]gencies provide extensive review of, and revisions to, statutory language drafted by outside agencies.”). Of course, it is not always possible for a court to determine whether an agency was involved in drafting. But, often it is. Although it is more difficult to tell when an agency proposed edits to a bill, it is not uncommon for agencies to submit edits as part of a views letter or through other public means. There are also strong indicators of when an agency did not draft language. For example, if an agency or administration issues a views letter or testifies in opposition of a bill, or certain sections of a bill, then it would seem likely that the agency did not draft, and was not heavily involved in reviewing, the bill or those sections of the bill.

52Id.

53Id. at 482.

54See, e.g., Aluminum Co. of Am. v. Cent. Lincoln Peoples’ Util. Dist., 467 U.S. 380, 390 (1984) (noting that the agency “was intimately involved in the drafting and consideration of the [Regional Act] by Congress”); Guardians Ass’n v. Civil Serv. Comm’n of the City of New York, 463 U.S. 582, 592 (1983); Pub. Util. Dist. No. 1 of Douglas County v. Bonneville Power Admin., 947 F.2d 386, 390 (9th Cir. 1991) (arguing that because the agency was involved in drafting the court should defer unless the interpretation is unreasonable); Aluminum Co. of Am. v. Bonneville Power Admin. 903 F.2d 585, 590 (9th Cir. 1990) (same); Cal. Energy Res. Conservation & Dev. Co. v. Johnson, 807 F.2d 1456, 1459 (5th Cir. 1987) (same); Almendarez v. Barrett-Fisher Co., 762 F.2d 1275, 1282 n.3 (5th Cir. 1985); Sweeny v. Murray, 732 F.2d 1022, 1029 (1st Cir. 1984) (arguing that greater deference should be given when agency played a role in drafting statute); McDannell v. U.S. Office of Pers. Mgmt., 716 F.2d 1063, 1066 (5th Cir. 1983) (giving greater deference to the Office of Personnel Management (defendant-appellant) in their interpretation of the statute because their predecessor agency had “an actual hand in its drafting and passage”); Int’l Nutrition, Inc. v. U.S. Dep’t of Health & Human Servs., 676 F.2d 338, 342 (8th Cir. 1982) (arguing that because the FDA participated in drafting the statute, the court defers to its interpretation, as long as that interpretation furthers goals of legislation); Comm. for Auto Responsibility v. Solomon, 603 F.2d 992, 1004 (D.C. Cir. 1979) (noting that the General Services Administration had been actively involved in drafting and adopting the statutory language that was in dispute in the case); Hercules, Inc. v. EPA, 598 F.2d 91, 125 (D.C. Cir. 1978); Morgan v. United States, No. CV 84-4664 (RR), 1991 WL 353371, at *4 (E.D.N.Y. June 30, 1991) (“It is axiomatic that where an agency assists in the drafting of legislation and aids in its passage, it views on that legislation are entitled to great deference.”); Air Courier Conference of Am./Int’l Comm. v. U.S. Postal Serv., 762 F. Supp. 86, 88 (D. Del. 1991) (noting that “due deference” was supported by the fact that Postal Department officials had participated in the Act’s drafting); Faught v. Heckler, 577 F. Supp. 1180, 1186 (S.D. Iowa 1983); Turney v. United States, 525 F. Supp. 675, 677 (D. Md. 1981) (“Administrative interpretations are especially persuasive where, as here, the agency participated in developing the provision.”); Am. Waterways Operators, Inc. v. United States, 386 F. Supp. 799, 804 (D.D.C. 1974) (“Of higher significance, however, is the construction placed on an act by those administrators who participated in its drafting and directly made known their views to Congress.”).

55See supra Section II.B.

56 440 U.S. 125, 144 (1979).

57 319 U.S. 312, 315 (1943).

58See cases cited supra note 54. Courts that have applied this canon of greater deference have not automatically accepted agency interpretations. Aluminum Co. of Am., 903 F.2d at 594, 598–99 (noting that the commission had a significant role in drafting the legislation); Watkins v. Blinzinger, 789 F.2d 474, 478 (7th Cir. 1986) (stating that the court should defer to agency interpretations where the agency drafted the language, unless the interpretation is “beyond the pale of reasonableness”); Cent. Lincoln Peoples’ Util. Dist. v. Johnson, 735 F.2d 1101, 1106 (9th Cir. 1984) (noting that the commission had a significant role in drafting the legislation); Watts v. Hadden, 686 F.2d 841, 843 (10th Cir. 1981) (same); Patagonia Corp. v. Bd. of Governors, 517 F.2d 803, 811, 813 (9th Cir. 1975) (noting that the agency suggested the language at issue but still relying on an analysis of the meaning of the words to rule against the agency). Much like in Chevron, these courts have still looked to whether the interpretation was “permissible” or “reasonable.” Chevron U.S.A. Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 840, 866 (1984). The Supreme Court in Zuber v. Allen implied that judges should look beyond the mere fact of agency involvement in drafting to other types of agency legislative history to make sure that no other evidence pointed to a disagreement between Congress and the responsible agency. 396 U.S. 168, 181–82, 184, 188 (1969).

59 818 F.2d 953, 960–61 (D.C. Cir. 1987) (“It is well established that the prestige of a statutory construction by an agency depends crucially upon whether it was promulgated contemporaneously with enactment of the statute.”).

60 818 F.2d 1148, 1155–56 n.4 (5th Cir. 1987); see also Aluminum Co. of Am. v. Cent. Lincoln Peoples’ Util. Dist., 467 U.S. 380, 390 (1984) (noting that the agency was involved in drafting the statute and its interpretation came immediately following the enactment of the statute); Howe v. Smith, 452 U.S. 473, 474 (1981) (noting that the agency’s interpretation was “contemporaneous”).

61 For examples of agency section-by-section analyses, see OLC Detailed Guidelines for Preparation of Legislative Proposals for the Fiscal Year 2016 DoD Legislative Program, Dep’t of Def. (2014), http://ogc.osd.mil/olc/docs/DetailedGuidelinesforPreparingProposalsFY16.pdf (discussing section-by-section analysis requirement for DoD proposals); Surface Transportation Reauthorization Bill, U.S. Dep’t of Transp. (Feb. 11, 2015), https://www.transportation.gov/content/surface-transportation-reauthorization-bill-laying-foundation-us-economic-growth-and-job (providing proposed legislative language and accompanying section-by-section analysis to Congress).

62Eskridge, Jr., Interpreting Law, supra note 21, at 242.

63 Archer-Daniels-Midland Co. v. United States, 37 F.3d 321, 323–24 (7th Cir. 1994) (Posner, C.J.). That members of Congress are more likely to read a committee report than actual statutory language is confirmed by some of those involved in the legislative process. See Gluck & Bressman, Part I, supra note 1, at 968–69. Judge Katzmann recently explained, in light of “the expanding, competing demands on legislators’ time,” “they cannot read every word of the bills they vote upon, but they, and certainly their staffs, become educated about the bill by reading the materials produced by the committees and conference committees from which the proposed legislation emanates.” Robert A. Katzmann, Statutes, 87 N.Y.U. L. Rev. 637, 653 (2012).

64See Lisa Schultz Bressman, Chevron’s Mistake, 58 Duke L.J. 549, 582 (2009) (“[A]n agency may be involved in drafting the legislation, sharing its interpretations with legislative staff. If those understandings are sufficiently specific, Congress may rely on them when enacting legislation.”).

65See, e.g., DoD Legislative Proposals, Fiscal Year 2017, Dep’t of Def., http://ogc.osd.mil/olc/legispro17.html (last visited Oct. 20, 2018).

66See, e.g., National Defense Authorization Act for Fiscal Year 2017, H.R. 4909, 114th Cong. §§ 113, 143 (2d Sess. 2016). The Farm Bill is an example of another must-pass bill that receives significant agency input. See, e.g., Agriculture, Rural Development, Food and Drug Administration, and Related Agencies Appropriations Act, H.R. 5054, 114th Cong. (2d Sess. 2016).

67DoD Legislative Proposals, Fiscal Year 2018, Dep’t of Def., http://ogc.osd.mil/olc/legispro18.html (last visited Oct. 20, 2018).

68See, e.g., Morgan v. United States, No. CV 84-4664 (RR), 1991 WL 353371, at *4 (E.D.N.Y. June 30, 1991) (citing to a views letter from the Veterans Administration showing the agency’s position and interpretation of what the amendment would do); see also Dep’t of Defense, Office of Legislative Counsel, supra note 40, at 4 (“Periodically, the Chairman of the House or Senate Armed Services Committee requests that DoD review a bill and provide an official position, or ‘views’ of the Department via a formal letter on the merits of the subject bill. Such a letter declaring DoD’s official position, or ‘views,’ is commonly referred to as a ‘Views Letter.’ OLC receives the requests for Departmental views letters and coordinates them with all concerned DoD components.”); Nat’l Oceanic & Atmospheric Admin., Congressional Testimony and Views Letters, Off. of Legis. & Intergovernmental Aff., http://www.legislative.noaa.gov/115testimony.html (last visited Oct. 20, 2018) (“Views letters are statements of the Agency’s, Department’s, and Administration’s position, thoughts, and comments, on specific issues or legislation being considered by Congress.”); Views Letters, U.S. Dep’t of Just., Off. of Legis. Aff., https://www.justice.gov/ola/views-letters (last visited Oct. 20, 2018); Letters to Congress, Dep’t of Com., Off. of the Gen. Couns., https://ogc.commerce.gov/letters (last visited Oct. 20, 2018); Copies of Certain Department of Justice (DOJ) Views Letters from the 109th and 110th Congresses, 2005–2007, Governmentattic.org (Nov. 14, 2014), http://www.governmentattic.org/13docs/DOJviewsLetters_2005-2007.pdf (compiling Department of Justice views letters through FOIA requests).

69See Negonsott v. Samuels, 507 U.S. 99, 106 (1993) (noting that the Department of Interior proposed amendments to the original bill in a letter to Congress).

70See supra note 68.

71See supra note 68.

72 For example, the America Invents Act (AIA) was enacted after many years of negotiation between the United States Patent and Trademark Office (USPTO) and various congressional committees. The USPTO prepared an early version of the legislation, then proceeded to send at least six views letters and various reports to congressional committees in the following years. Letter from Gary Locke to the Honorable Patrick J. Leahy, Chairman, Senate Comm. on the Judiciary, and the Honorable Jeff Sessions, Ranking Member, Senate Comm. on the Judiciary (Apr. 20, 2010) (on file with U.S. Dep’t of Commerce) (stating that the new post-grant review procedures “will serve as a faster, lower-cost alternative to litigation.”); Letter from John J. Sullivan to the Honorable Patrick J. Leahy, Chairman, Senate Comm. on the Judiciary, and the Honorable Arlen Specter, Ranking Member, Senate Comm. on the Judiciary (May 18, 2007) (on file with U.S. Dep’t of Commerce); Letter from Nathaniel F. Wienecke to the Honorable Patrick J. Leahy, Chairman, Senate Comm. on the Judiciary 4 (Feb. 4, 2008) (on file with U.S. Dep’t of Commerce) (“The Administration supports establishment of an effective, efficient post-grant patent review process that truly functions as a lower-cost alternative to litigation . . . .”). This bill was subject to much litigation, including a case that reached the Supreme Court very recently, Cuozzo Speed Technologies, LLC v. Lee, 136 S. Ct. 2131 (2016), but the Court made no mention of this agency legislative history.

73 Agencies generally post this testimony on their websites. See, e.g., Congressional Testimony, U.S. Dep’t of Just., Off. of Legis. Aff., https://www.justice.gov/ola/congressional-testimony (last visited Oct. 20, 2018); Congressional Testimony, U.S. Dep’t of Lab., http://www.dol.gov/newsroom/congress (last visited Oct. 20, 2018); Congressional Testimony, U.S. Dep’t of St., Bureau of Legis. Aff., https://www.state.gov/s/h/tst/ (last visited Oct. 20, 2018); Testimony, Dep’t of Homeland Security, https://www.dhs.gov/news-releases/testimony (last visited Oct. 20, 2018); Testimony, U.S. Dep’t of Transp., Off. of Inspector Gen., https://www.oig.dot.gov/testimony (last visited Oct. 20, 2018).

74See Dep’t of Def., Office of Legislative Counsel, supra note 40, at 5 (“Congress frequently invites DoD leaders to testify before various committees and subcommittees. Because the subject of the testimony often crosses jurisdictional boundaries with other DoD components and government agencies, once again extensive coordination is required to ensure that DoD, and ultimately the entire Administration, speak with one voice.”).

75 For example, the Government Publishing Office maintains a record of hearings for each House and Senate Committee. See Browse by Committee, Govinfo, https://www.govinfo.gov/browse/committee (last visited Oct. 20, 2018) (listing each House and Senate committee with links to congressional hearings for each committee).

76 Patricia M. Wald, Some Observations on the Use of Legislative History in the 1981 Supreme Court Term, 68 Iowa L. Rev. 195, 202 (1983) (“The hornbook rule that hearings are relevant only as background to show the purpose of the statute no longer holds. In many cases the best explanation of what the legislation is about comes from the executive department or outside witnesses at the hearings.”).

77 455 U.S. 16, 29 (1982).

78Id. at 28.

79Id. at 31–32.

80Id. at 32.

81 789 F.2d 474, 475 (7th Cir. 1986).

82Id. at 475 n.2.

83Id. at 479–80.

84Id. at 480.

85 943 F.2d 1121, 1126 (9th Cir. 1991).

86Id.

87Id. at 1122.

88Id. at 1126 (citing S. Rep. No. 92-1157 (1972), as reprinted in 1972 U.S.C.C.A.N. 3835, 3839).

89Id.

90Id.

91 436 U.S. 268, 270 (1978).

92Id. at 276.

93Id.

94Id. at 276–77 (explaining that the amendment was created “to meet the objection of Treasury to the discharge of so-called trust fund taxes”).

95Id. at 272.

96Id. at 277.

97 470 U.S. 768, 784–85 (1985).

98Id. at 780.

99Id. at 784 (“Thereafter, the full Committee adopted an amendment in the nature of a substitute to H.R. 2510 that limited full judicial review ‘to cases involving agency-filed applications for disability retirement based on an employee’s mental condition.’”).

100Id. at 784–85 (citing Letter from Alan K. Campbell, Dir., OPM, to Representative James M. Hanley, Chairman, House Comm. on Post Office and Civil Service (May 14, 1980), reprinted in H.R. Rep. No. 96-1080, at 2 (1980); Letter from Alan K. Campbell, Dir., OPM, to Senator Abraham A. Ribicoff, Chairman, Senate Comm. on Governmental Affairs (Sept. 25, 1980), reprinted in S. Rep. No. 96-1004, at 4–5 (1980).

101Id. at 785.

102 484 U.S. 174, 185 (1988).

103Id.

104Id.

105Id. at 185–86.

106 For examples other than those discussed in the text here, see Wright v. City of Roanoke Redev. & Hous. Auth., 479 U.S. 418, 425 n.7 (1987); Pension Benefit Guar. Corp. v. R.A. Gray & Co., 467 U.S. 717, 722 n.2 (1984); Monroe v. Standard Oil Co., 452 U.S. 549, 558 (1981) (“The legislation was proposed by the Department of Labor. Accordingly, the testimony [of agency staff] . . . is instructive.”); Dawson Chem. Co. v. Rohm & Haas Co., 448 U.S. 176, 204 (1980) (relying on DOJ testimony as evidence that the legislature intended to expand patentee protection under Section 271(d) of the Patent Act); Friedsam v. Nicholson, 19 Vet. App. 222, 225–26 (2005) (reviewing agency testimony and determining that it was relevant to the case).

107 437 U.S. 153, 179–80 (1978).

108Id. (“These provisions were designed, in the words of an administration witness, ‘for the first time [to] prohibit [a] federal agency from taking action which does jeopardize the status of endangered species,’ . . . . [T]he proposed bills would ‘[direct] all . . . Federal agencies to utilize their authorities for carrying out programs for the protection of endangered animals.’”).

109 396 U.S. 168, 192 (1969).

110 377 U.S. 46, 48–49 (1964).

111Id. at 51.

112Id. at 52 n.8, 53 n.9.

113 430 U.S. 1, 33 (1977).

114Id.

115Id. at 34–35.

116 416 U.S. 505, 507–08 (1974).

117Id. at 512.

118Id. at 516–17.

119Id. at 516.

120Id. at 520.

121Id. at 517.

122Id. at 520.

123See supra Section II.E.

124 This is confirmed by agency staff who are involved in the legislative process. See Shobe, Agencies as Legislators, supra note 5, at 491 n.156 (“[A]gencies offer section-by-section analysis of bills drafted in an agency and supply other documents and studies to Congress, and respondents reported that these types of agency-produced documents can end up in a committee or conference report.”). Interestingly, Congress has, on occasion, incorporated agency legislative history into enacted statutory language. See, e.g., Medical Device User Fee and Modernization Act of 2002, Pub. L. No. 107-250, § 101(3), 116 Stat. 1588, 1589 (2002) (“[T]he fees authorized by this title will be dedicated to meeting the goals identified in the letters from the Secretary of Health and Human Services to the Committee on Energy and Commerce of the House of Representatives and the Committee on Health, Education, Labor, and Pensions of the Senate, as set forth in the Congressional Record.”); Food and Drug Administration Modernization Act of 1997, Pub. L. No. 105-115, § 101(4) , 111 Stat. 2296, 2298 (1997) (“[T]he fees authorized by amendments made in this subtitle will be dedicated toward expediting the drug development process and the review of human drug applications as set forth in the goals identified, for purposes of part 2 of subchapter C of chapter VII of the Federal Food, Drug, and Cosmetic Act, in the letters from the Secretary of Health and Human Services to the chairman of the Committee on Commerce of the House of Representatives and the chairman of the Committee on Labor and Human Resources of the Senate, as set forth in the Congressional Record.”).

125See, e.g., H.R. Rep. No. 112-98, at 85–88 (2011) (incorporating a views letter from the Department of Commerce); S. Rep. No. 110-143, at 17–23 (2007) (including both excerpts from a statement of Under Secretary for Benefits of the Department of Veterans Affairs and a letter from the Secretary of the Department of Veterans Affairs); S. Rep. No. 108-264, at 28–29 (2004) (including a letter from Director of Office of Congressional and Legislative Affairs of Dept. of Interior); S. Rep. No. 106-325, at 5–9 (2000) (including a letter from Henry H. Shelton, Chairman of the Joint Chiefs of Staff, and William S. Cohen, Secretary of Defense); S. Rep. No. 105-207, at 14–17 (1998) (including two letters from Kevin Gover, Assistant Secretary–Indian Affairs); H.R. Rep. No. 99-660, at 63–67 (1986) (including letters from John R. Bolton, Assistant Attorney General, United States Department of Justice).

126See, e.g., H.R. Rep. No. 114-396, at 13–14 (2016) (“[T]he Department of Homeland Security Inspector General recently released a report entitled ‘TSA Can Improve Aviation Worker Vetting’ . . . . The report made six recommendations to strengthen the vetting of credentialed aviation workers. This legislation codifies several of those recommendations and ensures that TSA has access to the necessary data to properly vet aviation employees, strengthen its criminal background check capabilities, and better-resolve issues of lawful status for credential applicants.”); S. Rep. No. 105-20, at 8–12 (1997) (including testimony from Gary Niles Kimble, Commissioner, Administration for Native Americans, Administration for Children and Families–U.S. Department of Health and Human Services); NAACP v. Detroit Police Officers Ass’n, 900 F.2d 903, 909 (6th Cir. 1990) (considering a statement from the Justice Department that had been placed in the Congressional Record); Krinsk v. Fund Asset Mgmt., Inc., No. 85 Civ. 8428 (JMW), 1986 WL 205, at *4 n.6 (S.D.N.Y. 1986) (citing to a memo drafted by the SEC and included in the House committee report).

127 Negonsott v. Samuels, 507 U.S. 99, 106–09 (1993).

128See, e.g., H.R. Rep. No. 114-107, pt. 1, at 245 (2015) (“We even received a letter from the widely respected Secretary of Energy. This may well be the first time in the history of this Committee that a sitting Cabinet member has provided formal opposition to a piece of legislation that we are considering, certainly at this stage of the legislative process. That should be a strong indication of just how bad this bill really is.”); H.R. Rep. No. 111-97, at 4 (2009) (“Following the hearings, the legislation was refined to take into account concerns raised by the Department of Justice and potential defendants, and the False Claims Amendments Act of 1986 was enacted on October 27, 1986.”); S. Rep. No. 105-379, at 10 (1998) (“At the hearing held on July 8, 1998, the Department of the Interior testified that it could not support S. 391 unless section 9(d) (Affirmative Defenses Waived) was removed.”); Am. Tobacco Co. v. Patterson, 456 U.S. 63, 73 (1982) (noting that a Justice Department statement was placed in the Congressional Record by one of the bill’s sponsors); Int’l Bhd. of Teamsters v. United States, 431 U.S. 324, 351 (1977) (same); Franks v. Bowman Transp. Co., 424 U.S. 747, 759–60 (1976) (same).

129See supra note 65.

130Compare Dep’t of Def., 2008 Section-by-Section Analysis 10, 16 (2008), http://ogc.osd.mil/olc/docs/26April2013NDAASectionalAnalysis.pdf, with S. Rep. No. 110-77, at 398–99 (2008); compare Dep’t of Def., 2013 Section-by-Section Analysis 27 (2013), http://ogc.osd.mil/olc/docs/26April2013NDAASectionalAnalysis.pdf, with S. Rep. No. 112-173, at 103 (2012); compare Dep’t of Def., 2014 Section-by-Section Analysis 14, 68, 136 (2014), http://ogc.osd.mil/olc/docs/26April2013NDAASectionalAnalysis.pdf, with H.R. Rep. No. 113-102, at 173–74, 187, 362 (2014); compare Dep’t of Def., 2016 Section-by-Section Analysis 141 (2016), http://ogc.osd.mil/olc/docs/26April2013NDAASectionalAnalysis.pdf, with S. Rep. No. 114-49, at 178 (2016).

131 For example, in the America Invents Act, the USPTO was especially influential in drafting and revising various versions of the bill. The USPTO sent a number of views letters to the relevant committees and these views letters used language very similar to language that ended up in the congressional committee reports about the purpose and function of the post-grant review proceedings that were an important part of the AIA. Compare Letter from Gary Locke to the Honorable Patrick J. Leahy, Chairman, Senate Comm. on the Judiciary, and the Honorable Jeff Sessions, Ranking Member, Senate Comm. On the Judiciary, supra note 72, and Letter from Nathaniel F. Wienecke to the Honorable Patrick J. Leahy, Chairman, Senate Comm. on the Judiciary, supra note 72, with Cuozzo Speed Techs., LLC v. Lee, 136 S. Ct. 2131, 2143 (2016) (“Inter partes review is a ‘quick and cost effective alternativ[e] to litigation.”) (citing to the House Committee Report), and id. (“Inter partes review is ‘a quick, inexpensive, and reliable alternative to district court litigation”) (citing S. Rep. No. 110-259, at 20 (2008)). The House Committee Report for the AIA included a views letter from the Department of Commerce. See H.R. Rep. No. 112-98, at 85–88 (2011).

132 The website governmentattic.org has posted certain DOJ views letters acquired through FOIA requests on its website. See, e.g., Copies of Certain Views Letters, 2001-2005, supra note 38. As part of this project the author made a number of FOIA requests for agency legislative history and while some agencies were able to provide the relevant documents, other agencies were unable to find them, especially those created many years ago. Requiring litigants to use FOIA requests to uncover agency legislative history is an imperfect solution to the access problem.

133 One solution to level the playing field would be to make the types of formal agency legislative history discussed here available in easily searchable online sources like the U.S. Code Congressional and Administrative News, although this would obviously be a costly and burdensome project. Precedent exists for this type of project. The Reagan Administration did this with presidential signing statements when it entered into an agreement with West Publishing Company to publish these statements. Marc N. Garber & Kurt A. Wimmer, Presidential Signing Statements as Interpretations of Legislative Intent: An Executive Aggrandizement of Power, 24 Harv. J. on Legis. 363, 367 (1987) (citing Att’y Gen. Edwin Meese III, Address to the National Press Club, Washington, D.C. (Feb. 25, 1986)) (“[W]e have now arranged with the West Publishing Company that the presidential statement on the signing of a bill will accompany the legislative history from Congress so that all can be available to the court for future construction of what the statute really means.”). However, this would not necessarily address the concerns raised here, because it would still allow agencies to fill the legislative record with their preferred interpretations without congressional oversight.

134See Abbe R. Gluck et al., Unorthodox Lawmaking, Unorthodox Rulemaking, 115 Colum. L. Rev. 1789, 1857–58 (2015) (“[M]any of the Court’s interpretive rules aim to improve how Congress drafts. . . . Textualists have long argued that one salutary effect of a text-centric approach is that it teaches Congress to draft better the next time.”). Of course, there is a debate over whether courts should be trying to shape congressional behavior at all, or merely reflect it. See id. at 1835–36.

135 Critics of legislative history have argued that the more legislative history is used, the less reliable it becomes, and this same argument could be leveled against agency legislative history. Blanchard v. Bergeron, 489 U.S. 87, 98–99 (1989) (Scalia, J., concurring in part and concurring in the judgment); see also Int’l Bhd. of Elec. Workers, Local Union No. 474 v. NLRB, 814 F.2d 697, 717 (D.C. Cir. 1987) (Buckley, J., concurring) (arguing that the use of legislative history by courts creates an incentive “to salt the legislative record with unilateral interpretations of statutory provisions they were unable to persuade their colleagues to accept”).

136See infra Part III (discussing courts’ use of agency legislative history, including instances where the agency legislative history is included in congressional legislative history).

137 It could be that members of Congress or their staff want to take credit for drafting the legislative history because of some kind of prestige or status benefits associated with being the drafter. It seems that this benefit would be outweighed by the benefits described here to Congress from memorializing agency–Congress legislative deals.

138Vermeule, supra note 12, at 206; Eskridge, Jr., supra note 10; Mashaw, supra note 10, at 93–95; Spence & Cross, supra note 10, at 138–41.

139See supra note 138.

140See infra notes 147–57 and accompanying text.

141See, e.g., SAS Inst., Inc. v. Iancu, 138 S. Ct. 1348, 1358 (2018).

142See infra Sections III.A.1.–III.A.3.

143See supra Section II.D.

144See, e.g., Frank H. Easterbrook, The Role of Original Intent in Statutory Construction, 11 Harv. J.L. & Pub. Pol’y 59, 64–65 (1998) (discussing complexities of the legislative process); Kenneth A. Shepsle & Barry R. Weingast, The Institutional Foundations of Committee Power, 81 Am. Pol. Sci. Rev. 85, 89 (1987) (same).

145Vermeule, supra note 12, at 190; Frank H. Easterbrook, Statutes’ Domains, 50 U. Chi. L. Rev. 533, 548 (1983) (arguing that the many complexities of the legislative process are “so integral to the legislative process that judicial predictions of how the legislature would have decided issues it did not in fact decide are bound to be little more than wild guesses”).

146Vermeule, supra note 12, at 113.

147 Antonin Scalia, Judicial Deference to Administrative Interpretations of Law, 1989 Duke. L.J. 511, 517 (1989).

148 City of Arlington v. FCC, 569 U.S. 290, 306 (2013).

149See Henry M. Hart, Jr. & Albert M. Sacks, The Legal Process: Basic Problems in the Making and Application of Law 1183 (William N. Eskridge, Jr. & Philip P. Frickey eds., 1994); William N. Eskridge, Jr. & Philip P. Frickey, Statutory Interpretation as Practical Reasoning, 42 Stan. L. Rev. 321, 348–49 (1990).

150 Easterbrook, supra note 145, at 547–48.

151 Pildes, supra note 13.

152 United States v. Mead Corp., 533 U.S. 218, 240–45 (2001) (Scalia, J., dissenting).

153See, e.g., Vermeule, supra note 12; Mashaw, supra note 12 (“There are persuasive grounds for believing that legitimate techniques and standards for agency statutory interpretation diverge sharply from the legitimate techniques and standards for judicial statutory interpretation”); Sunstein & Vermeule, supra note 12.

154 Sitaraman, supra note 5, at 128 (“The executive’s role in legislative drafting provides additional support to the Strauss-Mashaw thesis that agency interpretive practice can and should diverge from judicial interpretive process.”); Walker, Legislating in the Shadows, supra note 5, at 1403 (arguing that the role of agencies in the legislative process provides “additional empirical support for a more purposivist approach to agency statutory interpretation”).

155See infra notes 156–58.

156 Stephen Breyer, Judicial Review of Questions of Law and Policy, 38 Admin. L. Rev. 363, 368 (1986); see also Peter L. Strauss, When the Judge Is Not the Primary Official with Responsibility to Read: Agency Interpretation and the Problem of Legislative History, 66 Chi.-Kent L. Rev. 321, 329–30 (1990) [hereinafter Strauss, When the Judge Is Not the Primary] (noting that agencies maintain “transcripts of relevant hearings, correspondence, and other informal traces of the continuing interactions that go on between an agency and Capitol Hill as a statute is being shaped in the legislative process . . . .”).

157 Scalia, supra note 147, at 514.

158Vermeule, supra note 12, at 209.

159 Daniel B. Rodriguez & Barry R. Weingast, The Positive Political Theory of Legislative History: New Perspectives on the 1964 Civil Rights Act and Its Interpretation, 151 U. Pa. L. Rev. 1417, 1445–46, 1448 (2003); Strauss, When the Judge Is Not the Primary, supra note 156, at 347 (arguing that agencies are better able “to distinguish reliably those considerations that served to shape the legislation, the legislative history wheat, from the more manipulative chaff”).

160 The clearest articulation of this at the Supreme Court is in Smiley v. Citibank (South Dakota), N.A.: We accord deference to agencies under Chevron . . . because of a presumption that Congress, when it left ambiguity in a statute meant for implementation by an agency, understood that the ambiguity would be resolved . . . by the agency, and desired the agency (rather than the courts) to possess whatever degree of discretion the ambiguity allows.517 U.S. 735, 740–41 (1996).

161 Gluck and Bressman, Part II, supra note 5, at 765.

162See, e.g., Michigan v. EPA, 135 S. Ct. 2699, 2712 (2015) (Thomas, J., concurring in the judgment) (arguing that the EPA’s “request for deference raises serious questions about the constitutionality of our broader practice of deferring to agency interpretations of federal statutes”); King v. Burwell, 135 S. Ct. 2480, 2488–89 (2015); Dep’t of Transp. v. Ass’n of Am. R.Rs., 135 S. Ct. 1225, 1254 (2015) (Thomas, J., concurring) (describing the administrative state as a system that “concentrates the power to make laws and the power to enforce them in the hands of a vast and unaccountable administrative apparatus”). For an excellent discussion of the rise of the administrative state, see Lawson, supra note 16.

163 City of Arlington v. FCC, 569 U.S. 290, 314–15 (2013).

164See, e.g., Lisa Schultz Bressman, supra note 64 (arguing against automatic deference). The House recently passed a bill aimed at regulatory agencies that directly repudiated Chevron. Separation of Powers Restoration Act, H.R. 5, 115th Cong. § 202(1)(B) (2017) (“If the reviewing court determines that a statutory or regulatory provision relevant to its decision contains a gap or ambiguity, the court shall not interpret that gap or ambiguity as [1] an implicit delegation to the agency of legislative rulemaking authority . . . or [2] a justification for interpreting agency authority expansively or for deferring to the agency’s interpretation on the question of law.”). It is possible that broad judicial deference to agencies makes Congress less willing to pass legislation, especially in times of divided government, because Congress cannot rely on courts to ensure agencies carry out Congress’s will. If courts began to consider agency legislative history, then legislative gridlock might be, at least somewhat, reduced.

165See, e.g., Michigan, 135 S. Ct. at 2712 (Thomas, J., concurring in the judgment) (arguing that the transfer of interpretive authority from courts to agencies “is in tension with Article III’s Vesting Clause, which vests the judicial power exclusively in Article III courts, not administrative agencies”); Gutierrez-Brizuela v. Lynch, 834 F.3d 1142, 1149 (10th Cir. 2016) (Gorsuch, J., concurring) (“[T]he fact is Chevron and Brand X permit executive bureaucracies to swallow huge amounts of core judicial and legislative power and concentrate federal power in a way that seems more than a little difficult to square with the Constitution of the framers’ design. Maybe the time has come to face the behemoth.”); Philip Hamburger, Is Administrative Law Unlawful? 315–17 (2014).

166 Walker, supra note 5, at 1419.

167 Auer v. Robbins, 519 U.S. 452, 461 (1997).

168 Talk Am., Inc. v. Mich. Bell Tel. Co., 564 U.S. 50, 69 (2011) (Scalia, J., concurring).

169 Walker, supra note 5, at 1411 (“If the agency is indeed a partner with Congress in legislative drafting, Justice Scalia’s concern about an agency legislating and executing the law should apply with some force to legislative drafting. The executive and legislative functions are, in essence, combined via agency legislating in the shadows.”).

170Id. at 1380–81.

171See, e.g., James J. Brudney, Essay, Contextualizing Shadow Conversations, 166 U. Pa. L. Rev. Online 37, 41–42 (2017) (recounting the author’s experience as a congressional staffer and that this experience did not indicate that Congress ceded authority to agencies and instead indicated that “the congressional drafting process is complex, messy, and far from uniform”); Gluck & Bressman, Part I, supra note 1, at 1000 (reporting “that agencies often . . . participate in drafting and can be very useful partners,” but that congressional staff “sometimes avoided the agency, particularly when they were aware of a conflicting position”).

172 Gluck & Bressman, Part II, supra note 5 (“[O]ur respondents told us that first drafts are typically written by, respectively, the White House and agencies, or policy experts and outside groups, like lobbyists. Empirical work is lacking for the details of this account . . . .”); see Kenneth W. Starr, Observations About the Use of Legislative History, 1987 Duke L.J. 371, 376 (“Lobbyists maneuver to get their clients’ opinions into the mass of legislative materials . . . .”).

173 Victoria F. Nourse & Jane S. Schacter, The Politics of Legislative Drafting: A Congressional Case Study, 77 N.Y.U. L. Rev. 575, 611 (2002) (“Lobbyists are the closest to the people who will be affected by the bill . . . .”).

174 Shobe, Intertemporal Statutory Interpretation, supra note 5, at 847–49. Spending on lobbyists has increased significantly in recent years, so there is reason to believe they are taking an active role in policing legislative language. Id. at 847 (“Recent reports show that spending on lobbying has more than doubled in the last fifteen years, from $1.45 billion in 1998 to $3.3 billion in 2012. Long-serving committee staff and legislative counsel anecdotally report that lobbyist involvement in the drafting process has increased significantly over the last twenty years.”).

175See supra Section II.B.

176 550 U.S. 81, 98, 107 (2007).

177See Barnett v. Weinberger, 818 F.2d 953, 960–61 (D.C. Cir. 1987); cases cited supra note 60 and accompanying text.

178See supra Section III.B. (discussing Nacirema Operating Co. v. Johnson, 396 U.S. 212 (1969)).

179See Lisa Schultz Bressman, Essay, Reclaiming the Legal Fiction of Congressional Delegation, 97 Va. L. Rev. 2009, 2015, 2025 (2011) (demonstrating that a consensus exists in the academic literature that Chevron is based on a fiction); Sitaraman, supra note 5, at 129; Strauss, Chevron Space” and “Skidmore Weight”, supra note 26, at 1145 (arguing that courts should use a case-by-case common law approach to determine when to give deference); Walker, supra note 5, at 1409.

180 City of Arlington v. FCC, 569 U.S. 290, 318 (2013) (Roberts, C.J., dissenting).

181Id. at 309 (Breyer, J., concurring in part and concurring in the judgment) (quoting Barnhart v. Walton, 535 U.S. 212, 222 (2002)).

182 Breyer, supra note 156, at 371–73.

183 United States v. Mead Corp., 533 U.S. 218, 231 (2001). Mead famously restricted the application of Chevron to cases where Congress explicitly delegated, and agencies followed, relatively formal procedures. See id. at 226–27; Cass R. Sunstein, Beyond Marbury: The Executive’s Power to Say What the Law Is, 115 Yale L.J. 2580, 2602–03 (2006).

184 Peter L. Strauss, Courts or Tribunals? Federal Courts and the Common Law, 53 Ala. L. Rev. 891, 893 (2002).

185Id.

186 Gluck & Bressman, Part II, supra note 5, at 791.

187 Walker, supra note 5, at 1421.

188 Gluck et al., supra note 134, at 1847; see also David J. Barron & Elena Kagan, Chevron’s Nondelegation Doctrine, 2001 Sup. Ct. Rev. 201, 203 (2001) (“Given the difficulty of determining actual congressional intent, some version of constructive—or perhaps more frankly said, fictional—intent must operate in judicial efforts to delineate the scope of Chevron.”). It is worth noting that many of the factors used in the various cases cited above did receive some support from respondents in Gluck and Bressman’s surveys. See Gluck & Bressman, Part I, supra note 1, at 999–1004. Unfortunately, Gluck and Bressman did not ask about agency legislative history, so we do not know how congressional staff feel about agency legislative communications, although if the author’s interviews with agency staffers are correct, those communications are an integral part of the legislative process through which congressional intent is formed.

189See supra Part II.

190 Chevron U.S.A. Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 865–66 (1984); see also Cuozzo Speed Techs., LLC v. Lee, 136 S. Ct. 2131, 2142 (2016) (“But where a statute leaves a ‘gap’ or is ‘ambigu[ous],’ we typically interpret it as granting the agency leeway to enact rules that are reasonable in light of the text, nature, and purpose of the statute.”).

191 For example, where there was conflict between the agency and Congress during the legislative process, a court could apply the Skidmore standard, which means it would independently judge the quality of the agency’s interpretation where there is conflict between the agency and Congress. On the other hand, if it appeared that the agency and Congress were on the same page with respect to the legislation, then a court could grant Chevron-type deference.

192See supra Sections II.B.–II.C.

193See supra Section II.B.

194See cases cited supra note 60.

195See infra Section II.E.

196 Chevron U.S.A. Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 843–44 (1984).

197 As discussed in Section II.A., agencies also engage in many informal communications with Congress on the staff level, and it is possible that these types of communications could be more likely to be “cheap talk.” That is one reason why this Article focuses only on formal agency legislative history. These formal communications require a substantial agency drafting process and formal clearance that is likely to weed out empty rhetoric about statutory language.

198 This is different from the concern often raised by textualists about a sneaky congressional staffer being able to modify legislative history all on their own. As Justice Scalia argued, “[a]s anyone familiar with modern-day drafting of congressional committee reports is well aware, the references to the cases were inserted, at best by a committee staff member on his or her own initiative, and at worst by a committee staff member at the suggestion of a lawyer-lobbyist; and the purpose of those references was not primarily to inform the Members of Congress what the bill meant . . . but rather to influence judicial construction.” Blanchard v. Bergeron, 489 U.S. 87, 98–99 (1989) (Scalia, J., concurring in part and concurring in the judgment); see also Frank H. Easterbrook, Text, History, and Structure in Statutory Interpretation, 17 Harv. J.L. & Pub. Pol’y 61, 61 (1994) (“These clues [in legislative history] are slanted, drafted by the staff and perhaps by private interest groups.”); Gluck & Bressman, Part I, supra note 1, at 973 (showing that legislative drafters use legislative history to get “something we couldn’t get in the statute in order to make key stakeholders happy”); Antonin Scalia & John F. Manning, A Dialogue on Statutory and Constitutional Interpretation, 80 Geo. Wash. L. Rev. 1610, 1612 (2012); Starr, supra note 172 (“Lobbyists maneuver to get their clients’ opinions into the mass of legislative materials . . . .”).

199 Those working on legislation inside an agency and inside Congress would have to coordinate with regulators to generate the false legislative history to ensure that the regulators knew what to do and then they would have to work with litigators to make sure they knew to point out the false agency legislative history to a court if the agency’s interpretation was challenged. All of this would have to happen over a period of many years because of the time it takes to enact legislation and for legal challenges to arise.

200 Agencies are repeat actors full of career employees who rely on positive relationships with Congress to achieve their agency’s goals. Being uncovered as a dishonest actor would reduce the agency’s influence and potentially subject staffers and the agency to increased scrutiny from Congress. Commentators have made similar arguments about those within Congress who might attempt to use legislative history in devious ways. McNollgast, Legislative Intent: The Use of Positive Political Theory in Statutory Interpretation, 57 L. & Contemp. Probs. 3, 26 (1994) (“In practice, political actors have two routes to enforce truthfulness. First, members who prevaricate can be and occasionally are removed from gatekeeping positions, such as party leadership or committee chairs. Second, because Congress passes a very large number of legislative provisions each year, the same member is likely to be in a position of delegated authority on many occasions. To succeed in accomplishing numerous legislative objectives over a lifetime in politics, a legislator will find it valuable to develop a reputation for not taking strategic advantage when acting as an agent for other members.”); see also Katzmann, supra note 63, at 654 (“The system works because committee members and their staffs will lose influence with their colleagues as to future bills if they do not accurately represent the bills under consideration within their jurisdiction.”).

201 547 U.S. 47, 51 (2006).

202Id. at 52.

203Id. at 55–56 (“Certain law professors participating as amici, however, argue that the Government and FAIR misinterpret the statute.”).

204Id. at 54. For the agency’s proposal, see Dep’t of Def., National Defense Authorization Act for Fiscal Year 2005 (2004), http://ogc.osd.mil/olc/docs/March12-Text.pdf.

205Compare Dep’t of Def., supra note 204 (showing date legislation was sent to Congress), with H.R. Rep. No. 108-443, pt. 1, at 4 (2004) (showing date legislation was proposed in Congress).

206H.R. Rep. No. 108-443, pt. 1, at 7 (2004).

207See Hamdan v. Rumsfeld, 548 U.S. 557, 579 (2006); FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120, 125–26 (2000); Patterson v. McLean Credit Union, 491 U.S. 164, 176–78 (1989); INS v. Cardoza-Fonseca, 480 U.S. 421, 442–43 (1987); United States v. Riverside Bayview Homes, 474 U.S. 121, 137 (1985); United States v. Yermian, 468 U.S. 63, 72 (1984) (relying on congressional adoption of broadening language of false swearing statute after earlier bill was vetoed by President); Runyon v. McCrary, 427 U.S. 160, 174–75 (1976) (relying on the fact that the Senate had rejected attempts to override an interpretation as support for that interpretation); Blau v. Lehman, 368 U.S. 403, 411–12 (1962); Mont. Wilderness Ass’n v. U.S. Forest Serv., 314 F.3d 1146, 1150 (9th Cir. 2003). For a deeper discussion of the rejected proposal rule, see Eskridge, Jr. et al., supra note 3, at 831. Not all judges find the rejected proposal canon persuasive because legislative inaction can occur for a number of reasons that are unclear to courts. See, e.g., Rapanos v. United States, 547 U.S. 715, 750 (2006) (“Congress takes no governmental action except by legislation. What the dissent refers to as ‘Congress’[s] deliberate acquiescence’ should more appropriately be called Congress’s failure to express any opinion.”); Solid Waste Agency of N. Cook Cty. v. U.S. Army Corps of Eng’rs, 531 U.S. 159, 170 (2001) (“A bill can be proposed for any number of reasons, and it can be rejected for just as many others. The relationship between the actions and inactions of the 95th Congress and the intent of the 92d Congress in passing Section 404(a) is also considerably attenuated.”); Cent. Bank of Denver, N.A. v. First Interstate Bank of Denver, N.A., 511 U.S. 164, 187 (1994) (“[F]ailed legislative proposals are ‘a particularly dangerous ground on which to rest an interpretation of a prior statute.’”); Pattern Makers’ League of N. Am. v. NLRB, 473 U.S. 95, 110–12 (1985) (finding the legislative history too ambiguous to influence the Court’s decision). Some scholars have also been critical of the rule. See, e.g., William N. Eskridge, Jr., Interpreting Legislative Inaction, 87 Mich. L. Rev. 67, 84–95 (1988) (collecting cases using and declining to use the rejected proposal rule). These challenges could also apply to rejected agency legislative proposals, because agencies make many legislative proposals and it is hard to know why any particular proposal is rejected. Not every rejected proposal is significant, but many are, especially when Congress proceeds to pass legislation on the same topic.

208 On the other hand, recall the Cuozzo case, described above, where the Court relied on Chevron to defer to the USPTO. See supra Section II.B. In that case, Chevron’s presumption made sense in light of the agency legislative history. There was ample evidence of a cooperative relationship between the USPTO and Congress and an intent to delegate to the USPTO. Although the agency legislative history would not have changed the Court’s decision, it would have strengthened it by situating it within the empirical reality of how the bill came to be rather than an assumption that may or may not have been true. See supra Section II.B.

209See supra Part III.

210 One concrete way in which courts have attempted to approach Chevron on a case-by-case basis is through the application of the major questions doctrine. This doctrine presumes that Congress does not intend to delegate issues of major importance to agencies without explicitly doing so, and instead keeps the interpretation of those provisions to the court. See Kevin O. Leske, Major Questions About the “Major Questions” Doctrine, 5 Mich. J. Envtl. & Admin. L. 479, 480 (2016). This doctrine was an important part of the majority’s decision in King v. Burwell, which stated that it was inconceivable that Congress would have left a question of such “deep ‘economic and political significance’” to an agency without expressly doing so. 135 S. Ct. 2480, 2489 (2015). While the wisdom of allowing courts to determine whether an issue is “major” enough to do away with deference is certainly debatable, if courts are to engage in this type of inquiry then they would be better served by considering agency legislative history to determine whether such an issue had actually been contemplated. Perhaps the agency legislative history contains a discussion of the major issue or perhaps the agency pointed out the major issue and Congress still chose not to amend the legislation. Before a court decides that an issue is too major for Congress to have left it to an agency, it could check the agency legislative history to confirm that Congress in fact had not considered the issue, in much the same way courts check congressional legislative history before applying the absurdity doctrine to make sure that the absurdity was not intentional.

211 The Court has accepted evidence of statutory purpose, “including those revealed in part by legislative and regulatory history,” to determine whether a statute is ambiguous at Chevron step one. See City of Arlington v. FCC, 569 U.S. 290, 310 (2013); see also Brown & Williamson Tobacco Corp., 529 U.S. at 143–47.

212Cf. Strauss, Chevron Space” and “Skidmore Weight”, supra note 26, at 1163 (arguing that determinations of whether an agency interpretation is “reasonable” might consider whether the agency’s interpretation is consistent with statutory purposes). Courts have disagreed about whether to consult legislative history at step one or step two of Chevron. Compare Halverson v. Slater, 129 F.3d 180, 184 (D.C. Cir. 1997) (holding that “the lower court erred by failing to ‘exhaust the traditional tools of statutory construction,” including legislative history, at Chevron step one), with Cohen v. JP Morgan Chase & Co., 498 F.3d 111, 122 (2d Cir. 2007) (“This court has generally been reluctant to employ legislative history at step one of Chevron analysis, mindful that the ‘interpretive clues’ to be found in such history will rarely speak with sufficient clarity to permit us to conclude ‘beyond reasonable doubt’ that Congress has directly spoken to the precise question at issue.”). The distinction is unlikely to matter in whether a court upholds an agency interpretation.

213 Courts would still need to distinguish reliable from unreliable agency legislative history in much the same way they already do for congressional legislative history.

214 Congressional staff confirm that legislative history is often written with agencies as the intended audience. See Gluck & Bressman, Part II, supra note 5, at 789–90.

215 Manning, supra note 34, at 1534.

216See supra Sections II.E.1.–II.E.2.

217See supra Sections II.E.1.–II.E.2.

218See supra Part II.

219See William N. Eskridge, Jr. & Lauren E. Baer, The Continuum of Deference: Supreme Court Treatment of Agency Statutory Interpretations from Chevron to Hamdan, 96 Geo. L.J. 1083, 1136 (2008) (collecting statistics on the use of legislative history in Chevron cases).

220 The conventional wisdom is that committee and conference reports are by far the most reliable legislative history, followed by on-the-record statements by the members of Congress who sponsored or supported the bill. Eskridge, Jr. et al., supra note 3, at 631; Eskridge, Jr., supra note 4, at 221–22; Robert A. Katzmann, Judging Statutes 35–54 (2014); Manning & Stephenson, supra note 3, at 136 (“The conventional wisdom has been that the most reliable form of legislative history consists of the reports prepared by the House and Senate committees, which accompany bills favorably reported to the chamber, and the conference committee reports which accompany the reconciled version of the House and Senate bills.”); Costello, supra note 1, at 43–50 (discussing the importance of conference committee reports). For example, Justice Jackson objected to the use of legislative history generally but made an exception for committee reports. See Schwegmann Bros. v. Calvert Distillers Corp., 341 U.S. 384, 395 (1951) (Jackson, J., concurring). As Justice Frankfurter remarked, “[w]hatever we may think about the loose use of legislative history, it has never been questioned that reports of committees and utterances of those in charge of legislation constitute authoritative exposition of the meaning of legislation.” Orloff v. Willoughby, 345 U.S. 83, 98 (1953) (Frankfurter, J., dissenting). Quantitative analyses confirm that committee reports are relied on far more than other kinds of legislative history. See, e.g., James J. Brudney & Corey Ditslear, The Warp and Woof of Statutory Interpretation: Comparing Supreme Court Approaches in Tax Law and Workplace Law, 58 Duke L.J. 1231, 1262 (2009) (discussing the prevalence of committee reports used in tax law and workplace law cases); Jorge L. Carro & Andrew R. Brann, The U.S. Supreme Court and the Use of Legislative Histories: A Statistical Analysis, 22 Jurimetrics J. 294, 304 (1982) (reporting that, over the forty year period analyzed by the authors, over 60% of the Supreme Court’s citations to legislative history were references to committee reports).

221See Shobe, Intertemporal Statutory Interpretation, supra note 5, at 861–65; Abbe R. Gluck, Congress, Statutory Interpretation, and the Failure of Formalism: The CBO Canon and Other Ways that Courts Can Improve on What They Are Already Trying to Do, 84 U. Chi. L. Rev. 177, 208–09 (2017) [hereinafter Gluck, Congress] (describing the Gluck and Bressman study and its findings that congressional staffers “make the policy and sketch out statutory contours, often in the form of policy ‘bullet points.’ The nonpartisan Legislative Counsels then take over and turn the ‘asks’ into statute-ese”).

222See supra Part II.

223See Gluck, Congress, supra note 221, at 209 (“The section-by-section is drafted by committee staff (the policy staff, not the Legislative Counsels) . . . .”).

224 Blanchard v. Bergeron, 489 U.S. 87, 98–99 (1989) (Scalia, J., concurring in part and concurring in the judgment) (“As anyone familiar with modern-day drafting of congressional committee reports is well aware, the references to the cases were inserted, at best by a committee staff member on his or her own initiative, and at worst by a committee staff member at the suggestion of a lawyer-lobbyist; and the purpose of those references was not primarily to inform the Members of Congress what the bill meant . . . but rather to influence judicial construction.”).

225 Strauss, When the Judge Is Not the Primary, supra note 156, at 329–35. Professors Gluck and Bressman’s respondents almost uniformly agreed with the idea that one purpose of legislative history is to shape the way agencies interpret statutory ambiguities. Gluck & Bressman, Part II, supra note 5, at 768. Congress may adopt agency legislative history as a tool of oversight that is contemporaneous with the enactment of the statute, unlike other means of post-enactment oversight like congressional hearings. For a discussion of tools of congressional oversight, see Mathew D. McCubbins et al., Structure and Process, Politics and Policy: Administrative Arrangements and the Political Control of Agencies, 75 Va. L. Rev. 431, 432–45 (1989).