Agency Legislative Fixes

I.     Introduction

At the end of 2017, the Republican party celebrated the passage of the sweeping tax reform generally known as the Tax Cuts and Jobs Act (“TCJA”).1 House Speaker Paul Ryan declared that it had been a “promise made and [a] promise kept.”2 The next day, President Trump announced “[i]t’s always a lot of fun when you win.”3 This was the glory (for some) of the legislative process coming to fruition.

While the news coverage largely moved on after legislative passage, the story was far from over. In the aftermath of the TCJA came the messy reality of enacted legislation. Part of the way Congress passed the TCJA notwithstanding united Democratic opposition was by doing so at breakneck speed and utilizing a variety of unprecedented procedures to pass major tax reform.4 The predictable result was that many mistakes had been made in drafting it. Congress had inadvertently eliminated the ability of sexual abuse victims to deduct their attorney fees,5 taxed low-income college students like trust fund babies,6 and imposed steep tax hikes on Gold Star Families,7 among many other cringe-worthy errors.8

This ping-ponging between difficult yet exuberant legislative passage and subsequent realization of critical drafting errors is far from isolated to the recent TCJA. In 2015, the fate of the Affordable Care Act (“ACA”), President Obama’s signature legislative achievement, famously hung in the balance, turning on whether the Supreme Court would fix four words of the voluminous statute. If read literally, the four words, which seemed like a simple legislative drafting mistake,9 threatened to unravel insurance markets essential to the entire legislative scheme.10 Major environmental issues, such as whether California can set the standard for certain emissions for all other states, have turned on whether or not the statute omitted words.11 And the functionality of the Dodd–Frank whistleblower regime has turned on a potential, failed cross-reference in the statute.12 A list of similar examples could go on and on.13

The explanation for such drafting errors is quite similar to the explanation in the case of the TCJA. In an era of hyper-partisanship, political parties are turning to extraordinary procedures to push through legislative agendas without bipartisan support.14 These procedures, while a potentially useful way to break through gridlock, have eroded the traditional order of legislation, control of subject matter experts, and the quality of the legislative product. The result is that today’s legislation is highly vulnerable to errors. Making matters even worse, the use of extraordinary procedural methods to break through gridlock and get legislation passed often means that it is more difficult for Congress to fix the mistakes than it was to pass them in the first instance.15

So, what should be done about such mistakes? In examining this question, scholars have typically focused on the judiciary: to what extent will courts correct legislative drafting mistakes or allow agencies to do so?16 The answer has been: not much. Courts tread very carefully in openly acknowledging that they are fixing any such mistakes or allowing agencies to do so. Indeed, as an illustration of this phenomenon, in two of the three cases mentioned above—the preemption of state emission standards,17 and the Dodd–Frank whistleblower regime18—the courts rejected agencies’ attempts to correct what they viewed to be errors, or omissions, in the statute.19 In the case of the ACA, the Supreme Court famously saved the ACA by essentially reading the mistake out of the statute, although, in doing so, it refused to defer to the agency’s interpretation of the statute or acknowledge that it was correcting a mistake.20 The lengths to which the Court went to avoid acknowledging that it was essentially correcting a legislative drafting mistake or allowing an agency to do so underscores the reticence of courts to openly acknowledge they are correcting mistakes.21

Scholars’ focus on judicial approaches overlooks a crucial, less-noticed reality. Agencies are quietly going about fixing legislative drafting mistakes, or engaging in what this Article calls “agency legislative fixes,” irrespective of what courts say. Agencies routinely use informal guidance to fix legislative drafting mistakes, or otherwise fix such mistakes in a way that is unlikely to be challenged by members of Congress. Indeed, in the case of the TCJA, after Congress sputtered in its attempts to fix its own legislative drafting mistakes in a timely fashion, the Treasury Department fixed some of the mistakes, while leaving others in place.22 Many members of Congress even appealed to Treasury to fix Congress’ legislative drafting mistakes.23 Yet, existing legal theory offers no framework for evaluating this phenomenon.

This Article fills this gap by shining a light on agency legislative fixes and the thorny normative considerations they raise. Agency legislative fixes implicate fundamental constitutional and democratic values such as legislative supremacy and agency legitimacy. In terms of legislative supremacy, agency legislative fixes undermine the formal allocation of all lawmaking power to Congress. A more functional perspective yields a more nuanced analysis, which suggests that agency legislative fixes paradoxically may present a supremacy tradeoff. Under this tradeoff, agency legislative fixes may undermine Congress’ formal role as lawmaker, but also encourage more decisions to be made at the congressional level. However, this tradeoff is contingent and ultimately may not justify the formal infringement on Congress’ power. Agency legislative fixes also cannot clearly be justified as a legitimate exercise of agency power. In some ways, it may be possible to view agency legislative fixes as a legitimate exercise of agency expertise or, alternatively, agency authority to fill in ambiguity in statutory schemes. But these possibilities turn on deep, contested questions about agency legitimacy, as well as methods of statutory interpretation.

There are some practical arguments in favor of agency legislative fixes, including concerns about the relationship between agencies and the public. But there are also practical problems with agency legislative fixes, including encouraging problematic agency guidance practices and systematically biasing the law in a pro-regulated party fashion. In any event, it is not clear that any practical case for agency legislative fixes could outweigh the unresolved constitutional and democratic problems.

In light of the normative problems with agency legislative fixes, the Article concludes by evaluating potential paths forward. The easiest option would be to stick Congress with its own mistakes, forcing it to bear the costs of not making legislative fixes. But this option would impose significant costs on both agencies and the public. More proactively, the first-best solution would be for Congress to clean up its act by changing its legislative drafting practices to make fewer mistakes in the first instance. While ideal, this possibility also unrealistically assumes away the very disfunctions in legislative practices that make agency legislative fixes necessary.

This Article argues for a more plausible second-best solution: changing the baseline for measuring eligibility for reconciliation legislation to allow Congress to fix its own mistakes pursuant to the same vote count that was necessary to make them. This fix has the potential to ameliorate the serious catch-22 presented by current legislative drafting practices.

For those agency legislative fixes that will no doubt continue to some extent, agencies can make them more transparent and systematic. These changes at the agency level, without any changes in congressional practice, may promote some rule of law values, but at the cost of regularizing infringements on congressional power. This suggests that there may be some value in Congress being more explicit about when agencies have authority to correct legislative drafting mistakes.

The Article concludes that there are no easy ways to resolve the difficulties presented by agency legislative fixes. Agency legislative fixes are a pervasive phenomenon that responds to entrenched legislative problems. Exploring them, and the values undermined or vindicated by different paths forward, is essential to improving our system of governance. In the process of doing so, this Article advances conversations in a variety of recent literatures, including literatures regarding legislative drafting mistakes,24 congressional inaction,25 and the realities of legislation and regulation.26

This Article proceeds as follows. Part II describes both the conditions that systematically create legislative drafting mistakes and general judicial disapproval of non-congressional fixes of such mistakes. Part III explores how agencies can use informal guidance and other tactics strategically to offer agency legislative fixes, notwithstanding judicial disapproval. Part IV explores agency legislative fixes from a normative perspective. Part V examines potential paths forward. Part VI then offers a brief conclusion.

II.     Legislative Dysfunction Paired with Judicial Aversion to
Non-Congressional Fixes

Agency legislative fixes occur because, in the aftermath of legislation, agencies often find themselves stuck between a rock (legislative dysfunction) and a hard place (judicial aversion to non-congressional legislative fixes). This Part explains how current legislative practices are increasingly producing legislative drafting mistakes that Congress is unable, itself, to fix, and how courts are averse to non-congressional legislative fixes.

A.     Legislative Drafting Mistakes

A legislative drafting mistake exists when a drafting error results in the legislative text failing to carry out Congress’ intent. The easiest way to understand this category is to define it narrowly. For instance, in the context of judicial fixes, Justice Scalia distinguished “mistake[s] of expression,” which may be seen as synonymous with the category of legislative drafting mistakes, from mistakes of “legislative wisdom,” which may not.27

While it is easy to provide this definition of legislative drafting mistakes, it is much harder to identify them and, in some ways, they are in the eye of the beholder. Mistakes in policy, for instance, seemingly should not be classified as mere legislative drafting mistakes because, while perhaps ill-decided, mistakes in policy reflect Congress’ intent at the time of drafting the statute. But how to draw the line between what is a mere mistake in drafting (which presumably failed to reflect Congress’ intent at the time of drafting) and what is a mistake in policy is not always entirely clear. Some mistakes are obviously mere drafting errors, because the existence of the mistakes makes the statute absurd or nonsensical.28 But, it is equally possible for a statute to make sense, but for its text to still fail to reflect Congress’ intent. For instance, by accidentally omitting a word (such as “not”), the statute may reach a sensical outcome, even if it is the exact opposite of what Congress intended. In situations in which legislative history or purpose are clear, such a mistake may be clear to interpreters willing to look to such sources. But many interpreters may not be willing to look to such sources to interpret the text, and, in any event, legislative history and legislative purpose may not always be available. This may be especially true in an era of frenzied legislative passage.29 The fact that there is no easy, or agreed-upon, way to identify a legislative drafting mistake makes it more difficult for courts and agencies to act appropriately or uniformly when Congress does not fix its own mistakes.

Legislative drafting mistakes can come from a variety of sources. Least controversially, many would agree that typos constitute legislative drafting mistakes. Many would also agree that legislative drafting mistakes include failures to appreciate how interlinkages between different provisions produce a result that conflicts with what Congress intended. But there also has to be some limit to what counts as a legislative drafting mistake.

B.     Hyper-Partisanship and Non-Textbook Legislative Process

In textbook versions of the legislative process, legislation flows through Congress in an orderly and deliberative fashion.30 Such legislation would originate in either the House of Representatives or the Senate.31 In whichever chamber it originates, the legislation would be referred to a committee that is a subject matter expert.32 After the committee finalized the proposed legislation, it would bring the legislation to the rest of the chamber for a vote.33 The chamber would defer at least to some extent to the committee’s expertise, and the committee ideally would have sorted out how to get the chamber’s approval of the proposed legislation. The chamber would then vote on the proposed legislation without making significant amendments. If the chamber voted to approve the legislation, the legislation would then pass to the other chamber, which would replicate the same process using its own expert committee.34 After the second chamber voted on the legislation, a conference committee would reconcile the legislation, if need be.35 The conference committee would be staffed by both Congresspeople and professional staffers with subject matter expertise, and the work they produced, by and large, would be the proposal considered by the entire Congress.36 Congress would then vote on the product of the conference committee which, if approved, would pass to the President to sign or veto.37

Today’s legislative process deviates significantly from this textbook portrayal. There are many reasons for such deviations, but hyper-partisanship is a particularly important one. Political scientist Barbara Sinclair has described that “[i]ntense partisan polarization is the single most salient characteristic of contemporary politics and one that increasingly shapes the legislative process.”38 Hyper-partisanship has become both a cause and an effect of congressional dysfunction. As a cause, hyper-partisanship contributes to gridlock, which makes it difficult to pass legislation through ordinary procedures.39 As an effect, the political parties have increasingly turned to partisan means of passing legislation. Party leadership and their staffers, as well as the President, are now often involved in not only shepherding legislation through Congress, but also promoting their own versions of legislation and entertaining suggestions from members to make changes.40 As a result, there is much less deference to expert committees. Legislation is often passed in ways that bypass committees entirely.41 The deterioration of expert committee control has also made post-committee amendments more common, especially in passing major legislation.42 The rise of post-committee amendments means members may make a whole slew of rapid-fire changes after committee approval of the legislation. There will in many cases be little time or inclination to consider how such amendments fit within the scheme of the new proposed legislation, much less the entire existing legislative scheme.43

There may be some reasons to cheer some of these developments. As some scholars have pointed out, the Constitution does not require Congress to use any specific procedures.44 In some ways, Congress’ ability to pass legislation in an age of hyper-partisanship may be seen as a testament to Congress’ ability to get things done.45 Furthermore, the deterioration of committee control may arguably allow for more diffuse participation by outside individuals and groups, which might bring broader perspectives to bear.46

But it is also clear that these changes have costs. Much less control by subject-matter experts (both at the congressional member and staffer level) may reduce the legislation’s quality.47 Party leadership and their political staffers often have less of an eye toward the integrity of the legislation than committee members and nonpartisan staffers who are subject-matter experts.48 And legislating outside of the textbook process, such as through post-committee amendments and extreme legislative bundling, is also more likely to be chaotic, with inputs coming from many different sources, and a lower ability for a single coordinator (or multiple coordinators) to sort through the inputs carefully. Short timelines exacerbate these difficulties. The result is that, relative to legislation passed through the textbook process, today’s legislation is less likely to cohere in sensible ways, and specific provisions may be drafted erroneously in ways that fail to carry out Congress’ intent.49

This is not to say that legislation passed in a slower, more deliberative, bipartisan fashion would not also have its share of significant problems. Mistakes may be inevitable in any major piece of legislation.50 But the combination of increasingly complex legislation being passed in a rushed fashion, with decreased expert control, often through the use of chaotic congressional procedures, exacerbates legislative drafting mistakes.

C.     Congress’ Increased Use of Reconciliation Makes it More Difficult to Fix Legislative Drafting Mistakes

Not only is Congress more likely to make mistakes in today’s legislative process, but it is also harder for Congress to fix mistakes. This is the underappreciated result of Congress using special congressional procedures to break through partisan gridlock.

A particularly important procedure in this regard has been Congress’ increasing use of reconciliation. Reconciliation was a procedure created by the Congressional Budget and Impoundment Control Act of 1974 (“Budget Act”).51 Prior to the Budget Act, Congress had no official procedure to bring all of its revenue and spending decisions together in a comprehensive budget.52 The Budget Act provided Congress the ability to do so.53 The Budget Act requires Congress to first pass a budget resolution.54 While not itself law, the budget resolution creates a framework that enables Congress to pass laws in a manner consistent with the budget resolution.55 After passing a budget resolution, Congress can then direct congressional committees to pass reconciliation legislation to bring revenues and direct spending in line with the budget resolution.56 To ensure that pressing budget legislation can be passed expeditiously, the Budget Act made reconciliation legislation immune from the Senate filibuster, as well as certain other legislative obstacles.57 This effectively means that reconciliation legislation is subject to a majority, rather than supermajority, vote requirement in the Senate.

The lower required vote count prompted Congress to quickly begin using reconciliation to pass legislation in a more expansive fashion than originally intended when Congress passed the Budget Act.58 This more expansive use of reconciliation, in turn, led to congressional concerns that the reconciliation process was being manipulated to pass non-budget legislation. In 1985, Senator Byrd explained that “we are in the process now of seeing . . . the Pandora’s box which has been opened to the abuse of the reconciliation process. That process was never meant to be used as it is being used.”59

In order to preserve the efficacy of reconciliation and the integrity of congressional procedures generally, Senator Byrd offered an amendment to the Consolidated Omnibus Budget Reconciliation Act of 1985, which was designed to limit reconciliation to its original budgeting purposes.60 The so-called Byrd rule allows Senators to raise “points of order,” objecting to provisions they believe are extraneous.61 If an objection is sustained, the provision will be dropped from the bill (in what is often colloquially referred to as Byrd droppings),62 or otherwise stricken from the bill.63 Over time, the Byrd rule was expanded, made permanent, and codified. In its present form, the Byrd rule defines as extraneous any provision that, among other things “does not produce a change in outlays or revenues . . . [or that] produces changes in outlays or revenues which are merely incidental to the non-budgetary components of the provision.”64

However, the Byrd rule has not functioned well over time and has proven easily manipulated.65 Since passage of the Byrd rule, reconciliation has been used repeatedly to pass major legislation, oftentimes in ways that is difficult to square with its intended, original, limited scope.66 This has reached somewhat of a zenith in recent years. The Obama administration used reconciliation to pass the ACA—major, transformative health care legislation.67 The Trump administration then used reconciliation to pass the TJCA—a massive overhaul of the tax code.68 Both uses of reconciliation go well beyond short-term budgeting adjustments.

The fact that the Byrd rule is highly manipulable does not mean it has no bite at all. It does shape what legislation can go through reconciliation, even if it does so in unsatisfying ways. The process for determining whether a provision is “extraneous” within the meaning of the Byrd rule is complicated and opaque.69 According to the Byrd rule, the Presiding Officer of the Senate makes determinations about whether a provision is extraneous.70 Since the question turns on how the proposed legislation affects the budget, the Presiding Officer has to assess what the budget score (or impact) for the proposed legislation will be. In practice, the Presiding Officer turns to the Budget Committee Chair, who, by historical practice, turns to the Congressional Budget Office (“CBO”) and Joint Committee on Taxation (“JCT”), to determine whether the legislation follows the reconciliation instructions and whether it increases the deficit beyond the budget window.71 The Presiding Officer turns to the Senate Parliamentarian for other Byrd rule points of order, including whether the legislation produces a non-incidental change in outlays or revenues.72 The Senate Parliamentarian, a somewhat obscure figure deep inside the Senate, makes these decisions largely out of public view based on criteria that are not entirely apparent.73 Senator Judd Gregg, the one-time top Republican on the Budget Committee, described the Senate Parliamentarian as “the defense counsel, he’s the prosecution, he’s the judge, he’s the jury and he’s the hangman.”74

The increasing use of reconciliation to pass major legislation, and the accompanying limitations of the Byrd rule, have had perverse and surprising impacts on Congress’ ability to correct its own legislative drafting mistakes. To fix legislation through reconciliation, the fix itself must meet reconciliation requirements, including the Byrd rule.75 This means that the corrective legislation, among other things, must have a non-incidental impact on revenues or outlays. Making this determination depends on whether the fix impacts revenue or outlays relative to the original legislation, which, after its passage, represents the baseline for evaluating future changes. A critical question in making this evaluation is whether the baseline should be the original legislation as it was actually enacted, or the original legislation as it was erroneously scored when it was enacted. This is an issue because, when a mistake was made in the original legislation, the budget score for the original legislation often reflects Congress’ intent in passing the legislation, not the mistake that was made in passing it, because even budget scorers often do not realize that drafting mistakes have been made until after the scoring.76

For instance, imagine an original piece of legislation provides a new tax credit for taxpayers who provide financial support to qualifying relatives.77 Among other things, qualifying relatives must have income that falls below an exemption amount.78 “Exemption amount” is defined elsewhere, in an existing section of the tax code.79 Unrelatedly, in the same, original legislation, Congress temporarily reduces the “exemption amount” to zero.80 In enacting the legislation, neither Congress nor the budget scorers pick up on how the latter change would affect the newly enacted qualifying relative tax credit. This yields the unintended consequence that, as a technical matter, Congress has just inadvertently created a new tax credit for which essentially no individuals will qualify. But because the scorers, like the drafters of the original legislation, did not recognize this at the time of enactment, the score for the original legislation took into account the full cost for the qualifying relatives tax credit as Congress intended for it to operate.

If Congress wants to pass new legislation to fix the mistake made in the original legislation, should the baseline for determining the cost of the new legislation be the legislation as enacted (in which case the new legislation affects revenue because, as a matter of law, essentially no one is actually eligible for the tax credit)? Or should the baseline be the legislation as it was erroneously scored (in which case the new legislation would not affect revenue because the score for the original legislation reflected the cost of this new tax credit)? This seemingly technical question has extraordinary importance. In an era in which Congress increasingly passes major, flawed legislation through reconciliation, how this question is answered plays a significant role in determining Congress’ ability to fix its own mistakes.

Notwithstanding the importance of the question, there is surprisingly little attention that has been paid to it. There are no clear rules for how the Byrd rule should be applied in this case, though current practice is to use the original legislation as it was erroneously scored as a baseline.81 As applied to the example above, the baseline is the full cost for the qualifying relatives tax credit, as Congress had intended for it to operate. But this yields a conundrum: Relative to the baseline, fixing the mistake would not impact revenue, because the erroneous score assumed the fix Congress now wants to make.82 Current practice thus treats fixes of legislative drafting mistakes as ineligible for reconciliation.83 As a result, Congress cannot fix in reconciliation mistakes that apply as a matter of law.

By itself, this conundrum would not necessarily thwart attempts to fix the legislative drafting mistakes. In theory, Congress could just pass legislation through the regular legislative process (i.e., outside of reconciliation) to fix the flaws in the original legislation. But, as a practical matter, the majority party often does not have the votes to fix the legislation through the regular legislative process. The majority party likely used reconciliation to begin with because it lacked the votes necessary to push the legislation through the regular process. Even to the extent the minority party agrees the fixes would be a net improvement, the minority party is often unwilling to participate in fixing mistakes because the majority originally passed the legislation in a hyper-partisan fashion.84 Without the ability to correct the legislation it passed through reconciliation, and without the assent of the minority party, the majority party may be unable to fix the drafting mistakes. This has turned the correction of legislation—previously thought to be a routine part of the legislative process85—into a surprisingly difficult task. This is not to say that corrections can never happen.86 The critical point is that Congress’ increasingly expansive use of reconciliation in hyper-partisan times is systematically making it more difficult for Congress to fix the mistakes that it is increasingly likely to make.

This set of dynamics strikingly occurred in the signature legislative achievements from the current and prior presidential administrations—the TCJA and the ACA. In the TCJA, the lack of any Democratic votes meant that the Republican Congress had to pass the original legislation through reconciliation.87 Democrats in part passed the ACA through reconciliation without any Republican votes.88 In both cases, Congress made significant legislative drafting mistakes. In the context of the TCJA, commentators described the massive tax reform as littered with glitches and errors as a result of the rushed legislative process.89 In the context of the ACA, the haphazard procedures used to secure legislative passage resulted in a number of critical mistakes, including one that threatened to imperil the entire legislative scheme.90 Yet, in both cases, as a result of the application of the Byrd rule and the scoring dynamics outlined above, the mistakes could not be fixed through reconciliation. In each case, the party excluded from the original process—Democrats in the case of the TCJA and Republicans in the case of the ACA—refused to participate in fixing the mistakes made in legislative passage in a timely fashion.91 In the case of the TCJA, the minority party, the Democrats, belatedly began suggesting that they may join in a technical corrections bill.92 However, they suggested this well after tax return filing season was already underway, and with no real promise in sight as to the extent of corrections that may occur.93 In both cases, the combination of hyper-partisanship and reconciliation to pass the original legislation, along with the inability to use reconciliation on the back-end to fix mistakes, left Congress surprisingly impotent to correct its own drafting mistakes.

D.     Judicial Aversion to Non-Congressional Fixes

Courts are typically willing to openly acknowledge they are correcting legislative drafting mistakes only in very narrow situations, in which there is a clear case of typographical error. For instance, in U.S. National Bank of Oregon, the Supreme Court was willing to correct what it believed to be an error in the placement of a quotation mark, explaining that “[a]gainst the overwhelming evidence from the structure, language, and subject matter of the 1916 Act there stands only the evidence from the Act’s punctuation, too weak to trump the rest.”94 When courts do correct these types of clear typographical errors, they often employ what is called the “scrivener’s error” doctrine. In dissent in X-Citement Video, Justice Scalia emphasized the narrowness of this doctrine, explaining that “the sine qua non of any ‘scrivener’s error’ doctrine . . . is that the meaning genuinely intended but inadequately expressed must be absolutely clear.”95

Sometimes courts’ willingness to correct mistakes is dependent on the determination that failing to correct a mistake would lead to absurd outcomes, leading to a blurring of the “scrivener’s error doctrine” with the “absurdity doctrine.”96 For instance, in Bock Laundry, the Supreme Court examined a Federal Rule of Evidence that, if read verbatim, compelled admission of evidence of a prior crime that was beneficial to civil defendants, but not to civil plaintiffs.97 The majority rejected this “odd result,” arguing that the Federal Rule “can’t mean what it says.”98 Justice Scalia concurred on the basis that the statute, “if interpreted literally, produces an absurd, and perhaps unconstitutional, result.”99

Recent cases have exhibited how courts’ own discomfort with fixing legislative drafting mistakes may leave very little leeway for courts to approve of agency fixes. First, in King, the case in which the Supreme Court decided the fate of the ACA in 2015,100 the Court displayed that, even when it might be willing to correct a mistake, it was not necessarily willing to do so openly. The ACA depended on a number of interlocking provisions, including the provision of tax credits to make insurance more affordable.101 In order for the statutory scheme to work, the tax credits had to be available to all eligible individuals whether they enrolled in insurance through a federal or state exchange.102 But the statute only authorized such credits for individuals who enrolled in insurance through an “[e]xchange established by the State.”103 It was pretty clear that this problematic language failed to carry out Congress’ intent but it was unclear whether the Court would be willing to treat the language as a mistake and, crucially, be willing to fix it.

The Supreme Court, in an opinion authored by Chief Justice Roberts, concluded that the best interpretation of the statute authorized tax credits for individuals who enrolled through an exchange established by states or by the federal government.104 In other words, the Court reached the conclusion that was consistent with the finding of a mistake. But in doing so, the Court assiduously avoided saying it was correcting a mistake.105 Rather, the Court relied on what many argued was a strained statutory interpretation to reach the same outcome.106 The Court also explicitly refused to delegate to the agency, the IRS (which had released a regulation correcting the arguable mistake), concluding that the extraordinarily important nature of the question indicated Congress had not intended to delegate it to the IRS.107

The Court thus both simultaneously seemed to correct the mistake and to leave unsettled the continued treatment of legislative drafting mistakes. The fact that the Court reached so hard to avoid deciding the issue on the grounds of mistake suggested discomfort with openly correcting legislative drafting mistakes.108 The fact that the Court refused to defer to the IRS made even less clear when agencies could correct mistakes. And yet, the fact that the Court reached an outcome consistent with correcting Congress’ likely mistake suggested that perhaps the Court had opened the door on courts correcting legislative drafting mistakes, if only just a little.

The Supreme Court has since signaled that King should not be read to open the door wider on courts correcting mistakes or allowing agencies to do so, at least in anything but an extraordinary case. In Digital Realty Trust v. Somers,109 the Court had to sort through the Dodd–Frank Act’s whistleblower and retaliation protection provisions. Somewhat curiously, the whistleblower definition under Dodd­–Frank only included whistleblowers who report to the SEC (as opposed to reporting internally within an organization), though the retaliation provisions protected whistleblowers who reported internally.110 This created somewhat of a statutory conundrum, whereby individuals were supposed to receive protection for reporting internally (which many employees are actually required to do prior to making any external reports), but they would not actually qualify as whistleblowers, and hence would not technically be eligible for such retaliation protections, unless they reported to the SEC.111 The oddity of the narrower whistleblower definition in light of the broader retaliation protections, and the seeming need for such broader protections in cases in which employees are required to report internally first, led the SEC to issue a broader regulation that included internal reporters as whistleblowers for the purposes of Dodd–Frank.112 Some courts of appeals upheld the SEC’s regulation,113 whereas the Fifth Circuit struck it down.114 The nature of the case115 seemed, at least in the view of some, to place King’s legacy in play.116

The Supreme Court unanimously found “the statute’s definition of ‘whistleblower’ clear and conclusive” and thus struck down the SEC’s regulation as being in conflict with a clear statute.117 While there were certainly arguments that the statutory provision at issue in Somers was not, in fact, a mistake,118 the Court’s decision suggested that it generally did not intend to be in the business of looking too hard at statutes to determine whether a mistake existed, or defer to agencies that had. The Court emphasized that, as long as the language of the statute was clear, it was binding on both the agency and the Court,119 a discussion that sat at least a bit uncomfortably with King.120 Some commentators celebrated what they viewed as the Court’s retreat from an inclination to fix legislative drafting mistakes.121 For instance, Paul Mahoney expressed that the fact that King could now be understood as an anomaly “is all to the good” because “[t]he Court should refuse to play along when Congress cannot enact legislative language sufficient to achieve what it wants.”122 The upshot is that courts’ willingness to facilitate mistake correction has been quite limited. Courts are sometimes comfortable correcting fairly obvious typographical errors but are often wary of deferring to or approving of agencies’ attempts to correct legislative drafting mistakes.

III.     Agency Legislative Fixes

Despite the Court’s failure to endorse agency legislative fixes, agencies often do fix legislative drafting mistakes. This Part explores how agencies can create legislative fixes through the use of informal guidance and other tactics and offers an example and counterexample of this phenomenon through the most recent major legislation, the highly-flawed TCJA. These examples reveal underexplored features of agency legislative fixes, including post-legislative attempts by Congress to influence such fixes.

A.     The Dynamics of Agency Legislative Fixes

When an agency becomes responsible for implementing flawed legislation, Congress members have both formal and informal means of influencing the agency’s implementation. First, Congress has formal channels of control over agencies including appropriations power, the ability to require agency reporting, conducting congressional investigations, the ability to confirm high-level officials, and the like.123 These formal mechanisms of control are paired with less studied, but perhaps equally important, informal mechanisms such as backchannel contacts between Congress members and agency officials.124 Additionally, congressional staffers may also participate in the administrative process to ensure administrative guidance reflects the intent of those who passed the legislation. For instance, House Ways and Means staffers acknowledged working with Treasury in crafting its post-TCJA guidance.125

Notwithstanding Congress members’ (or their staffers’) and agency officials’ coordinated efforts, agencies face obstacles when the correct implementation, according to Congress members or agency officials, would be inconsistent with the text of the statute. Namely, to the extent that an agency provides guidance inconsistent with the statute, if the guidance disadvantages a constituency with standing to sue, the agency risks its action being invalidated by courts under the judicial approaches discussed above.126

Agencies can get around this difficulty. If agencies issue guidance that benefits regulated parties, there may not be anyone with both the incentive and standing to sue the agency. Regulated parties who benefit from the guidance generally will not have the incentive to sue.127 And, because of the way standing rules work, other parties that are not directly regulated by the legislation, but rather only stand to benefit indirectly, generally do not have standing to sue for the agency’s failure to implement the legislation.128 Moreover, if the agency avoids issuing the guidance in official regulations, instead opting for less formal guidance (such as notices, press releases, and the like), the agency makes it even less likely that anyone will be able to sue, as such guidance often fails to meet judicial “ripeness” requirements.129 Thus, through informal guidance that benefits regulated parties, agency officials may fix legislative drafting mistakes.130

B.     Applying the Dynamics

The TCJA, the most recent major tax legislation and the largest overhaul of the tax system in over 30 years,131 provides numerous recent examples of how these dynamics play out. Due to a variety of legislative irregularities in the passage of the TCJA, including the incredible speed at which it was pushed through Congress, the TCJA contained numerous legislative drafting mistakes.132 Yet, Congress and the public placed immense pressure on the implementing agency, the Treasury Department (accompanied by the IRS), to ensure taxpayers could apply the new tax law.133

The IRS resolved this dilemma in some cases by providing agency legislative fixes. One such agency legislative fix addressed sexual abuse attorney fees.134 The TCJA included a provision, informally referred to as the “Weinstein tax,”135 which disallowed deductions for “(1) any settlement or payment related to sexual harassment or sexual abuse if such settlement or payment is subject to a nondisclosure agreement, or (2) attorney’s fees related to such a settlement or payment.”136 Democratic Senator Menendez had originally proposed this addition to the TCJA as a way to ensure that businesses could not deduct hush money paid to cover up sexual abuse.137 However, when tax law drafters translated this conceptual proposal into legislative language,138 they inadvertently broadened its scope (including by adding an overly broad cross-reference to “this chapter,” which referenced not only business, but also individual, deductions).139 This seemingly innocuous drafting mistake had significant consequences. It seemed to prevent not only businesses, but also sexual abuse victims, from deducting attorney fees they pay with respect to sexual abuse settlements.140 Senator Menendez quickly acknowledged a drafting mistake had been made, criticizing Republicans for the “outrageous and maddening” result, which “inevitably can occur when members are forced to vote on haphazardly rushed legislation before even getting a chance to read it.”141

Congress made various efforts in the aftermath of the passage of the TCJA to fix the mistake, but ultimately lacked the ability to do so. In December 2018, the Republican-led House of Representatives passed the Retirement, Savings, and Other Tax Relief Act of 2018, which principally addressed other issues but included a number of technical fixes to the TCJA, including the sexual abuse attorney fees mistake.142 That legislation failed to advance through the full Congress as a result of Democratic opposition to various provisions as well as the cost of the legislation.143 Republican representative Kevin Brady, then-Chair of the House Ways and Means Committee, then introduced a discussion draft of the Tax Technical and Clerical Corrections Act.144 The proposed legislation would have, among many other fixes, limited the new sexual abuse provision only to business deductions, which would have eliminated the unintended limitation for sexual abuse victims.145 However, none of the proposed technical corrections progressed under Brady’s tenure, even though Democratic and Republican lawmakers seemed to be in agreement that a mistake had been made, and bi-partisan support existed to change the legislation.146 Senator Menendez also moved forward on his own, introducing the “Repeal the Trump Tax Hike on Victims of Sexual Harassment Act of 2018.”147 This proposal faltered as well.

These legislative failures left Treasury in a real bind. It was clear that Congress did not intend for sexual abuse victims to lose the ability to deduct attorney fees for settlements they received, as underscored by Congress’ efforts to fix the mistake in the aftermath of legislative passage. Yet, the text of the legislation that was passed also seemed clear: It explicitly disallowed “attorney’s fees related” to specified sexual abuse settlements under the income tax code “chapter,” which included the deduction otherwise available for victims’ attorney fees. In other words, it was perfectly clear that Congress had not intended the result that it had created, and that Congress nonetheless seemed, at least at present, unable to fix it. And the clock was ticking—tax return filing season for 2018 (the first year in which the TCJA was in effect) was beginning. The question was: Would the IRS abide by what the text of the law seemed to require? Or would the IRS deviate from it? If the IRS deviated from the text, would it tell the public in advance?

On February 28, 2019, just as tax return filing season was about to kick into high gear, the IRS answered this question in the form of an “FAQ” on its website. The IRS’ FAQ asked, “Does section 162(q) preclude me from deducting my attorney’s fees related to the settlement of my sexual harassment claim if the settlement is subject to a nondisclosure agreement?”148 The IRS answered:

No, recipients of settlements or payments related to sexual harassment or sexual abuse, whose settlement or payment is subject to a nondisclosure agreement, are not precluded by section 162(q) from deducting attorney’s fees related to the settlement or payment, if otherwise deductible. See Publication 525, Taxable and Nontaxable Income, for additional information on when all or a portion of attorney’s fees may be deductible.149

And, simple as that, the IRS resolved the issue with a few short sentences. No muss. No fuss. This was the answer everyone knew accomplished what Congress had intended. It settled the matter for sexual abuse victims, who could now deduct their attorney fees without concern. And it would allow the IRS to administer the tax system without having to worry about its agents taking different positions in auditing taxpayers’ returns. For all these reasons, one could reasonably argue that the IRS had taken the only sensible path.

But the reasonableness of the IRS’ actions did not diminish the hubris that had accompanied it. Powerful Republican and Democratic representatives alike (including Senate Finance Committee Member Senator Menendez and House Ways and Means Committee Chairman Brady) tried to make this correction and yet lacked the ability to muster enough votes. In their stead, the IRS, an unelected agency, simply fixed the problem on its own through an “FAQ” on its website, an informal communication that lacked any status or process as a form of law.

In taking this tack, the IRS adopted a particularly notable, but also paradigmatic, agency legislative fix. Here, as elsewhere,150 the IRS used informal guidance, which, as a practical matter, no one could challenge,151 to fix a legislative drafting mistake in a way Congress could not, and in a way that was likely to be disapproved of by courts, had they had a chance to review the action.

While the IRS was willing to make a fix in some cases, it was not willing to do so in others. A good counterexample occurred with respect to qualified improvement property. This mistake arose out of very detailed changes that Congress made to the rules regarding the depreciation of property in the TCJA. Prior to the TCJA, various classes of “improvement property” were eligible for depreciation over a relatively short time period. As one of the housekeeping changes in the TCJA, Congress consolidated the various types of “improvement property” into one category—“qualified improvement property.”152 Congress had meant for such qualified improvement property to retain its prior, favorable depreciation designation, which, under the new TCJA rules, would also make it eligible for an even more favorable provision—bonus depreciation.153 However, in what was widely acknowledged by Congress members and commentators to be a drafting mistake, the TCJA drafters failed to include cross-references to the newly consolidated qualified improvement property category in the appropriate places in the depreciation rules.154 The unintended result was that qualified improvement property would be ineligible for the new bonus depreciation, and, even worse, it would actually be subject to depreciation over a much longer period of time than had been the case under the depreciation rules that preceded the TCJA.155

Shortly after the TCJA passed, the qualified improvement property mistake was widely identified as such, often grouped with the sexual abuse attorney fees mistake in this regard. Interest groups came out in droves, arguing that the qualified improvement property treatment in the TCJA was a mistake that should be fixed.156 Various government officials affirmed that, like with sexual abuse attorney fees, Congress made a mistake for qualified improvement property.157 For instance, a JCT legislation tax accountant explained that the final draft of the qualified improvement provision did not reflect the conference agreement.158

A delicate dance began regarding the possibility of a qualified improvement property fix. Treasury moved first by issuing proposed regulations on the new depreciation rules, which did not correct the legislative drafting error that had been made.159 In reaction, the Senate Finance Committee Republicans who had enacted the TCJA wrote a letter to Treasury and the IRS in order to “clarify the congressional intent of this recently enacted tax legislation.”160 The letter fleshed out various instances, including both the qualified improvement property mistake and the sexual abuse attorney fee mistake, in which, at least according to the Senate Finance Committee Republicans, drafting mistakes had thwarted Congress’ intent.161 The letter also pointed to legislative history in support of its claims.162 The Treasury Department demurred, responding that, while the letter from the Senate Finance Republicans was “welcome,” it “may not be enough” to get to an administrative solution to the problems and that, more generally, Treasury was still talking to Capitol Hill about the potential options.163 An IRS branch chief in the Office of Associate Chief Counsel explained that the agency typically looks for statements of intent from the “four horsemen—the chairs and ranking members of the Senate Finance and House Ways and Means committees,” and that the IRS was thus waiting to see if any statement came from House Ways and Means.164 Sixteen Senate Democrats got in on the act, writing a subsequent letter to Treasury arguing that, in light of the fact that a drafting mistake was clearly made, as reflected in the legislative history and the assumption of the more favorable depreciation rules that went into the Joint Committee’s scoring of the legislation,165 it would be “prudent for Treasury to address this issue and its interpretation through guidance.”166

Unlike with the sexual abuse attorney fees mistake, Treasury ultimately declined to make an administrative fix for the qualified improvement property mistake. A tax policy advisor in the Treasury Office of Tax Legislative Counsel stated that, “We’ve considered what authority we may have . . . and I think we have come to the point now where we need a technical correction.”167 In other words, it was Congress or bust.

Subsequent to Treasury’s indication that it would not fix the qualified improvement property mistake, Congress attempted to do so. Like with the sexual abuse attorney fee mistake, the House’s Retirement, Savings, and Other Tax Relief Act of 2018 contained a proposed fix for qualified improvement property,168 as did Representative Brady’s discussion draft of the Tax Technical and Clerical Corrections Act.169 Because Treasury refused to fix the mistake in the qualified improvement property context, the congressional efforts also continued in a variety of fashions. For instance, in March 2019, a bipartisan pair of senators joined to introduce yet another bill to fix the qualified improvement property mistake,170 an effort that was followed by a bipartisan House bill designed to do the same thing.171 But, while hope remained that Congress would eventually fix the qualified improvement property mistake, it was not done in a timely fashion—the 2018 tax filing season was well underway without a fix in sight.172

C.     Considerations Influencing Agency Action and Inaction in
Response to the TCJA

There were striking similarities between the sexual abuse attorney fee mistake and the qualified improvement property mistake. Both resulted from a failed cross-reference,173 an easy mistake to make in the rush of frenzied legislative passage. In both cases, members of Congress as well as other commentators came out shortly after the passage of the legislation to identify the mistake. Indeed, Congress and members of the public often mentioned the two mistakes in the same breath as glaring errors that had been made in the TCJA.174 In both cases, members of Congress tried to fix the mistake in time for the first tax return filing season affected by the TCJA but failed to do so. And, in both cases, all these factors joined together to put pressure on Treasury to fix the problem. But, while the IRS fixed the sexual abuse attorney fee mistake, neither Treasury nor the IRS were willing to do so for qualified improvement property, at least not in time for filing season.

So, what explains the agency’s willingness to step in and fix Congress’ legislative drafting mistake in the case of sexual abuse attorney fees but not for qualified improvement property? The short answer is that the answer simply is not clear. In this case, as in others, the lack of any systematic or transparent process for making and communicating decisions about agency legislative fixes created uncertainty about what motivated the agency’s decisions. However, some possible hypotheses are discussed below.

1.     The Statute

One possibility is that the statute dictated the different outcomes in the different cases. If this were true, there would be nothing that remarkable about Treasury making the fix in the former case but not the latter. Treasury would just be making fixes only when permitted by the statute.

This is certainly possible. Tellingly, in explaining why Treasury had refused to reach an agency legislative fix for qualified improvement property, an IRS associate chief counsel lamented that, despite the bad result, “there’s no authority for us, administratively, to step in and override the clear language of the statute.”175 Seemingly, under the statute, the failure to cross-reference the new qualified improvement property category as one of the types of property that would be eligible for more favorable depreciation rules left the new category subject to the default, undesirable treatment applicable to nonresidential real property.176 In contrast, the statutory prohibition on deducting “attorney’s fees related to [any settlement or payment related to sexual harassment or sexual abuse]” had some slight looseness in the statutory language, including the “related to” phrase, which may have given Treasury some interpretive space to make a fix.

But it also is not entirely clear that the statute provided Treasury substantially more wiggle room to make the fix in the case of the sexual abuse attorney fees. As described previously, notwithstanding the “related to” language, the sexual abuse provision seemed to broadly disallow deductions for attorney fees even for sexual abuse victims.177 In any event, the IRS, in its FAQs, used the exact same “related to” language to seemingly inexplicably reach the exact opposite conclusion of the statute. The IRS explained that “recipients of settlements or payments related to sexual harassment or sexual abuse, whose settlement or payment is subject to a nondisclosure agreement, are not precluded by section 162(q) from deducting attorney’s fees related to the settlement or payment, if otherwise deductible.”178 In using the exact same language of the statute but reaching the opposite conclusion, the IRS did not seem to be relying on statutory looseness. Rather, the IRS simply seemed to be setting the statute aside to reach a different result.

Moreover, there are arguments that a qualified improvement property fix would have been consistent with the statute. Specifically, the reason why qualified improvement property seems to be ineligible for a shorter depreciation schedule and bonus depreciation is because it is not explicitly listed as falling in a category of property that is eligible for a shorter depreciation period. As a result, it would appear, by default, to fall into the category of “nonresidential real property,” which is listed as having a longer, 39-year recovery period, making it ineligible for bonus depreciation.179 But, one could argue that, by creating a defined category of “qualified improvement property” separate and apart from “nonresidential real property,” Congress implicitly indicated that the new category should not just be treated as nonresidential real property by default.180 If interpreted this way, the statute simply does not specify the recovery period for qualified improvement property, perhaps opening the door to Treasury answering the question through the use of legislative history, which may enable a fix.

In any event, whether the differences in statutory language explain why the agency treated these two situations differently or not, the IRS did not openly say so. Rather, in its website FAQs, the IRS just asserted that sexual abuse victims could deduct attorney fees. The IRS never defended this decision publicly or subjected it, or its criteria generally, to transparent explanation.

2.     Legislative History

Even if the statute did not dictate the different treatments, the legislative history might. Depending on one’s method of interpretation, legislative history can be an important interpretive tool, and agencies may look to legislative history to provide support for an interpretation that would otherwise appear to conflict with the statute.181

However, this factor was not determinative in the case of the two TCJA examples. There was essentially no legislative history regarding the TCJA’s addition of the new sexual abuse deduction limitation.182 In contrast, the legislative history regarding qualified improvement property clearly supported the conclusion that an inadvertent drafting mistake had been made. The conference committee report for the TCJA had (incorrectly) asserted that qualified improvement property was subject to a 15-year recovery period.183 The Joint Explanatory Statement of the conference committee released in conjunction with the TCJA also asserted that the now consolidated category of qualified improvement property had a 15-year recovery period.184 The IRS’ willingness to fix the sexual abuse attorney fee mistake, but not the qualified improvement property mistake, indicates that while legislative history may matter in some cases, it is far from the only explanatory factor. Moreover, the qualified improvement property example indicates that, even when the legislative history shows the drafters made a mistake, an agency may not be willing to fix it.

3.     More Practical Considerations

If statutory language and legislative history alone may not explain an agency’s decisions, what other factors may serve as motivators? A variety of more practical factors are also likely to matter.

i.     Procedural Options for Making a Fix

One possibility is that procedural options for making a fix serve as a factor. The IRS likely found it easier to make an agency legislative fix in the case of the sexual abuse attorney fees because the IRS was able to make the fix in this case in the form of frequently asked questions (“FAQs”) on its website.185 Such FAQs have no formal status as or place in the law. Since they are not officially binding on anyone, no party would be in a position to challenge them.186 In contrast, Treasury had to provide guidance about the new depreciation and expensing regime in officially-issued regulations because of the complexity of the depreciation and expensing rules and the extensive changes the TCJA made to them. Notwithstanding the fact that few would have standing to challenge taxpayer-favorable regulations, such regulations constitute a substantial regulatory action that still must go through extensive internal and external procedures. These include the Office of Information and Regulatory Affairs (“OIRA”) review and the notice-and-comment process. Treasury would be hard-pressed to defend positions inconsistent with the statute in such procedural spaces.

ii.     Informal Dialogue with Congress

The examples from the TCJA reveal another factor in considering an agency legislative fix: informal dialogue with members of Congress after legislative passage. In the case of both mistakes, Congress members from both sides of the aisle reached out to encourage an agency legislative fix. There were clearly internal norms that helped shape the communications, although such norms have no formal place in the law. For instance, the IRS branch chief’s explanation that the agency typically looks for statements of intent from the “four horsemen—the chairs and ranking members of the Senate Finance and House Ways and Means committees,” and that the IRS was thus waiting to see if any statement came from House Ways and Means,187 has no place in formal law. According to clearly established precedent, only Congress can make law as a body, through the formal procedures of bicameralism and presentment.188 From this perspective, it is unclear why it would matter whether the four horsemen issued a letter indicating their agreement that a legislative drafting mistake had been made. If Congress believes a legislative drafting mistake has been made, Congress’ recourse, as a formal matter, would be to issue corrective legislation.

But this dialogue nonetheless matters in the context of agency legislative fixes. Marc Gerson, former majority tax counsel to the House Ways and Means Committee, described in a rare article about the subject how staff within Congress, Congress members themselves, and Treasury and IRS officials work closely together to determine how to address legislative drafting problems.189 These actors negotiate carefully over whether the agency may address the issue through administrative guidance, rather than requiring Congress to actually pass corrective legislation in the form of a technical correction.190 For this reason, taxpayers and their advisors often first look to Treasury and the IRS to see what legislative problems can be fixed administratively before looking to Congress.191 While this path would seem to be backwards, considering Congress’ constitutional role in passing and correcting legislation,192 it reflects the reality of agency legislative fixes and the informal dialogue with Congress that can motivate them.

iii.     Public Opinion and Other Factors

Other practical factors surely help explain the agency’s decisions about whether to correct mistakes. For instance, commentators suggested that Congress, JCT, and Treasury all have to have a meeting of the minds in order to pass a technical corrections bill.193 It is unclear why Treasury, for instance, would have a say in whether a technical corrections bill could pass, a question that would seem to lie solely with Congress. Perhaps the dialogue running from the agency to Congress about the likelihood of reaching a technical corrections solution in Congress may also motivate an agency’s decision about whether to try to reach a solution in the administrative context. In this regard, perhaps the fact that Congress tried and failed to reach a resolution on the sexual abuse attorney fees issue prior to Treasury acting at all encouraged Treasury to resolve it. In contrast, Treasury became relatively locked into its position that it would not resolve the qualified improvement property mistake prior to Congress indicating that it would not act either. 

Public opinion likely also played a role in motivating the agency legislative fix in the case of the sexual abuse drafting mistake. Increased attention to sexual abuse and a desire to support victims motivated Congress’ change in the law.194 As a policy matter, it would make little sense for the IRS to enforce a new rule that denies victims a deduction for attorney fees, thus accomplishing exactly the opposite of Congress’ goals.195 Moreover, asking the IRS to be the agency that does this is doubly troubling. The IRS is an agency already perpetually struggling for public support.196 The public is unlikely to understand (or at least fully accept) that the IRS, the public face of the tax law, is bound by a mistaken cross-reference in a statute.197 Asking the IRS to be the public face of a mistaken change in the law that would be likely to garner widespread outrage was likely too much for the IRS to bear. Moreover, as an individual income tax deduction affecting only sexual abuse victims who engaged in settlements subject to nondisclosure agreements, the sexual abuse legislative fix was actually a relatively low-cost fix. While the IRS clearly faced industry pressure to make the qualified improvement property fix as well, the IRS may have rightly judged that making the sexual abuse fix was even more important from a public-relations perspective, especially considering the relatively low cost of doing so, relative to the issue’s salience. Together, these sorts of practical considerations, which have no formal place in legal interpretation, nonetheless likely help motivate decisions about agency legislatives fixes.

IV.     Evaluating Agency Legislative Fixes

Having identified the practice of agency legislative fixes, the next question is how to analyze them normatively. This Part explores how, even though agency legislative fixes may fill an important function for agencies and the public once Congress makes a legislative drafting mistake, they are still problematic from a normative perspective. Agency legislative fixes infringe on formal visions of legislative supremacy in a way that may not be remedied from a functional perspective. It is controversial to justify them as legitimate exercises of agency power. And even the practical case for them is somewhat ambiguous.

A.     Legislative Supremacy

The first dimension for evaluating agency legislative fixes is how they impact legislative supremacy. From a formal perspective, agency legislative fixes are problematic. While there are some functional arguments in favor of agency legislative fixes, they are more contingent and contested.

1.     Formal Visions of Legislative Supremacy

Formal visions of legislative supremacy look to whether or not Congress is the body stating what the law is, in accordance with the Constitution’s grant of legislative power to Congress.198 It is easy to dismiss agency legislative fixes as violating formal visions of legislative supremacy. The reason is both fundamental and straightforward: Fixing mistakes arguably enables branches of government other than Congress to make law. The Supreme Court underscored this concern in Lamie v. United States Trustee.199 In Lamie, a variety of clues in the statute’s text created the impression of an “apparent legislative drafting error.”200 Legislative history supported this conclusion (although there were arguments to the contrary), and a leading treatise agreed.201 But the Court refused to deviate from a verbatim reading of the statute, explaining that, “[i]t results from ‘deference to the supremacy of the Legislature, as well as recognition that Congressmen typically vote on the language of a bill.’”202

Indeed, this formal infringement on Congress’ lawmaking power may seem even more problematic when agencies carry it out. Executive agencies formally sit in the executive branch of government, and thus their exertion of power may seem to aggrandize the President’s power.203 Concerns about presidential aggrandizement have increased in recent years, with both political parties alleging that various presidential actions have weakened Congress and threatened to give the President a sort of imperial power.204 Agency legislative fixes may exacerbate this tendency by allowing the President to direct agencies to say that the law is different than what Congress passed. In the extreme, agencies’ ability to make legislative fixes could render Congress powerless.

2.     Functional Visions of Legislative Supremacy

But looking to Congress’ formal lawmaking role is not the only way to assess legislative supremacy. More functional visions of legislative supremacy ask how agency legislative fixes, or the lack thereof, affect Congress’ power relative to other institutional players.205 This analysis would consider real-world dynamics, including the multifaceted role of agencies in government.206 From a functional perspective, there may be some arguments in favor of agency legislative fixes. In particular, the realities of how agencies operate and interact with Congress suggest that, in some ways, agency legislative fixes may promote Congress’ ability to legislate with specificity. To understand this point, it is essential to realize that when Congress passes legislation, it is not Congress members themselves, but rather an army of staffers, that is principally responsible for drafting the legislation.207 Recent research has revealed that Congress members pay little attention to the details of crafting legislation or the details of legislation itself, focusing instead on setting broad strokes of what the legislation will accomplish.208 They largely defer to staffers in turning these broad strokes into actual legislative language.209 Recent research has revealed that Congress members’ knowledge of what is actually in the text of enacted statutes typically comes from brief staffer-prepared summaries.210

This army of legislative drafters matters because they commit the mistakes examined in this Article. Over and over, the types of errors that undermined congressional intent were erroneous cross-references (or erroneous failures to include cross-references) in the statute. This was true with the legislative drafting mistakes that occurred in the TCJA,211 the ACA,212 and the Dodd–Frank Act.213 Prior research revealed professional drafters (principally Legislative Counsel), not Congress members themselves, make decisions about cross-references, including how new provisions relate to and should be cross-referenced relative to old ones.214 Thus, these drafters make decisions at the level that cause drafting mistakes.

So, what is the impact of agency legislative fixes on such drafters’ choices? While it is of course difficult to say for sure, it is at least possible to consider the incentives such drafters face. Without such fixes, it may be riskier for legislative drafters to draft in detail. The more detailed the legislation, the higher the likelihood may be of making inadvertent mistakes.215 The lower the ability of agencies to easily correct such mistakes, the higher the cost of drafting detailed legislation. Conversely, if agency legislative fixes are more likely, more detailed legislation can be more easily corrected, lowering the risk of drafting in greater detail. Lowering the likelihood of agency legislative fixes thereby may encourage drafters to draft less detailed legislation, instead delegating more decisions to agencies.

As scholars have recently argued, more congressional delegation to agencies may be problematic because it weakens Congress as a body.216 Congress is supposed to make law through bicameralism and presentment.217 When agencies have greater power to make legislative decisions, Congress members can avoid these onerous procedures and individually influence agencies through a variety of formal and informal mechanisms of control.218 Congress’ complicity in greater delegation to agencies means Congress may no longer serve as a sufficient check on the executive. The evils that bicameralism and presentment are supposed to minimize, such as the power of factions to disproportionately control government, may instead be maximized.219

Exacerbating the above incentives is the fact that legislative drafters may more acutely bear the error costs of detailed drafting, enhancing their incentives to delegate to a greater extent to agencies. Legislative drafters’ role, while crucial, is also typically largely invisible to the public. Drafters are supposed to carry out congressional will and, if all goes well, their role in the process will be relatively unseen. However, when legislative drafting errors occur, drafters’ roles become visible in the worst possible way. Indeed, a particularly notable example of this occurred in the context of the ACA and the legal challenge that threatened to imperil it. A New York Times article provided accounts of how the drafting error had come about, attributing it to a failure to add cross-references to the portion of the legislation providing tax subsidies.220 The article pointed out that “the [Senate] Finance Committee voted on a detailed conceptual description of the bill,” not on actual legislative language.221 The article focused on the role of staffers in drafting the actual legislative language, explaining that “[t]he words were written by professional drafters—skilled nonpartisan lawyers—from the office of the Senate legislative counsel,” even naming the individuals responsible for drafting the actual legislative language.222 While perhaps a particularly poignant example, all drafters would reasonably seek to avoid this type of attention for a critical drafting error. Indeed, in prior interview research of tax legislation drafters, one interviewee explained that staffers involved in drafting legislation are very concerned about their reputations among insiders who will evaluate the quality of the drafting.223 This is because staffers want to be able to gain employment later based on their reputations.224 This may create an agency cost problem in which legislating with greater specificity would be more desirable, but doing so increases the risk to legislative drafters in particular. These drafters thus may be inclined to legislate with less specificity to avoid such mistakes.

A more lenient approach to legislative drafting mistakes may provide particular comfort to legislative drafters because they often have close working relationships with their agency counterparts.225 A lenient approach would allow legislative drafters to draft with greater specificity, knowing they can work with their agency counterparts on the back-end to clean up any technical mistakes made in the legislative process. This could yield a legislative supremacy tradeoff, whereby agency legislative fixes might formally seem to infringe on Congress’ lawmaking supremacy, but functionally may encourage more specific decisions to be made by drafters under Congress’ auspices.

But this possibility underscores a deeper question about what legislative supremacy means in a world in which Congress does not know the details of the legislation it is passing.226 From a functional perspective, does it really promote legislative supremacy to a greater extent for detailed decisions to be made at the congressional level if Congress will be unaware of those decisions? Some might say yes, pointing to the fact that congressional drafters at least work for Congress directly and are tasked with promoting Congress’ will.227 But a functional approach must contend with what the difference really is in practice, especially to the extent that the alternative decisionmakers, agency staffers, are also career functionaries simply trying to make the legislation as technically workable as possible.228

In other words, one might argue that agency legislative fixes promote a functional vision of legislative supremacy, for instance as a result of various drafting incentives. But arguments along these lines are more contingent on how the legislative process actually plays out in reality. More deeply, the functional argument runs into the problem of articulating what legislative supremacy really means in the realities of the legislative process. Thus, while arguments can certainly be made that agency legislative fixes promote some functional visions of legislative supremacy, the case does not clearly overcome more formalist objections.

B.     Legitimate Exercise of Agency Authority?

It is also difficult to justify agency legislative fixes as exercises of legitimate agency power. This related lens focuses on the power agencies are exercising, instead of the power Congress is losing. This lens asks: When agencies issue legislative fixes, may these fixes be a legitimate exercise of agency authority? Here, too, the answer is problematic. This is in part due to deeply contested views about agency legitimacy, layered onto the complicated, multifaceted nature of agency legislative fixes.

As an initial matter, the justifications for agencies creating public rights and responsibilities are deeply contested.229 The Constitution does not mention agencies, and agency officials are not elected.230 Over time, scholars have developed a variety of theories to justify agencies’ exercises of power.231

In some ways, one model in particular, the expertise model of the administrative state, seems to implicate and potentially legitimate agency legislative fixes. Under this model, agency officials are “not political, but professional,” providing “an objective basis” that is achieved through exercise of agencies’ “specialized experience.”232 In essence, the expertise model suggests that agency discretion can be justified because the agency decisions are actually not discretionary. Rather, expert analysis compels agency decisions.233

The technical and somewhat limited nature of many agency legislative fixes in some ways seems to map onto this expertise model. While Congress and courts face substantial, sometimes self-imposed, barriers in correcting legislative drafting mistakes in a timely fashion, it is still the case that they will correct some.234 As an example, even if the Supreme Court did not say so, it was willing to correct a legislative drafting mistake when the viability of the signature legislative achievement of the Obama administration turned on it.235 And, while Congress may face significant difficulty in correcting its legislative drafting mistakes in the legislation’s aftermath, it too may muster the political will to do so when both the political consequences of failing to do so are extreme enough and enough other pieces of the legislative puzzle fall in place.236 This means the types of mistakes left to agencies alone will tend to be less politically sensitive. In such cases, agencies may be well-positioned, with their knowledge of the statute’s intricacies, to identify mistakes and make the fixes. If this narrative holds, then agency legislative fixes may be a legitimate exercise of agency expertise.

But the expertise model generally suffers from some weaknesses, which are certainly applicable in the case of agency legislative fixes. Namely, agency decisions are often value-laden policy choices rather than pure exercises of expertise.237 As applied to agency legislative fixes, the concern would be that many fixes are not necessarily obvious applications of expert analysis, but rather reflect agency officials exercising politically motivated policy judgment. The sexual abuse attorney fees fix illustrates this phenomenon well. On the one hand, the sexual abuse mistake seems like a mere technical drafting mistake—the drafters inserted a too-broad cross-reference to “this chapter”238—a mistake that the agency, through the use of its expertise, could correct. On the other hand, the insertion of the too-broad cross-reference to “this chapter” may reflect a deeper failure to think through how the new deduction disallowance would have ramifications throughout the code and disadvantage politically sensitive constituencies. One could just as easily characterize this as a policy, or even political, error. If this were the case, the fix would no longer be an inevitable or objective exercise of agency expertise.

If expertise alone may not justify the universe of agency legislative fixes, there are other options. One possibility is that agency legislative fixes may be characterized as part of an agency’s authority to fill in legislative ambiguity. Scholars have often defended the highly contested Chevron doctrine on the ground that agencies should have the authority to fill in ambiguities in statutes because agencies answer to the politically accountable executive department.239 If this justification holds for ambiguity generally, perhaps legislative drafting mistakes should be seen as creating statutory ambiguity, which agencies thus have the power to resolve.

There are multiple difficulties with this approach. First, as alluded to above, the Chevron doctrine itself is under sustained attack.240 Moreover, even if one subscribes to Chevron’s allocation of power to agencies in cases of statutory ambiguity, it is not clear whether legislative drafting mistakes constitute ambiguity. In a close analogue to the question, there is disagreement among judges about whether problems in statutory regimes create ambiguity that merits Chevron deference.241 One’s approach to this question may well turn on not only one’s view regarding whether agency power should be seen as legitimate only when Congress clearly intended to delegate to the agency, but also on the equally weighty question of what is the best mode of statutory interpretation. While a textualist may be less inclined to find any ambiguity when the statute seems clear, more purposivist approaches may find that suggestions of congressional intent in conflict with the statutory text may create ambiguity about the text, which then may supply an agency the power to identify a legislative drafting mistake and fix it.242

Clearly, questions about which mode of statutory interpretation is better, along with questions about what justifies agency power, are massive. Extensive amounts have been written about each of these subjects.243 The depth of these issues means that whether agencies can legitimately fix legislative drafting mistakes depends on inherently contested, rather than consensus, visions of agency power and methods of interpretation.

C.     Practical Considerations

Finally, turning away from grand constitutional and democratic theory, agency legislative fixes also have practical impacts. Practical considerations present the strongest case for agency legislative fixes because such fixes may play an important role in protecting relationships between agencies and the public. However, even along this practical dimension, there is ambiguity about whether agency legislative fixes have a positive or negative effect overall. And it is unclear whether any practical benefits could overcome more foundational problems.

1.     Benefits of Fixing the Law

From an agency perspective, once a mistake in legislation has been made, agency legislative fixes serve an important role. Failing to fix a legislative drafting mistake through guidance is costly, both financially and reputationally, to the agency. Not fixing the mistake means the agency may be expending resources to enforce the law in a manner Congress did not intend, sometimes in a fashion that runs strongly counter to public opinion. The sexual abuse attorney fee mistake is a prime example of this dilemma. As discussed previously, the lack of an agency legislative fix would have required the IRS to very publicly and likely painfully fall on its sword in defense of an outcome no one actually wanted.244

Absent an agency legislative fix, an agency could avoid the consequences of a legislative drafting mistake simply by not enforcing it. Indeed, the Senate Republicans who wrote to Treasury in the wake of the TCJA suggested that the IRS should rely on selective enforcement to carry out Congress’ intent.245 Various practitioners seconded that this outcome may be better than Treasury using its resources to enforce the law in a way that Congress did not intend.246

However, nonenforcement is problematic as well. It forces regulated parties to assume the regulating agency will ignore the law. It also has the potential, problematic effect of inadvertently encouraging regulated parties to engage in self-help more broadly when they believe that there is a problem with the law. Nonenforcement also places an arbitrarily higher burden on regulated parties who comply with the law as written, either because, by disposition, they are incapable of doing otherwise, or, even more problematically, because they do not have access to the informal sources of information that suggest they should do otherwise.

More generally, addressing legislative drafting mistakes through a policy of nonenforcement pushes the subordination of the text of the law deeper inside the agency, making the infringement on legislative text less transparent to the public. While some well-connected parties are likely to know of the nonenforcement policy, others are not, yielding inequities in access to the law as applied. Worst of all, if the agency relies on de facto nonenforcement, rather than an official, public policy against enforcement, the agency leaves individual agents leeway to deviate from the nonenforcement, setting the stage for inequitable treatment under the law.247

2.     Negative Impacts on Agency Guidance Practices

While agency legislative fixes provide some clear practical benefits, there are also some likely practical costs. As fleshed out previously, agencies are likely to be more willing to issue legislative fixes when they can do so without being challenged.248 This means agencies are systematically more likely to issue legislative fixes through informal guidance, and in a way that favors regulated parties.

Agencies’ use of informal guidance may undermine important values. The elaborate notice-and-comment procedures that accompany agency promulgation of regulations are designed to promote participation and accountability in the regulatory process.249 These values are thought to infuse agency decision-making with legitimacy that is lost when Congress delegates governance decisions to unelected agency officials.250 When an agency instead uses less formal means of making administrative decisions, these procedures, and the values they are supposed to promote, are lost, undermining legitimacy of agency decisions.251

Moreover, the fact that agencies are likely to feel freer in issuing legislative fixes that benefit, rather than hurt, regulated parties, is likely to skew the direction of legislation. All other things being equal, Congress is likely to pass laws with legislative drafting mistakes that both benefit and harm regulated parties. This means that, while particular legislative drafting mistakes may have particularized costs, legislative drafting mistakes in general should not be expected to advantage or disadvantage either regulated parties or the general public, relative to each other. But, if agency legislative fixes systematically occur in a pro-regulated party direction, they are likely to bias the law in a way that diverges from the law as passed by Congress.

The fact that even the practical case for agency legislative fixes is ambiguous is telling. There are certainly reasons why agencies would offer legislative fixes. They can save the agency and the public from having to take positions contrary to what Congress actually wanted. But this benefit comes at a significant cost in terms of other negative practical consequences, as well as deeper potential threats to legislative supremacy and agency legitimacy. In a world in which Congress is finding it easier to make mistakes but harder to correct them, agency legislative fixes fill an important, but problematic, role.

V.     Potential Paths Forward

The widespread nature of agency legislative fixes, combined with the normative problems with such fixes, begs the question: Are there any better paths forward? This Part suggests there are no easy, magic wand solutions. However, there are some discrete changes that may help both Congress and agencies address the legislative drafting mistakes that are inevitable in today’s legislative process.

A.     Stick Congress with its Mistakes

One possibility is for agencies to simply stop issuing agency legislative fixes. In other words, just stick Congress with its own mistakes. As outlined in Part II, there are often high costs in Congress correcting its own mistakes.252 In some cases, these high costs essentially mean there is no foreseeable path for Congress to fix its mistakes.253 However, as suggested in Section IV.B, sometimes the costs may not be insurmountable, especially in the long-term as political conditions continue to shift.254 Congress is more likely to be able to correct its mistakes when the pressure is high to make a fix and other legislative conditions smooth the way, such as the availability of another must-pass piece of bipartisan legislation on which the corrections can ride.255 To increase the likelihood of a fix, the party that wants the fix may be able to get it by giving the other party large concessions. Exemplifying this last possibility, Democratic Senator Sherrod Brown explained in the wake of the TCJA, “[w]e’re not going to say to Republicans, ‘Oh tell us what you want to do,’ . . . . We want to make the bill better, not just correct whatever technical fix is needed.”256

One particular fix from the TCJA exemplifies these dynamics well. The only legislative drafting mistake Congress was able to fix in relatively short order after the passage of the TCJA was the so-called “grain glitch.” The grain glitch was a drafting mistake made in a new provision for pass-through businesses, which created significant, unintended disadvantages for certain agricultural businesses as well as a loophole for aggressive tax planning.257 Congress passed a fix in March 2018, approximately three months after passage of the TCJA, by attaching it to a must-pass appropriations act.258 Democrats agreed to the addition of the grain glitch fix in exchange for an expansion of the low-income housing tax credit.259

To some, the notion that some price could be paid to make some number of fixes at the congressional level may be enough to reject agency legislative fixes entirely. The legislative enterprise is fundamentally one of concession and compromise.260 The fact that legislative fixes may be quite difficult for Congress to make perhaps should not excuse Congress from making them. Alleviating Congress of this obligation may be an abrogation of Congress’ burden to find compromises precisely when it is hard to do so.

But taking this perspective to the extreme shows there has to be some stopping point at which simply sticking Congress with mistakes becomes untenable. A legislative problem cannot be justified by the fact that some price, no matter how extreme, could theoretically be paid to fix it. The question is whether the cost is too high to fix the problem legislatively relative to other options. This Article has illustrated how the lack of agency legislative fixes creates real costs to the agency and the public in the many cases in which Congress simply cannot or will not alleviate them in a timely fashion. With the sexual abuse attorney fee mistake, for instance, it would not just be Congress, but also the IRS, and sexual abuse victims, who would bear the price for Congress not acting in time. Likewise, in the case of qualified improvement property, affected taxpayers complained bitterly about the lack of a timely qualified improvement property fix. The Retail Leaders Industry Association lamented that the failure to fix the mistake had “stifled growth and innovation across the industry,” and the National Retail Federation argued that the error affected “jobs in retail, restaurants and construction in every community in the country.”261 Moreover, others complained that the failure to get any sort of fix prior to the first tax return filing season impacted by the TCJA created “substantial uncertainty” for taxpayers about what they should do in filing their tax returns.262 The cost of such delay of course only compounds as time continues to pass without any resolution of the issue. The question, then, is not whether Congress can enact fixes at any cost, but rather what the best option is given all of the various costs of the differing, potential solutions.

Moreover, even to the extent that the right weighing of benefits and costs argues in favor of sticking Congress with its mistakes, it is not clear that this is the path agencies will actually take. Sticking Congress with its own mistakes may vindicate long-term constitutional and democratic values, but it is likely to do so by imposing significant costs on agencies and their relationships with the public, as in the case of the sexual abuse attorney fee mistake. As the decisionmaker, the agency is likely to focus on its own benefits from agency legislative fixes to a greater extent than is optimal. As a result, even in situations in which sticking Congress with its own mistakes is preferable, agency legislative fixes are likely to continue to some extent.

B.     First-Best Solution: Major Changes to Legislation

If sticking Congress with its own mistakes is an imperfect solution, then what are other options? The first-best solution would be for Congress to change its legislative drafting practices to make fewer mistakes. For instance, if part of the reason why Congress is making so many mistakes is that the legislative process is rushed, overly chaotic, or lacks a final point at which all of the different inputs are checked, Congress could add a mandatory deliberation period to the process or a time-period in which the legislation must be checked before Congress can vote. Congress could also make particular members, or a committee, responsible for overseeing this final deliberation period, to try to ensure that at least some Congress members have a vested stake in avoiding errors across legislation. Alternatively, Congress could implement an even broader review process, akin to the notice-and-comment procedures that apply in the administrative rulemaking context.263 Like with administrative rulemaking, providing draft legislation to the public may allow expert commentators to catch legislative drafting mistakes prior to their passage.264 Indeed, the Administrative Conference of the United States is currently engaged in a pilot program that attempts to improve legislative drafting.265

However, these and other first-best suggestions to change congressional procedures face a fundamental barrier: Congress itself. The reason why legislation looks the way it does today is because Congress has found its legislative practices advantageous to passing legislation. For instance, drafters have indicated that keeping legislation secret as long as possible can often be critical to keeping lobbyists at bay who would otherwise pounce and begin to unravel the legislation’s chance at passage.266 When Congress pushes legislation through at lightning speed, it is often because members used speed as a tactic to pass legislation.267 Thus, when mistakes happen it is often the case that more careful deliberation did not occur because drafters did not have time to slow down or the liberty to circulate the legislation in a way that would invite error correction. As for designated individuals or committees tasked with making sure all the legislation fits together in the right way, there are already numerous processes in place within Congress to ensure that drafting decisions are checked and cross-checked.268 But, Congress and its staffers are unable to catch all mistakes when they are operating under a crush of time pressure and other chaotic conditions. Adding additional committees or designated individuals will not change the basic dynamics that are currently producing drafting mistakes.

To generalize the problem, suggesting wholesale changes to legislative practices to prevent legislative drafting mistakes is an example of the inside/out fallacy. This fallacy involves diagnosing a problem by appreciating the real-world constraints an institution faces, and then proposing solutions that appear to magically solve the problem by ignoring all those constraints.269 This Article has sought to carefully explain the real-world dynamics that lead to legislative drafting mistakes and the accompanying agency legislative fixes. Having done so, it is untenable to then assume away these very realities to suggest ways to eliminate the causes of such mistakes.

C.     Second-Best Solutions

If first-best wholesale changes to legislative drafting practices are not plausible enough to eliminate the need for agency legislative fixes, what about second-best solutions? As a general matter, second-best solutions present well-known problems. According to the theory of the second best, if optimal solutions cannot be reached, it is unclear whether second-best solutions will move towards or away from the ideal.270 Moreover, second-best solutions arguably also can present inside/out fallacies. While being mindful of these limitations, the discussion below analyzes the extent to which several potential second-best solutions are actually plausible and identifies the values each of the possibilities vindicate and undermine. 

1.     Backend Legislative Fixes

While wholesale front-end legislative changes may not be realistic, perhaps Congress could plausibly make it easier to make backend legislative fixes. As explored previously, the current reconciliation rules stymie Congress’ ability to fix its own mistakes.271 The current practice is to treat mistakes as having no effect on outlays or revenues, because they do not affect the legislation as scored on original legislative passage.272 This means Congress is subject to a higher vote count in passing legislative fixes than it was subject to in passing the original legislation.273 The result is that it is extremely hard for Congress to fix mistakes in legislation it has passed.

Understanding this dynamic lends itself to a potential improvement: Congress could change its practices to allow Congress to fix legislative drafting mistakes through reconciliation. This would allow Congress to correct mistakes via the same vote count as the original legislation. This should improve Congress’ ability to correct its own drafting mistakes.

As a practical matter, this suggested change is more easily achievable than wholesale, front-end legislative reform. Budgeting rules in general are endogenous to Congress and have changed over time as Congress has seen fit to change them.274 The whole question of what should be used as the baseline for determining whether legislation would affect revenue (thereby making it eligible for reconciliation) is poorly understood.275 As suggested previously, budget estimators are not bound by formalized rules in making such determinations.276 Rather, estimators rely on informal practices, which the Senate Parliamentarian applies under a fair amount of opacity and subjectivity.277 As a result, if Congress wants to allow legislative drafting mistakes to be corrected in reconciliation, it could do so by changing the baseline for reconciliation eligibility from the existing approach of law as erroneously scored to an approach of law as enacted. This would mean that fixing legislative drafting mistakes would affect revenue, making the fixes eligible for reconciliation and the lower vote count it requires. The Senate Parliamentarian could make this change in consultation with members of Congress or the relevant advisory bodies such as CBO and JCT.278

This suggestion may seem audacious to some. Many believe Congress has already expanded reconciliation well beyond its original intention, to bad effect,279 and thus suggestions to increase eligibility for reconciliation may seem ill-advised.280 Some might think relying on the subjectivity and opacity of the budget rules to suggest Congress could easily make a change without requiring any formal process is a bad precedent to set.281 Some may worry that allowing the original legislation to pass based on a scoring assumption that the legislation is doing one thing, and then passing corrective legislation based on the fact that it really did another thing, may allow Congress to use shifting baselines in a way that games the legislative process. Scholars have forcefully raised the shifting baselines concern with respect to temporary legislation.282

However, this legitimate concern about shifting baselines does not apply well in the context of legislative drafting mistakes. Using the enacted legislation as the baseline, even when it was erroneously scored, holds Congress to an objective, fixed baseline—the law as actually enacted. As a result, the proposed baseline should provide Congress members little leeway to game the system. If Congress had intended to make the law x and scorers accordingly assumed it was x when it was enacted, but it really says y, and Congress wants to return it to the intended x, then x presumably was eligible for reconciliation in the first instance. So, using a baseline of the law as actually enacted, even when it was erroneously scored, should not improperly enable Congress to pass legislation that is only possible through shifting baselines.283

Moreover, the change being proposed in the context of legislative drafting mistakes is simply the right policy. While the budget rules are no model of consistency, they do generally look to current law as a baseline for budget analysis.284 This suggests that the practice of using the existing law, as previously scored, is actually a deviation from typical baseline budget policies rather than a necessary manifestation of them. Applying a baseline of the existing law, as actually enacted, is not only consistent with general baseline practices, but is also the only reasonable option when the actual law is different than what the scorers thought that it was. The alternative unreasonably holds, in an almost Kafkaesque fashion, that, as a matter of law, the law is the law as enacted, while, as a matter of budgeting, the law is the law as previously scored.

So, if the approach being offered is arguably plausible and would be a real improvement, what, if anything, might be the problems with it? First, while it might benefit the political party in power to change the baseline practices in order to fix their legislative drafting mistakes through reconciliation, doing so may also enable the other political party to then correct their mistakes more easily in the future. This concern about changing the rules in a way that benefits one’s own political party in the short term but may ultimately benefit the other political party in the long term arguably explains the filibuster’s durability.285 Whether this dynamic would serve as a significant enough obstacle to changing the baseline practices with respect to legislative drafting mistakes is not clear. One reason to be optimistic about the possibility of changing the baseline practices is that both parties may see this change as often advantageous to them whether the other party passed the legislation or not. This is because, once the original legislation is passed, both parties often would prefer to have the fixes made so that the legislation is at least as functional as possible, but may be hampered in their ability to actually vote for changes to legislation that they opposed. Making it easier for the party in power to make the changes through reconciliation may yield a win-win of allowing the opposing party to fix the legislation through reconciliation, without actually having to vote for it.286

However, even to the extent Congress changes the baseline practices for reconciliation eligibility, Congress may face political obstacles in making use of the change. An election on the heels of the passage of the original legislation may mean the majority party that passed the original legislation can no longer garner even the lower number of votes required to pass the original legislation. For example, in the case of both the ACA and the TCJA, the majority party that had passed the legislation lost control of the House in the election that followed the passage of the legislation.287 Even if the majority party maintains its majority, changed dynamics within that party may prevent even a reconciliation bill from coming to pass. One obstacle in particular is likely to be troublesome: the need to pass a new budget resolution in order to kick-off a new round of reconciliation.288 Relatedly, even when the majority party maintains its votes for long enough to pass corrective legislation and its internal dynamics remain stable, the majority party may worry that offering new legislation on the heels of the original legislation may offer the minority party an opportunity to make unforeseeable amendments to the new legislative vehicle, which may imperil the legislative victory that was just won.289 Alternatively, as a political matter, the majority party simply may want to move on after legislative passage, even if the cost of doing so is leaving mistakes in place. Finally, to the extent the baseline is changed to allow fixing mistakes to be scored as affecting revenue, making the changes may be revenue-reducing in certain circumstances. It may be politically difficult for Congress to pass legislation that reduces revenue on the heels of the recently enacted, major legislation.290 

All of these potential political obstacles to fixing mistakes, even with the new suggested baseline, are real. As a result, changing the baseline for fixing mistakes through reconciliation will not result in Congress fixing all mistakes. But, changing the baseline should still substantially increase Congress’ ability to fix mistakes. The change will thus respond more sensibly to the mistakes that are not only inevitable in any legislative process, but also are more likely to occur given the nature of today’s legislative process.

One final objection to this suggested change to reconciliation practices is that it is not ambitious enough. It would only enable legislation that was passed through reconciliation to get fixed in reconciliation. And this is true, as far as it goes. But, as illustrated by recent legislative history, reconciliation is important.291 Congress has been using reconciliation in more and more expansive ways, leading the most recent two administrations to pass signature legislative achievements at least partially through reconciliation.292 In other words, even if this suggested change to reconciliation would not solve every problem with the legislative process, it could be a consequential change. Moreover, when legislation is not passed through reconciliation in the first instance, it is not nearly as problematic to fix. When legislation is passed through ordinary procedures, the Congress that passed it presumably had the higher vote count required to do so, meaning the Congress that passed it should have an easier time getting the votes to make fixes through ordinary procedures. Changes to reconciliation therefore may not only be consequential, but also may appropriately target the conditions that serve as a crux of the problem.

Rather than making changes to reconciliation, an alternative change would be to create a separate procedure whereby a majority vote could pass technical corrections, with technical corrections defined as any changes that do not have a revenue effect, using the original legislation as a baseline. Other limitations might help narrow the definition of technical corrections, for instance by cordoning it to technical mistakes identified within a specified period of time. However, in many ways, this is already supposed to be the practice. For instance, in the tax context, JCT already routinely identifies technical corrections after legislative passage.293

Congress used to routinely pass such technical corrections,294 but the dynamics endemic to today’s legislative dysfunction stymie routine passage of such corrections. Given today’s hyper-partisanship and general legislative dysfunction, Congress members now worry that other members will use technical corrections bills as a way to sneak in nontechnical changes to the prior legislation.295 This level of concern about passing technical corrections makes it unlikely that Congress will more formally lower the vote count for this somewhat vaguely specified category. Relatedly, many new pieces of legislation may not have a revenue effect relative to prior legislation, simply because they do not affect revenue at all. Thus, allowing any changes that do “not affect revenue” relative to prior legislation to be passed by majority vote would essentially kill the supermajority requirement in the Senate, an outcome that could have widespread, destabilizing effects. In contrast, changing the baseline for reconciliation preserves the existing limitations (as manipulable as those limitations might be) on reconciliation, but makes them better suited to address legislative drafting mistakes. The reconciliation proposal thus serves as the most plausible, though surely not a comprehensive or foolproof, way to improve Congress’ ability to fix its own mistakes.

2.     Changes to Agency Practice

Finally, agency legislative fixes inevitably will continue to some extent, notwithstanding changes made to the legislative process. Agencies could try to make these fixes more transparent and systematic. This Article has revealed that, in assessing whether to issue legislative fixes, agencies use criteria for which there is no clear legal authority—such as whether the agency has heard from the “four horsemen” that a fix should be made.296 If agencies believe certain criteria should control decisions about making agency legislative fixes, they could make such criteria transparent and systematize decisions to ensure they comport with such criteria. Indeed, agencies could even do so through a notice-and-comment period, in order to garner the benefits of public input and legitimacy that accompany such procedures.297 Even short of notice and comment, agencies can do more to publicize and justify the agency legislative fixes they make.

The intuition here is that transparent and internally-binding agency guidelines can recreate some of the rule-of-law values that are lost when agencies act in ways that either appear to contravene, or are not otherwise controlled by, governing law. At first blush, agencies overriding statutory dictates through informal guidance, such as FAQs on a website, appears to be completely inconsistent with rule-of-law values. However, as scholars have increasingly identified in other contexts, agencies often have to make decisions that either appear inconsistent with, or at least are outside of, governing law.298 When agencies publicize these decisions in ways that bind agency officials, they may actually increase legitimacy by promoting values such as notice, fairness, and justification that are central to rule of law.299

Moreover, by making the criteria for agency legislative fixes public, agencies can invite accountability into a process that is otherwise shrouded from it. Agency officials are not elected and do not have to answer directly at the ballot box for their administration of the law. This makes agencies’ power to subvert Congress particularly dangerous. However, agencies can reduce this danger by making public the policies used to assess whether to fix a legislative drafting mistake, such as considering Congress member’s own efforts to fix the drafting mistakes. Doing so allows accountable elected officials, as well as the public, to weigh in on and ultimately shape the criteria for decisions.300

Making such policies public also puts members of the public on more equal footing. For instance, well-connected parties would be aware of Treasury’s practice of issuing legislative fixes when Treasury has heard from the four horsemen. Such parties could then direct efforts toward the four horsemen, to get them to pressure Treasury to make an agency legislative fix. If agency policies about legislative fixes are not publicized, less well-connected parties would be systematically disadvantaged. While power and privilege of course affect legislative and regulatory outcomes and cannot be neutralized entirely,301 agencies publicizing their criteria for fixes nonetheless would at least increase the potential for less well-connected parties to make efficacious appeals for fixes.

Relatedly, pre-committing to policies that dictate when to create agency legislative fixes may help ameliorate agency legislative fixes’ negative tendency to systematically advantage regulated parties in informal guidance. For instance, indicating that agency legislative fixes should be provided when certain criteria are met may embolden agencies to offer fixes even when they may be challenged by regulated parties. Even if an agency’s fixes are challenged and ultimately struck down, the agency’s willingness to take action may better encourage Congress to make a needed fix.302

However, while more transparency and systematic decisionmaking around agency legislative fixes may be an improvement in some respects, there would be some accompanying problems with this approach. The first is whether agencies would adopt it. Part of the reason agencies are using formats like FAQs (without any accompanying explanation) is to deal with the matter expeditiously without attracting too much public scrutiny.303 The notion that agencies may voluntarily draw significantly more attention to guidance that seems to directly conflict with the statute may seem somewhat fanciful. It is possible that central executive policy, such as mandates handed down from OIRA to be more forthcoming about agency legislative fixes, might be used to force agencies’ hands and yield more transparency.304 But, if this outcome is achieved, there would be yet another cost. More transparent and systematized agency legislative fixes, perhaps happening under OIRA’s auspices, would also seem to formalize what at least some would think was an affront to Congress’ constitutional power, and even do so by involving higher levels of the executive branch.

This leads to one final, potential change. Aware of its likelihood of making legislative drafting mistakes, Congress could explicitly delegate to agencies the authority to fix such mistakes.305 This approach could, in theory, help embolden agencies to issue legislative fixes through notice-and-comment regulations, potentially even when they disadvantage parties with standing to sue. This outcome would help ameliorate some of the problematic impacts of agency legislative fixes on agency guidance practices. Congress’ explicit delegation to agencies to fix mistakes arguably may help legitimate agencies’ fixes and ease encroachments on legislative supremacy.

Some may argue that this approach would be preferable only in the most formal sense. Such a delegation would beg the question of what constitutes a legislative drafting mistake. Congress’ inherent inability to provide any ex ante guidance about what agencies should fix would arguably recreate all the same problems that already exist with agency legislative fixes. And it may even do so in a way that would raise new constitutional infirmities, such as the problem of excessive delegation.306 Moreover, precisely because it is hard to say what the limits of such a delegation would be, it is unclear whether Congress, fearful of its hard-won legislation being undone, would actually ever explicitly delegate power to agencies to fix legislative drafting mistakes.

But the possibility of such delegations could be better than nothing. Congress could tailor such delegations in ways that might make them useful and more likely to be used by Congress. For instance, rather than just broadly delegating a legislative fix power to agencies in all circumstances, Congress could do so only when legislative passage is particularly harried and mistakes therefore seem particularly likely. Congress could enhance its control by creating timeframes for mistake correction and using other creative innovations to enhance its own power over and participation in the process. Internal executive review (and potentially even judicial review) may benefit from having some standards against which to evaluate an agency’s claim that it is fixing a mistake. Agencies may be able to point to an explicit congressional statement that it was worried about making a mistake in certain circumstances, thus enabling agencies to root fixes in legislative authority.

Even if Congress tailors the fix-it delegation to certain situations, there would be accompanying problems. For instance, explicit delegations to fix mistakes in certain instances may have the undesirable effect of creating an unintended inference against a mistake in situations in which Congress does not so delegate. And when Congress does delegate this power, it may irresponsibly seem to give Congress license to draft poorly. Fundamentally, this proposal’s outside-the-box contemplation of the relationship between Congress and agencies fits uncomfortably with formal views of separation of powers. But understanding the current state of agency legislative fixes suggests that this, and other possibilities discussed above, nonetheless may be improvements over the current threats that agency legislative fixes pose to legislative supremacy, agency legitimacy, and agency guidance practices.

VI.     Conclusion

Evaluating agency legislative fixes and charting a better path forward is no easy task. Agency legislative fixes not only play a critical role in the legislative process, but are also in tension with fundamental constitutional, democratic, and practical values.

This Article has illustrated how, after Congress passes flawed legislation, a dialogue ensues between Congress and agencies that helps motivate agencies to fix legislative drafting mistakes. While these fixes conflict with the text of the legislation, agencies nonetheless make them through the use of informal guidance and other tactics. Agencies do so based in part on a shadowy set of criteria, such as the assent of the “four horsemen,” which sit outside of the legal framework and have escaped attention of existing scholarship.

The most promising path forward is to reorient the baseline for determining eligibility for reconciliation when legislative drafting mistakes have been made. This may put Congress’ ability to correct its mistakes on more equal footing with Congress’ ability to make them. Other changes are possible, each of which offers tradeoffs between different, conflicting values.

In an ideal world, Congress would not make mistakes in passing legislation and, when Congress did make such mistakes, it would fix them. The realities of the legislative process are far from this ideal. We need to recognize, evaluate, and seek to reform the role agency legislative fixes are playing in light of the real-world constraints today’s legislative process presents.

  1. [1]. The official title of the Act is “An Act [t]o provide for reconciliation pursuant to titles II and V of the concurrent resolution on the budget for fiscal year 2018.” Pub. L. No. 115-97, 131 Stat. 2054 (2017) [hereinafter Tax Cuts and Jobs Act]. The title “Tax Cuts and Jobs Act” could not be passed through reconciliation as a result of the application of the Byrd rule. Ellen P. Aprill & Daniel J. Hemel, The Tax Legislative Process: A Byrd’s Eye View, 81 Law & Contemp. Probs. 99, 100 (2018). For ease of exposition, the 2017 tax reform will nonetheless be referred to by its popular abbreviation of “TCJA.”

  2. [2]. Bob Bryan, Paul Ryan Yelled in Delight and Triumphantly Slammed His Gavel as He Announced the GOP Tax Bill’s Passage, Bus. Insider (Dec. 19, 2017, 2:51 PM),
    paul-ryan-gavel-trump-gop-tax-reform-bill-passes-house-2017-12 [].

  3. [3]. Deirdre Walsh et al., White House, GOP Celebrate Passing Sweeping Tax Bill, CNN, [https://] (last updated Dec. 20, 2017, 5:37 PM).

  4. [4]. Jim Tankersley & Alan Rappeport, How Republicans Rallied Together to Deliver a Tax
    , N.Y. Times (Dec. 19, 2017),
    republican-tax-bill.html [].

  5. [5]. Robert W. Wood, IRS Gives Tax Break to Sexual Harassment Victims, Forbes (Mar. 4, 2019, 8:45 AM),
    -harassment-victims [].

  6. [6]. Erica L. Green, Low-Income College Students Are Being Taxed like Trust-Fund Babies,
    N.Y. Times (May 17, 2019), [].

  7. [7]. Kate Sullivan, Tax Overhaul Means Some Gold Star Families Paying Much Higher Taxes on Survivor Benefits, CNN, [] (last updated Apr. 25, 2019, 9:45 PM).

  8. [8]. See, e.g., Brian Faler, ‘This Is Not Normal’: Glitches Mar New Tax Law, Politico (Feb. 24, 2018, 6:50 AM), [] (describing various errors in the TCJA).

  9. [9]. Robert Pear, Four Words that Imperil Health Care Law Were All a Mistake, Writers Now Say, N.Y. Times (May 25, 2015), [] (detailing accounts from members of Congress explaining that the four words in question were a mere drafting mistake).

  10. [10]. Id.

  11. [11]. See Engine Mfrs. Ass’n v. EPA, 88 F.3d 1075, 1085–95 (D.C. Cir. 1996).

  12. [12]. See Dig. Realty Tr., Inc. v. Somers, 138 S. Ct. 767, 778 (2018).

  13. [13]. See, e.g., Nat. Res. Def. Council v. EPA, 489 F.3d 1364, 1371–73 (D.C. Cir. 2007) (examining whether Congress used the word “category” as a stand-in for “category or subcategory” in a way that would allow the EPA to delist subcategories of carcinogens, and concluding that the EPA did not show that Congress did not mean what it said).

  14. [14]. See, e.g., infra text accompanying notes 38–43.

  15. [15]. See infra Section II.C for discussion of this understudied phenomenon.

  16. [16]. See generally Jesse M. Cross, The Staffer’s Error Doctrine, 56 Harv. J. on Legis. 83 (2019) (examining how staffers’ roles in the legislative process should affect judicial treatment of statutes); Ryan D. Doerfler, The Scrivener’s Error, 110 Nw. U. L. Rev. 811 (2016) (examining judicial treatment of legislative drafting mistakes and arguing for more expansive approach); Abbe R. Gluck, Imperfect Statutes, Imperfect Courts: Understanding Congress’s Plan in the Era of Unorthodox Lawmaking, 129 Harv. L. Rev. 62 (2015) (examining options the Court had and the path it took in King v. Burwell); Michael S. Greve & Ashley C. Parrish, Administrative Law Without Congress, 22 Geo. Mason L. Rev. 501 (2015) (examining “de facto” rewrites of statutes by agencies as part of a study of agency deviations from the law, with a focus on the judicial response and doctrines, rather than the agency framework and potential legislative and regulatory solutions); Paul G. Mahoney, Canons of Construction for Dysfunctional Statutes: A Comment on Bennett, 75 Wash. & Lee L. Rev. 577 (2018) (examining how courts should construe dysfunctional statutes); John F. Manning, The Absurdity Doctrine, 116 Harv. L. Rev. 2387 (2005) (examining related judicial application of absurdity doctrine to absurd statutory results).

  17. [17]. See Engine Mfrs. Ass’n v. EPA, 88 F.3d 1075, 1091, 1100–05 (D.C. Cir. 1996).

  18. [18]. See Dig. Realty Tr., Inc. v. Somers, 138 S. Ct. 767, 778–82 (2018).

  19. [19]. See, e.g., Somers v. Dig. Realty Tr. Inc., 850 F.3d 1045, 1049 (9th Cir. 2017) (reasoning that refusing to allow the SEC to correct the potential drafting error would be “illogical” and lead to “absurdity,” and citing King v. Burwell, 135 S. Ct. 2480, 2493 n.3 (2015) in support of its reasoning), rev’d and remanded, 138 S. Ct. 767 (2018)); see also Engine Mfrs. Ass’n, 88 F.3d at 1100
    –05 (Tatel, J., concurring) (rejecting majority approach and instead arguing that the Court should recognize that Congress made a “clerical mistake” in drafting the statute).

  20. [20]. See King, 135 S. Ct. at 2489–95 (explaining that “this [was] not a case for the IRS” and concluding that “an Exchange established by the State” refers to both state and federal exchanges for the purposes of tax credits, in part because the statute would not make sense absent such a reading).

  21. [21]. See, e.g., id. at 2497 (Scalia, J., dissenting) (arguing that “[w]ords no longer have meaning if an Exchange that is not established by a State is ‘established by the State’”); Doerfler, supra note 16, at 843–57 (arguing that reticence to openly correct legislative drafting errors led litigants to make and the Court to adopt “deeply distorted interpretive rationales” in King and other cases).

  22. [22]. See discussion infra Section III.B.

  23. [23]. See infra notes 152–66 and accompanying text (describing the phenomenon in the case of qualified improvement property mistake).

  24. [24]. See sources cited supra note 16.

  25. [25]. A number of recent works have examined legal outgrowths of this problem, see generally Jessica Bulman-Pozen, Preemption and Commandeering Without Congress, 70 Stan. L. Rev. 2029 (2018) (explaining how congressional gridlock makes states a focal point in national policymaking and outlining potential threats to their ability to act); Daniel T. Deacon, Administrative Forbearance, 125 Yale L.J. 1548 (2016) (investigating the case for administrative forbearance authority as a form of congressional delegation); and Jody Freeman & David B. Spence, Old Statutes, New Problems, 163 U. Pa. L. Rev. 1 (2014) (arguing that agencies are best suited to update statutes in the face of congressional inaction).

  26. [26]. The empirical legal literature here is significant and growing, see generally 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 (2014) (exploring, empirically, the factors that drive the legislative drafting process); Victoria F. Nourse & Jane S. Schacter, The Politics of Legislative Drafting: A Congressional Case Study, 77 N.Y.U. L. Rev. 575 (2002) (exploring, empirically, differences in the way the judiciary and the legislature view the drafting process and how that may affect statutory interpretation); Shu-Yi Oei & Leigh Z. Osofsky, Constituencies and Control in Statutory Drafting: Interviews with Government Tax Counsels, 104 Iowa L. Rev. 1291 (2019) (examining, empirically, tax law drafting and formulation decisions); Jarrod Shobe, Intertemporal Statutory Interpretation and the Evolution of Legislative Drafting, 114 Colum. L. Rev. 807 (2014) (examining the steps in statutory drafting and contending that drafting has evolved to become far more specialized); Ganesh Sitaraman, The Origins of Legislation, 91 Notre Dame L. Rev. 79 (2015) (outlining the essential elements of the legislative drafting process); and Christopher J. Walker, Legislating in the Shadows, 165 U. Pa. L. Rev. 1377 (2017) (describing the technical drafting assistance that agencies provide to the legislative drafting process and its potential effects).

  27. [27]. Antonin Scalia, Common-Law Courts in a Civil-Law System: The Role of United States Federal Courts in Interpreting the Constitution and Laws, in A Matter of Interpretation: Federal Courts and the Law 3, 20 (Amy Gutmann ed., 1997).

  28. [28]. See infra text accompanying notes 87–93.

  29. [29]. See, e.g., Stephen K. Cooper, Technical Corrections Bill May Ride on December Tax Extenders, 160 Tax Notes 1162, 1163 (2018) (discussing limited legislative history available in the case of the TCJA).

  30. [30]. Walter J. Oleszek, Congressional Procedures and the Policy Process 14–17 (9th ed. 2014).

  31. [31]. John V. Sullivan, How Our Laws are Made, H.R. Doc. No. 110-49, at 5, 8–9 (2007); Brian Nutting, The Legislative Process, in 1 CQ Press, Guide to Congress 576–77 (7th ed. 2013).

  32. [32]. Sullivan, supra note 31, at 9.

  33. [33]. Nutting, supra note 31, at 576.

  34. [34]. Id. at 580.

  35. [35]. Christopher J. Deering, The Committee System, in 1 CQ Press, Guide to Congress, supra note 31, at 666.

  36. [36]. See Sullivan, supra note 31, at 14–15 (explaining final committee actions); Oleszek, supra note 30, at 30–31 (describing the expertise of lawmakers and committee staff).

  37. [37]. See Sullivan, supra note 31, at 50–51; Nutting, supra note 31, at 576.

  38. [38]. Barbara Sinclair, Unorthodox Lawmaking: New Legislative Processes in the U.S. Congress 166 (5th ed. 2017); see, e.g., Richard H. Pildes, Why the Center Does Not Hold: The Causes of Hyperpolarized Democracy in America, 99 Calif. L. Rev. 273, 276–81 (2011) (describing emergence of hyperpolarized democracy); George K. Yin, Crafting Structural Tax Legislation in a Highly Polarized Congress, 81 Law & Contemp. Probs. 241, 243–45 (2018) (describing evidence of increasing polarization in Congress); Political Polarization in the American Public, Pew Res.
    Ctr. (June 12, 2014), [] (detailing high levels of partisan divides and antipathy).

  39. [39]. See, e.g., Allen Schick, The Federal Budget: Politics, Policy, Process 220–22 (3d ed. 2007) (describing how party polarization has interfered with the ability of the appropriations committees to complete their basic appropriations functions); Sinclair, supra note 38, at 159
    –66 (linking rise in polarization to changes in legislative procedures); David R. Jones, Party Polarization and Legislative Gridlock, 54 Pol. Res. Q. 125, 133–37 (2001) (finding that polarization increases the likelihood of gridlock).

  40. [40]. Schick, supra note 39, at 219; Barbara Sinclair, Party Wars: Polarization and the Politics of National Policy Making 1–35 (2006); John H. Aldrich & David W. Rohde, The Consequences of Party Organization in the House: The Role of the Majority and Minority Parties in Conditional Party Government, in Polarized Politics: Congress and the President in a Partisan Era 31, 33–34 (Jon R. Bond & Richard Fleisher eds., 2000); Yin, supra note 38, at 250.

  41. [41]. Individual legislative products that could not each make their way through Congress may be combined into extremely large, complex “omnibus” bills. See, e.g., James V. Saturno
    & Jessica Tollestrup, Cong. Research Serv., RL32473, Omnibus Appropriations Acts: Overview of Recent Practices 9 (2016), available at
    RL32473.pdf [] (describing the rise of omnibus appropriations acts and the inclusion of significant legislative provisions within them). These omnibus bills are passed through multiple committees, in a way that undermines any one committee’s ability to ensure the functionality of the legislation. Bressman & Gluck, supra note 26, at 760–61.

  42. [42]. See generally Cong. Research Serv., R44362, Post-Committee Adjustment in the Modern House: The Use of Rules Committee Prints (2016), available at https://crsreports. [] (discussing how changes to proposed legislation are often discussed after the legislation is introduced from a committee).

  43. [43]. The TCJA provided a particularly notable example of this phenomenon, as the media widely circulated reports of handwritten, difficult-to-decipher notes being added to the legislation in its final stages. Jim Tankersley & Alan Rappeport, A Hasty, Hand-Scribbled Tax Bill Sets Off
    an Outcry
    , N.Y. Times (Dec. 1, 2017),
    hand-scribbled-tax-bill-outcry.html [].

  44. [44]. Abbe R. Gluck, Anne Joseph O’Connell & Rosa Po, Unorthodox Lawmaking, Unorthodox Rulemaking, 115 Colum. L. Rev. 1789, 1795 (2015) (citing U.S. Const. art. I, § 5, cl. 2 (“Each House may determine the Rules of its Proceedings ....”)).

  45. [45]. R. Shep Melnick, The Conventional Misdiagnosis: Why “Gridlock” Is Not Our Central Problem and Constitutional Revision Is Not the Solution, 94 B.U. L. Rev. 767, 775–81 (2014).

  46. [46]. See, e.g., Sinclair, supra note 38, at 268 (surmising as much); Michael Doran, Legislative Entrenchment and Federal Fiscal Policy, 81 Law & Contemp. Probs. 27, 46–49 (2018) (examining entrenching and its impact on the committee system).

  47. [47]. See, e.g., Victor Fleischer, The State of America’s Tax Institutions, 81 Law & Contemp. Probs. 7, 7, 22 (2018) (lamenting that “[i]deology increasingly carries the day over the influence of experts”).

  48. [48]. Yin, supra note 38, at 255–69.

  49. [49]. Mahoney, supra note 16, at 588 (“[T]he Court is likely to confront more and more cases asking whether the word ‘blue’ actually means ‘orange.’ We should discuss the elephant in the room: Congress no longer legislates carefully, and what it produces is often an incoherent mess.”).

  50. [50]. See, e.g., Robert D. Hershey Jr., A NEW ERA IN AMERICAN TAX POLICY; Congress to Tackle Taxes, Again, N.Y. Times (Mar. 8, 1987), [] (pointing to the technical corrections needed for the Tax Reform Act of 1986, a bipartisan effort enacted with significantly more deliberation than the TCJA).

  51. [51]. Congressional Budget and Impoundment Control Act of 1974, Pub. L. No. 93-344,
    § 310, 88 Stat. 297, 315–16.

  52. [52]. Anita S. Krishnakumar, Note, Reconciliation and the Fiscal Constitution: The Anatomy of the 1995–96 Budget “Train Wreck, 35 Harv. J. on Legis. 589, 591–92 (1998).

  53. [53]. Congressional Budget and Impoundment Control Act § 2; Schick, supra note 39, at 18–19.

  54. [54]. 2 U.S.C. § 632 (2012).

  55. [55]. Schick, supra note 39, at 118–19.

  56. [56]. 2 U.S.C. § 641.

  57. [57]. See, e.g., id. § 641(e)(2) (limiting debate in the Senate on reconciliation bills to not more than 20 hours); Bill Heniff Jr., Cong. Research Serv., RL30862, The Budget Reconciliation Process: The Senate’s “Byrd Rule” 1 (2016), available at https:// [] (describing the expedited procedures); Rebecca M. Kysar, Tax Law and the Eroding Budget Process, 81 Law & Contemp.
    Probs. 61, 67–68 (2018) (explaining how reconciliation is not subject to the filibuster and thus, paradoxically, has likely allowed the filibuster to last longer).

  58. [58]. See Heniff Jr., supra note 57, at 7 (listing budget reconciliation measures enacted or vetoed between 1980 and 2016).

  59. [59]. 131 Cong. Rec. 28,968 (1985) (statement of Sen. Byrd).

  60. [60]. Heniff Jr., supra note 57, at 2.

  61. [61]. James V. Saturno, Cong. Research Serv., 97-865, Points of Order in the Congressional Budget Process 1 (2015), available at
    97-865.pdf [].

  62. [62]. Jay Newton-Small, Alan Frumin: Health Reform’s Reconciliation Ref, TIME (Mar. 3, 2010),,8599,1969267,00.html [

  63. [63]. 2 U.S.C. § 644(e) (2012); Aprill & Hemel, supra note 1, at 105–07.

  64. [64]. 2 U.S.C. § 644(b).

  65. [65]. See, e.g., Aprill & Hemel, supra note 1, at 108–26 (outlining in detail the effects and perversions of the Byrd rule); George K. Yin, Temporary-Effect Legislation, Political Accountability,
    and Fiscal Restraint
    , 84 N.Y.U. L. Rev. 174, 222–24 (2009) (lamenting a particularly perverse application of the Byrd rule, which ultimately resulted in the inclusion of an additional tax cut).

  66. [66]. See Heniff Jr., supra note 57, at 7 (listing budget reconciliation measures enacted into law through 2016). But see, e.g., Aprill & Hemel, supra note 1, at 101–02 (arguing that the Byrd rule may have some benefits, such as serving as an obstacle to special interests or helping to “preserve[] a [negotiating seat] for the minority party”).

  67. [67]. See, e.g., Kysar, supra note 57, at 67 (citing “Democrats[’] controversial[] use[] [of] reconciliation to enact parts of [the ACA’s] major healthcare reform”); Newton-Small, supra note 62 (quoting former Senate parliamentarian Bob Dove’s explanation that the “use of [sidecar] reconciliation [in the ACA was] highly unusual”).

  68. [68]. Kysar, supra note 57, at 69 (citing “unprecedented” use of reconciliation to pass major tax reform in the case of the TCJA).

  69. [69]. Aprill & Hemel, supra note 1, at 136; Cheryl D. Block, Pathologies at the Intersection of the Budget and Tax Legislative Processes, 43 B.C. L. Rev. 863, 882 (2002).

  70. [70]. 2 U.S.C. § 644(e) (2012); Aprill & Hemel, supra note 1, at 105–07.

  71. [71]. Aprill & Hemel, supra note 1, at 105–06; see Heniff, Jr., supra note 57, at 1–2; James V. Saturno, Bill Heniff Jr. & Megan S. Lynch, Cong. Research Serv., R42388, The Congressional Appropriations Process: An Introduction 9 (2016), available at https:// [

  72. [72]. See, e.g., Ida A. Brudnick, Cong. Research Serv., R43532, Offices and Officials
    in the Senate: Roles and Duties 14 (2015), available at
    R43532.pdf [] (describing the role of the Parliamentarian).

  73. [73]. Aprill & Hemel, supra note 1, at 107.

  74. [74]. Newton-Small, supra note 62.

  75. [75]. 2 U.S.C. § 644.

  76. [76]. See, e.g., Halbig v. Sebelius, 27 F. Supp. 3d 1, 24 (D.D.C.), rev’d and remanded sub nom. Halbig v. Burwell, 758 F.3d 390 (D.C. Cir. 2014), reh’g en banc granted and judgment vacated, No. 14-5018, 2014 WL 4627181 (D.C. Cir. Sept. 4, 2014) (discussing how CBO assumptions differed from actual text in the case of the ACA).

  77. [77]. This example actually occurred in the TCJA. See I.R.C. § 24(h)(4) (2018).

  78. [78]. Id. § 152(d)(1)(B).

  79. [79]. Id. § 151(d)(5)(A).

  80. [80]. Id.

  81. [81]. See, e.g., Marc J. Gerson, Technically Speaking: The Art of Tax Technical Corrections, 114 Tax Notes 927, 935 n.79 (2007) (pointing out that “[u]nder the Senate’s so-called Byrd rule, technical corrections generally cannot be included as part of a revenue reconciliation bill because such corrections by definition do not have a revenue impact”); Dustin Stamper, Taxwriters Making Tough Choices on Reconciliation Tax Bills, 109 Tax Notes 714, 716 (2005) (explaining that 60 votes are needed to move technical corrections through the Senate because “the so-called Byrd rule does not allow provisions without revenue implications to move in a reconciliation bill”). Why, exactly, this is the case is generally unarticulated. See infra Section V.C.1 for further discussion of why the results reached under current practices are not clearly required and how they might be changed.

  82. [82]. See, e.g., Business Groups Urge Congress to Support Bill to Fix TCJA Error, Tax Notes Today (Apr. 26, 2019) [hereinafter Letter from Businesses], [] (noting that “current law was scored as if the QIP [qualified improvement property] mistake were not made”).

  83. [83]. See, e.g., Jim Tankersley & Alan Rappeport, G.O.P. Rushed to Pass Tax Overhaul. Now It May Need to Be Altered., N.Y. Times (Mar. 11, 2018),
    11/us/politics/tax-cut-law-problems.html [] (explaining the congressional consensus in the context of the TCJA).

  84. [84]. See, e.g., Jonathan Bernstein, Why the Tax Bill’s Errors Won’t Be Fixed, Bloomberg (Dec. 12, 2017, 4:00 AM), [] (describing this dynamic in the context of ACA and TCJA).

  85. [85]. Id. (describing technical corrections as previously a “routine” part of the process).

  86. [86]. See infra text accompanying notes 241–42 for discussion of when Congress may fix its own mistakes.

  87. [87]. Jasmine C. Lee & Sara Simon, How Every Senator Voted on the Tax Bill, N.Y. Times (Dec. 19, 2017), [].

  88. [88]. The procedural story with the ACA was particularly complicated. Democrats held a supermajority in the Senate when the House and Senate passed separate versions of the legislation in 2009. However, the death of Senator Kennedy and his replacement by Republican Scott Brown meant that the House passed the Senate version of the legislation along with a reconciliation “sidecar” that made the legislation palatable to the House and passable by the Senate through reconciliation, given the Senate’s new political situation. Heniff Jr., supra note 57, at 18.

  89. [89]. See Faler, supra note 8 (quoting expert tax analyst Marty Sullivan, as saying, “[t]his is not normal... . There’s always this kind of stuff, but the order of magnitude is entirely different.”).

  90. [90]. Pear, supra note 9.

  91. [91]. Bernstein, supra note 84; Emily L. Foster, TCJA-ACA Trade-Off Is Possible ‘Glimmer of Hope’ for Tax Fixes, 163 Tax Notes 145, 145–46 (2019) (explaining how the dynamic was the same post-ACA and post-TCJA); Tory Newmyer, The Finance 202: The Next Fight over the GOP Tax Package Is Just Around the Corner, Wash. Post (Dec. 20, 2017),
    powerpost/paloma/the-finance-202/2017/12/20/the-finance-202-the-next-fight-over-the-gop-tax-package-is-just-around-the-corner/5a39549330fb0469e883fcf1/?utm_term=.48d6ddc9f74c [] (discussing the phenomenon of a minority party refusing to take part in fixing the legislative mistakes made by the majority, even before TCJA passage); Tankersley & Rappeport, supra note 83.

  92. [92]. Jad Chamseddine, Technical Corrections Package Likely This Year, 164 Tax Notes Fed. 120, 120 (2019) (quoting Democratic Ways and Means Chair saying that technical corrections are a “pressing matter,” after the most recent failed set of legislative fixes).

  93. [93]. Id.

  94. [94]. U.S. Nat’l Bank of Or. v. Indep. Ins. Agents of Am., Inc., 508 U.S. 439, 462 (1993).

  95. [95]. United States v. X-Citement Video, Inc., 513 U.S. 64, 82 (1994) (Scalia, J., dissenting); see also Doerfler, supra note 16, at 828–34 (describing the narrowness of the judicial approach).

  96. [96]. See Manning, supra note 16, at 2420 n.124 (noting that the two doctrines are sometimes linked); Adam N. Steinman, “Less” Is “More”? Textualism, Intentionalism, and a Better Solution to the Class Action Fairness Act’s Appellate Deadline Riddle, 92 Iowa L. Rev. 1183, 1210 n.158 (2007) (same); see also The Honorable Antonin Scalia & John F. Manning, A Dialogue on Statutory and Constitutional Interpretation, 80 Geo. Wash. L. Rev. 1610, 1614–15 (2012) (describing when absurdity doctrine and scrivener’s error doctrine do and do not overlap).

  97. [97]. Green v. Bock Laundry Mach. Co., 490 U.S. 504, 509–10 (1989).

  98. [98]. Id. at 509, 511 (quoting Campbell v. Greer, 831 F.2d 700, 703 (7th Cir. 1987)).

  99. [99]. Id. at 527 (Scalia, J., concurring).

  100. [100]. King v. Burwell, 135 S. Ct. 2480, 2496 (2015).

  101. [101]. Id. at 2485.

  102. [102]. Id. at 2494.

  103. [103]. Id. at 2489 (quoting 42 U.S.C. § 18031 (2012)).

  104. [104]. Id. at 2488–96.

  105. [105]. See id. The Court did acknowledge Congress’ “inartful drafting” of the ACA and the messy legislative process that had created it, but nonetheless did not address or dispose of the issue as a mistake. Id. at 2492.

  106. [106]. See, e.g., Doerfler, supra note 16, at 843–58 (arguing that judicial disinclination to correct legislative drafting mistakes led to distorted argumentation and judicial decisions in King and other cases).

  107. [107]. King, 135 S. Ct. at 2488–99.

  108. [108]. The parties themselves did not brief the case as one of mistake. This likely reflected perceived discomfort with an issue so major turning on a doctrine of questionable application. Gluck, supra note 16, at 67 (describing the possibility of “a statutory mistake [as] the enormous elephant that neither party dared mention throughout the litigation”).

  109. [109]. Dig. Realty Tr., Inc. v. Somers, 138 S. Ct. 767, 772–73 (2018).

  110. [110]. 15 U.S.C. § 78u-6(a)(6), (h) (2012).

  111. [111]. Id. § 78u-6(a)(6); see also Berman v. Neo@Ogilvy L.L.C., 801 F.3d 145, 151 (2d Cir. 2015) (exploring how “there are categories of whistleblowers who cannot report wrongdoing to the Commission until after they have reported the wrongdoing to their employer”), abrogated by Dig. Realty Tr., Inc. v. Somers, 138 S. Ct. 767 (2018).

  112. [112]. 17 C.F.R. § 240.21F-2(b)(1)(i)–(ii) (2012).

  113. [113]. Somers v. Dig. Realty Tr. Inc., 850 F.3d 1045, 1050–51 (9th Cir. 2017), rev’d and remanded, 138 S. Ct. 767 (2018); Berman, 801 F.3d at 155.

  114. [114]. Asadi v. G.E. Energy (USA), L.L.C., 720 F.3d 620, 630 (5th Cir. 2013).

  115. [115]. See, e.g., Berman, 801 F.3d at 154 (noting that “[w]hen conferees are hastily trying to reconcile House and Senate bills, each of which number hundreds of pages, and someone succeeds in inserting a new provision like subdivision (iii) into subsection 21F(h)(1)(A), it is not at all surprising that no one noticed that the new subdivision and the definition of ‘whistleblower’ do not fit together neatly” (footnote omitted)).

  116. [116]. See Somers, 850 F.3d at 1051 (Owens, J., dissenting) (“In my view, we should quarantine King and its potentially dangerous shapeshifting nature to the specific facts of that case to avoid jurisprudential disruption on a cellular level.”), rev’d and remanded, 138 S. Ct. 767 (2018). The majority and dissenting opinions in the lower courts sparred over King. Compare, e.g., Berman, 801 F.3d at 150, 155 (relying on King in the majority opinion), with id. at 159–60 (Jacobs, J., dissenting) (arguing that “[t]he majority relies almost [exclusively] on [King],” but that King was motivated by “the upending of a ramified, hugely consequential enactment” and did not indicate a broader license to “cast aside [clear] statutory text”).

  117. [117]. Dig. Realty Tr., Inc. v. Somers, 138 S. Ct. 767, 781–82 (2018).

  118. [118]. See, e.g., id. at 777–78 (arguing that “Dodd–Frank’s purpose and design” were consistent with the Court’s reading of the statute).

  119. [119]. Id. at 782 (“The statute’s unambiguous whistleblower definition, in short, precludes the Commission from more expansively interpreting that term.” (citation omitted)).

  120. [120]. See Berman, 801 F.3d at 150 (“The interpretation issue facing the Supreme Court in King was far more problematic than the issue we face here.”).

  121. [121]. See, e.g., Dan McLaughlin, The Supreme Court Proves It Didn’t Mean What It Said in King v. Burwell, Nat’l Rev. (March 28, 2018, 6:30 AM),
    supreme-court-king-v-burwell-decision-political [] (arguing that Somers revealed the truth about King: that it was politically motivated, rather than “a ... sea change in” statutory interpretation).

  122. [122]. Mahoney, supra note 16, at 590.

  123. [123]. For foundational work on congressional control of agency policy outcomes, see generally Mathew D. McCubbins, Roger G. Noll & Barry R. Weingast, Administrative Procedures as Instruments of Political Control, 3 J.L. Econ. & Org. 243 (1987); Mathew D. McCubbins & Thomas Schwartz, Congressional Oversight Overlooked: Police Patrols Versus Fire Alarms, 28 Am. J. Pol. Sci. 165 (1984); and Mathew D. McCubbins, Roger G. Noll & Barry R. Weingast, Structure and Process, Politics and Policy: Administrative Arrangements and the Political Control of Agencies, 75 Va. L. Rev. 431 (1989). For some more recent work, see generally Curtis W. Copeland, Cong. Research Serv., RL34354, Congressional Influence on Rulemaking and Regulation Through Appropriations Restrictions (2008), available at [] (detailing Congress’ methods for influencing agency regulatory activity); Jack M. Beermann, Congressional Administration, 43 San Diego L. Rev. 61 (2006) (drawing on political science literature to detail Congress’ deep involvement in the administration of the law); and Neomi Rao, Administrative Collusion: How Delegation Diminishes the Collective Congress, 90 N.Y.U. L. Rev. 1463 (2015) (fleshing out how individual members of Congress can influence agency exercises of discretion).

  124. [124]. See, e.g., Terry M. Moe & Scott A. Wilson, Presidents and the Politics of Structure, 57 Law & Contemp. Probs. 1, 10 (1994) (explaining that legislators intervene in the administrative process on behalf of specific constituencies, often “behind the scenes”); Rao, supra note 123, at 1482–83 (discussing such informal controls and citing example of Senator Warren’s informal control over the CFPB as well as the practice of Representative Dingell issuing letters seeking information
    —“colloquially ... referred to as ‘Dingell-grams’”—to agencies).

  125. [125]. David van den Berg, ABA Section of Taxation Meeting: Ways and Means Still Assisting Treasury with TCJA Guidance, 159 Tax Notes 1246, 1246 (2018).

  126. [126]. See supra Section II.D. It is also possible that agency guidance that conflicts with the statute may be disallowed during centralized executive review. See infra text accompanying note 304 for discussion of the potential role of OIRA with respect to agency legislative fixes.

  127. [127]. Leigh Osofsky, The Case for Categorical Nonenforcement, 69 Tax L. Rev. 73, 108 (2015); Lawrence Zelenak, Custom and the Rule of Law in the Administration of the Income Tax, 62 Duke L.J. 829, 848–50 (2012).

  128. [128]. See, e.g., David L. Franklin, Legislative Rules, Nonlegislative Rules, and the Perils of the Short Cut, 120 Yale L.J. 276, 308–09 (2010) (examining difficulties for regulatory beneficiaries); Nina A. Mendelson, Regulatory Beneficiaries and Informal Agency Policymaking, 92 Cornell L. Rev. 397, 420–24 (2007) (same); Mark Seidenfeld, Substituting Substantive for Procedural Review of Guidance Documents, 90 Tex. L. Rev. 331, 344 (2011) (same).

  129. [129]. See, e.g., Abbott Labs. v. Gardner, 387 U.S. 136, 148–49 (1967) (“Without undertaking to survey the intricacies of the ripeness doctrine it is fair to say that its basic rationale is to prevent the courts, through avoidance of premature adjudication, from entangling themselves in abstract disagreements over administrative policies, and also to protect the agencies from judicial interference until an administrative decision has been formalized and its effects felt in a concrete way by the challenging parties.” (footnote omitted)), abrogated by Califano v. Sanders, 430 U.S. 99 (1977); Util. Air Regulatory Grp. v. EPA, 320 F.3d 272, 278 (D.C. Cir. 2003) (finding that the EPA’s interpretation in its manual does not create any imminent or redressable injury); Aulenback, Inc. v. Fed. Highway Admin., 103 F.3d 156, 166 (D.C. Cir. 1997) (finding that the Federal Highway Administration training manual which was arguably inconsistent with a statute was not ripe for review); Greve & Parrish, supra note 16, at 532 (discussing how agencies’ turn to guidance documents avoids judicial review); Seidenfeld, supra note 128, at 376–85 (discussing ripeness issues generally, as well as the related judicial doctrine of finality).

  130. [130]. See, e.g., Greve & Parrish, supra note 16, at 512 (“A Congress that enacts massive statutes under frantic conditions is bound to make mistakes—and unlikely to correct them after the fact. Against this backdrop, even ordinarily cautious agencies may resort to de facto rewrites and exotic canons, at least so long as the litigation risks are minimal.”).

  131. [131]. Mike DeBonis & Erica Werner, How Republicans Pulled Off the Biggest Tax Overhaul in 30 Years, Wash. Post (Dec. 20, 2017, 12:20 PM),
    efcba3c4-e54e-11e7-ab50-621fe0588340_story.html [].

  132. [132]. Indeed, in its post-legislation description of the TCJA, known widely as its “Bluebook,” the JCT identified over 70 instances in which technical corrections may be needed for the TCJA. See supra note 89 (commenting on the greater than usual number of mistakes in the TCJA). See generally Staff of Joint Comm. on Taxation, 115th Cong., General Explanation of Public Law 115-97 (J. Comm. Print 2018) (identifying numerous opportunities for technical corrections throughout the Act).

  133. [133]. See, e.g., William Hoffman, TCJA Reg Writers Earn Tax Notes’ 2018 Person of the Year, 161 Tax Notes 1409, 1409 (2018) (describing enormous pressure on Treasury and IRS in the year after the legislation’s passage to ensure its administrability).

  134. [134]. As will be described in the text, the provision addresses deductions for payments related to sexual harassment or sexual abuse. For ease of exposition, the issue will be referred under the umbrella “sexual abuse” term.

  135. [135]. See Robert W. Wood, Ironically, Weinstein Tax on Sexual Harassment Settlements May Hurt Plaintiffs Too, Forbes (Jan. 3, 2018, 8:51 AM),
    2018/01/03/ironically-weinstein-tax-on-sexual-harassment-settlements-may-hurt-plaintiffs-too [].

  136. [136]. I.R.C. § 162(q) (2018).

  137. [137]. Menendez Calls on GOP to Fix Its Tax Bill to Protect Victims of Workplace Sexual Misconduct, Bob Menendez for N.J. (Dec. 21, 2017) [hereinafter Menendez News Release], https:// [].

  138. [138]. See 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, 940 (2013) (exploring how Congress members are involved in legislation at the conceptual level, whereas professional drafters actually craft legislation); Nourse & Schacter, supra note 26, at 585
    –86 (same). See generally Oei & Osofsky, supra note 26 (examining the conceptual/drafting divide in the tax context).

  139. [139]. Kristen A. Parillo, IRS Clarifies TCJA’s Sexual Harassment Provision, 162 Tax Notes 1255, 1255 (2019); Menendez News Release, supra note 137.

  140. [140]. See, e.g., Ryan Sit, Republican Tax Bill Will Hurt Sexual Harassment Victims, Experts Say, Newsweek (Dec. 20, 2017, 3:58 PM), [].

  141. [141]. Menendez News Release, supra note 137.

  142. [142]. Retirement, Savings, and Other Tax Relief Act of 2018, H.R. 88, 115th Cong. § 501(d) (2018).

  143. [143]. Ted Godbout, House Attempts ‘Hail Mary’ with Tax, Retirement Legislation, Nat’l Ass’n Plan Advisors (Dec. 21, 2018), [

  144. [144]. House Ways and Means Comm., 115th Cong., 2d Session, Tax Technical and Clerical Corrections Act Discussion Draft, (Jan. 2, 2019), available at [].

  145. [145]. Id. § 4(r); Staff of Joint Comm. on Taxation, 115th Cong., Technical Explanation of the House Ways and Means Comm. Chairman’s Discussion Draft of the “Tax Technical and Clerical Corrections Act” 9 (Comm. Print 2019).

  146. [146]. See, e.g., Parillo, supra note 139, at 1255–56 (discussing efforts by Congress members on both sides of the aisle to fix the mistake).

  147. [147]. S. 2820, 115th Cong. (2018).

  148. [148]. Section 162(q) FAQ, IRS, [https://] (last updated Jan. 17, 2020).

  149. [149]. Id.

  150. [150]. The examples offered here are emblematic of a broader set of decisions. For instance, as suggested in the text accompanying notes 78–81, another TCJA legislative drafting mistake involved creating a new tax credit for qualifying relatives but simultaneously rendering it inoperable. In the case of the new qualifying relatives tax credit, Treasury released a notice indicating that the reduction of the zero-exemption amount shall not be taken into account for defining a qualifying relative. I.R.S. Notice 2018-70, 2018-38 I.R.B. 441, available at https:// [].

  151. [151]. Sexual abuse victims would not challenge the IRS’ position because it advantaged them. And no one else would have standing to challenge the application of the rule.

  152. [152]. Tax Cuts and Jobs Act, Pub. L. No. 115-97, § 13204(a)(6), 131 Stat. 2054, 2110–11 (2017).

  153. [153]. See, e.g., Nathan J. Richman, ABA Section of Taxation Meeting: Qualified Improvement Property Ripe for Technical Correction Fix, 159 Tax Notes 1236, 1236 (2018).

  154. [154]. See, e.g., id. (explaining that “[t]here is broad agreement that qualified improvement property was excluded from bonus depreciation because of a drafting error”).

  155. [155]. See id.

  156. [156]. See, e.g., David van den Berg, Coalition of Companies and Groups Seeks Technical Fixes for TCJA, 159 Tax Notes 1667, 1667–68 (2018) (pointing to various constituencies seeking correction).

  157. [157]. See, e.g., Richman, supra note 153, at 1236.

  158. [158]. Id.

  159. [159]. 83 Fed. Reg. 39,292, 39,293 (Aug. 8, 2018). Since the effective date of the TCJA’s new depreciation rules applied to property placed in service after December 31, 2017, Treasury was able to preserve the former, more favorable treatment for qualified improvement property for property placed in service prior to January 1, 2018. Id.

  160. [160]. Finance Committee Republicans Clarify Intent of 3 TCJA Sections, Tax Notes Today (Aug. 16, 2018) [hereinafter Finance Committee Republicans],
    depreciation-amortization-and-depletion/finance-committee-republicans-clarify-intent-3-tcja-sections/2018/08/17/28bxz []. This was far from the only time that Congress had sent such a letter in an effort to get an agency legislative fix. See, e.g., Income Attributable to Domestic Production Activities, 70 Fed. Reg. 67,220, 67,222 (Nov. 4, 2005) (letter to the Secretary of the Treasury from the House Ways and Means Chair, the Senate Finance Committee Chair, and the Senate Finance Committee ranking member providing clarification of congressional intent “so that appropriate regulatory guidance may be issued reflecting [their] intention”).

  161. [161]. Finance Committee Republicans, supra note 160.

  162. [162]. Id.

  163. [163]. Nathan J. Richman, Regulators Anticipate Legislative Changes to Bonus Depreciation, 160 Tax Notes 1304, 1304 (2018).

  164. [164]. Id.

  165. [165]. Cf. 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, 187–91 (2017) (suggesting “that an ambiguous statut[ory provision] should be [interpreted] in [a manner consistent] with the ... ‘budget score’”).

  166. [166]. Senators Look to Treasury for TCJA Depreciation Fix, Tax Notes Today (Sept. 24, 2018),
    %20to%20Treasury%20for%20TCJA%20Depreciation [].

  167. [167]. Nathan J. Richman, Treasury: Qualified Improvement Property Needs Technical Correction, Tax Notes Today (Oct. 9, 2018),
    28hbq?highlight=Qualified%20Improvement%20Property%20Needs%20Technical%20Correction [].

  168. [168]. Retirement, Savings, and Other Tax Relief Act of 2018, H.R. 88, 115th Cong. § 501(b) (2018).

  169. [169]. House Ways and Means Comm., 115th Cong., 2d Session, Tax Technical and Clerical Corrections Act Discussion Draft, (Jan. 2, 2019), available at [].

  170. [170]. Restoring Investment in Improvements Act, S. 803, 116th Cong. § 2 (2019).

  171. [171]. Restoring Investment in Improvements Act, H.R. 1689, 116th Cong. § 2 (2019).

  172. [172]. See Foster, supra note 91, at 146 (suggesting that changes may not come until Congress has to pass must-pass legislation such as legislation addressing the debt ceiling and even reporting that TCJA technical corrections may eventually be made only in exchange for still-unmade technical corrections for the ACA). These prognostications ended up being right. As this Article went to press in the spring of 2020, Congress finally passed the qualified improvement property fix as part of the Coronavirus Aid, Relief and Economic Security (CARES) Act. CARES Act, S. 3548, 116th Cong. § 2207 (2020).

  173. [173]. While the too-broad cross-reference to “this chapter” in § 162(q) created the problem in the case of the sexual abuse provision, it could be corrected in a number of ways including, for instance, adding a narrowing clause to the provision addressing attorney fees. See, e.g., House Ways and Means Comm., 115th Cong., 2d Session, Tax Technical and Clerical Corrections Act Discussion Draft (adopting such an approach).

  174. [174]. The August 16th, 2018 letter from Republicans on the Senate Finance Committee identified qualified improvement property and sexual abuse lawyer fees as two of the three issues in need of technical corrections. Finance Committee Republicans, supra note 160.

  175. [175]. Richman, supra note 167.

  176. [176]. See, e.g., I.R.C. § 168(c) (2018) (listing nonresidential real property as having a recovery period of 39 years, a default applicable to qualified improvement property absent its inclusion in some other class of property).

  177. [177]. See supra notes 135–41.

  178. [178]. Section 162(q) FAQ, supra note 148 (emphasis added).

  179. [179]. I.R.C. § 168(c).

  180. [180]. See, e.g., id. § 168(b)(3) (listing separately nonresidential real property and qualified improvement property as categories of property to which the straight line method of depreciation applies).

  181. [181]. See, e.g., Gluck & Bressman, supra note 138, at 972 (examining Congress’ use of legislative history as a means of influencing agency implementation of statutes); Christopher J. Walker, Inside Agency Statutory Interpretation, 67 Stan. L. Rev. 999, 1034–48 (2015) (examining agency use of legislative history).

  182. [182]. The conference committee report simply describes, almost verbatim, the legislative provision and then notes that “[t]he conference agreement follows the Senate amendment.” H.R. Rep. No. 115-466, at 431 (2017) (Conf. Rep.).

  183. [183]. Id. at 366–67.

  184. [184]. Joint Explanatory Statement of the Committee of Conference to H.R. 1, “Tax Cuts and Jobs Act” 205 (Dec. 2017), available at
    20171218/JointExplanatoryStatement.pdf [].

  185. [185]. See Section 162(q) FAQ, supra note 148 (answering the frequently asked question of whether “section 162(q) preclude[s] me from deducting my attorney’s fees related to the settlement of my sexual harassment claim if the settlement is subject to a nondisclosure agreement”).

  186. [186]. See cases cited supra note 129.

  187. [187]. Nathan J. Richman, GOP Taxwriters Ask Treasury to Accommodate 3 Technical Corrections, 160 Tax Notes 1161, 1161 (2018) [hereinafter Richman, GOP Taxwriters]; Nathan J. Richman, Practitioners See More Qualified Improvement Property Fix Options, 160 Tax Notes 1618, 1618 (2018) [hereinafter Richman, Practitioners].

  188. [188]. I.N.S. v. Chadha, 462 U.S. 919, 951 (1983).

  189. [189]. Gerson, supra note 81, at 931.

  190. [190]. Id.

  191. [191]. See Michael Cohn, Congressional Staff Aims to Finish Technical Corrections to Tax Reform
    , Acct. Today (Apr. 25, 2018, 5:53 PM),
    congressional-staff-aims-to-finish-technical-corrections-to-tax-reform-bill [
    N7A8-CHK6] (“[M]any taxpayers are awaiting guidance from the Treasury Department and the Internal Revenue Service while a technical corrections bill is being prepared.”).

  192. [192]. U.S. Const. art. I, § 1 (“All legislative Powers herein granted shall be vested in a Congress of the United States, which shall consist of a Senate and House of Representatives.”).

  193. [193]. Cooper, supra note 29, at 1162.

  194. [194]. See, e.g., Catharine A. MacKinnon, Where #MeToo Came from, and Where It’s Going: The Movement Is Moving the Culture Beneath the Law of Sexual Abuse, Atlantic (Mar. 24, 2019), [] (describing the rise and trajectory of the #MeToo movement).

  195. [195]. See I.R.C. § 62(a)(20) (2018).

  196. [196]. See generally Leandra Lederman, IRS Reform: Politics as Usual?, 7 Colum. J. Tax L. 36 (2016) (describing historical and recent lack of support for the IRS and the negative consequences).

  197. [197]. Cf. Sullivan, supra note 7 (quoting Treasury response to the Gold Star families mistake from the TCJA, which was, “[w]e are deeply grateful for the sacrifices made by our brave service members and their families and we are evaluating what can be done to solve this issue and provide relief to the families of our fallen heroes”).

  198. [198]. U.S. Const. art. I, § 1 (“All legislative Powers herein granted shall be vested in a Congress of the United States, which shall consist of a Senate and House of Representatives.”).

  199. [199]. Lamie v. U.S. Trustee, 540 U.S. 526 (2004).

  200. [200]. Id. at 530.

  201. [201]. Id. at 539–40.

  202. [202]. Id. at 538 (quoting United States v. Locke, 471 U.S. 84, 95 (1985)); see also King v. Burwell, 135 S. Ct. 2480, 2505 (2015) (Scalia, J., dissenting) (arguing that the notion “that judges should endure whatever interpretive distortions it takes in order to correct a supposed flaw in the statutory machinery... . [I]gnores the American people’s decision to give Congress ‘[a]ll legislative Powers’ enumerated in the Constitution. They made Congress, not this Court, responsible for both making laws and mending them.” (citation omitted)).

  203. [203]. See Daniel E. Walters, Litigation-Fostered Bureaucratic Autonomy: Administrative Law Against Political Control, 28 J.L. & Pol. 129, 130 (2013) (explaining both how “many executive agencies are formally housed in the executive branch and are, at least in theory, subject to the control of political appointees (who are themselves directly accountable only to the executive branch and removable at will)” and how the reality may differ from this formal story of executive control over agency action (footnote omitted)).

  204. [204]. See, e.g., Meagan Flynn, ‘Subvert the Constitution’: Trump’s 2014 Remarks on Obama’s Executive Actions Show ‘Hypocrisy,’ Critics Say, Wash. post (Feb. 15, 2019, 5:59 AM), [] (comparing Obama’s actions on immigration with Trump’s actions on the border wall); Kevin M. Kruse & Julian E. Zelizer, Have We Had Enough of the
    Imperial Presidency Yet?
    , N.Y. Times (Jan. 9, 2019),
    opinion/president-trump-border-wall-weak.html [] (arguing that Democratic and Republican presidents alike have overused executive power). See generally Kathryn A. Watts, Controlling Presidential Control, 114 Mich. L. Rev. 683 (2016) (examining expansion of presidential control, in particular under the George W. Bush and Barack Obama presidencies).

  205. [205]. See, e.g., Deacon, supra note 25, at 1564 (explaining that delegations of power to agencies are often examined based “in part on a set of functional considerations that depend on a particular understanding of the roles played by Congress and agencies in the policymaking process” (footnote omitted)); Rafael I. Pardo & Kathryn A. Watts, The Structural Exceptionalism of Bankruptcy Administration, 60 UCLA L. Rev. 384, 423 (2012) (exploring the “functionalist considerations that make agencies well suited to receive and exercise delegations of policymaking power”); cf. Peter L. Strauss, The Place of Agencies in Government: Separation of Powers and the Fourth Branch, 84 Colum. L. Rev. 573, 616 (1984) (describing a functional approach to separation of powers as one that “assess[es] structural arrangements ... in terms of their contribution to or detraction from the maintenance of tensions among the named branches”).

  206. [206]. See Deacon, supra note 25, at 1564–65.

  207. [207]. Christina L. Lyons, Congressional Staff, in 1 CQ Press, Guide to Congress, supra note 31, at 698–700. See generally Cong. Research Serv., RL34545, Congressional Staff: Duties and Functions of Selected Positions (2012), available at
    pdf/RL/RL34545/7 [] (describing various staff positions).

  208. [208]. See supra note 137 and accompanying text.

  209. [209]. See supra note 137 and accompanying text.

  210. [210]. Cross, supra note 16, at 85.

  211. [211]. See supra notes 173–74 and accompanying text.

  212. [212]. See, e.g., Gluck, supra note 16, at 76–77 (explaining how the problematic four words central to the King litigation were erroneously taken from the Senate Finance Committee draft, thereby failing to make the necessary drafting changes to dovetail with the portion of the legislation that came from the Senate Health, Education, Labor and Pensions Committee).

  213. [213]. See generally Dig. Realty Tr., Inc. v. Somers, 138 S. Ct. 767 (2018) (finding that the definition of whistleblower was unambiguous).

  214. [214]. Oei & Osofsky, supra note 26, at 1324.

  215. [215]. Cf. Louis Kaplow, Rules Versus Standards: An Economic Analysis, 42 Duke L.J. 557, 577 (1992) (concluding that “[r]ules cost more to promulgate” than standards); David A. Weisbach, Formalism in the Tax Law, 66 U. Chi. L. Rev. 860, 871 (1999) (explaining that “as tax rules become complex, interaction costs increase rapidly and rules quickly become unmanageable”).

  216. [216]. See, e.g., Thomas W. Merrill, Rethinking Article I, Section 1: From Nondelegation to Exclusive Delegation, 104 Colum. L. Rev. 2097, 2141 (2004) (raising democratic accountability concerns with Congress delegating decisions to agencies); Rao, supra note 123, at 1491–506 (positing that delegations to agencies fracture the collective Congress and yield various threats to separation of powers and related values); David Schoenbrod, Delegation and Democracy: A Reply to My Critics, 20 Cardozo L. Rev. 731, 744–45 (1999) (same). See generally David Epstein & Sharyn O’Halloran, Delegating Powers: A Transaction Cost Politics Approach to Policy Making Under Separate Powers (1999) (providing theory and literature review of congressional delegation).

  217. [217]. U.S. Const. art. 1, § 7; I.N.S. v. Chadha, 462 U.S. 919, 949–59 (1983) (striking down assertion of unicameral veto power).

  218. [218]. See supra notes 123–24 for discussion of formal and informal mechanisms of control. See generally Rao, supra note 123 (discussing these issues in the context of agency exercises of delegated power).

  219. [219]. The Federalist No. 48 (James Madison) (worrying over congressional aggrandizement of its powers); The Federalist No. 51, at 280–84 (James Madison) (J.R. Pole ed., 2005) (discussing separation of powers); cf. Frank H. Easterbrook, Statutes’ Domains, 50 U. Chi. L. Rev. 533, 539 (1983) (“Under article I of the Constitution, not to mention the rules of the chambers of Congress, support is not enough for legislation.”).

  220. [220]. Pear, supra note 9.

  221. [221]. Id.

  222. [222]. Id.

  223. [223]. See Oei & Osofsky, supra note 26, at 1318.

  224. [224]. See id. As anecdotal evidence of this phenomenon, there was a notable path from drafting the TCJA to private sector employment. See Cooper, supra note 29, at 1163 (pointing out that many of the chief staff architects of the TCJA had left for private industry); Asha Glover,
    Tax Staffers Move on from Capitol After Tax Reform, Tax Notes Today (May 23, 2018),
    n%20from%20Capitol%20After%20Reform [].

  225. [225]. See, e.g., Jarrod Shobe, Agency Legislative History, 68 Emory L.J. 283, 285–86, 291 (2018) (exploring generally how there is a “blurred line” between the legislative and regulatory processes, as agency officials are “intimately involved in drafting, revising, and negotiating legislation”); Walker, supra note 26, at 1388–89 (describing working relationships between congressional staffers and agency officials).

  226. [226]. Oei & Osofsky, supra note 26, at 1353–55 (questioning how to legitimate staffers’ extensive role in the legislative drafting process).

  227. [227]. See, e.g., Shobe, supra note 26, at 829 (noting how Legislative Counsel is careful to note that its job is to implement Congress’ policy decisions).

  228. [228]. See, e.g., Jarrod Shobe, Agencies as Legislators: An Empirical Study of the Role of Agencies in the Legislative Process, 85 Geo. Wash. L. Rev. 451, 499 (2017) (pointing out that “the vast majority of agency staffers are career employees who spend the bulk of their career working for one agency,” and that such agency staffers often have even greater knowledge about their governing legislation than Congress does).

  229. [229]. See, e.g., Evan J. Criddle, Fiduciary Foundations of Administrative Law, 54 UCLA L. Rev. 117, 123 (2006) (describing agency legitimacy as “a contested concept resting on other contested concepts”).

  230. [230]. Mendelson, supra note 128, at 417.

  231. [231]. For further discussion of these and other models and the evolution of views regarding agency legitimacy over time, see, for example, Lisa Schultz Bressman, Beyond Accountability: Arbitrariness and Legitimacy in the Administrative State, 78 N.Y.U. L. Rev. 461, 469–91 (2003).

  232. [232]. See, e.g., Richard B. Stewart, The Reformation of American Administrative Law, 88 Harv. L. Rev. 1667, 1678 (1975) (describing the expertise model).

  233. [233]. See, e.g., id. (describing that, under the expertise model, “the discretion that the administrator enjoys is more apparent than real”).

  234. [234]. In the case of Congress, they may correct many in the long-run, as political winds eventually allow opportunities to do so.

  235. [235]. See supra text accompanying notes 100–08.

  236. [236]. See infra text accompanying notes 251–58.

  237. [237]. See, e.g., Stewart, supra note 232, at 1683 (“Today, the exercise of agency discretion is inevitably seen as the essentially legislative process of adjusting the competing claims of various private interests affected by agency policy.” (footnote omitted)).

  238. [238]. See sources cited supra note 139.

  239. [239]. See, e.g., Chevron, U.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 865 (1984) (“While agencies are not directly accountable to the people, the Chief Executive is, and it is entirely appropriate for this political branch of the Government to make such policy choices
    ....”); Eric A. Posner & Cass R. Sunstein, Chevronizing Foreign Relations Law, 116 Yale L.J. 1170, 1202 (2007) (“As emphasized in Chevron, the executive, unlike the judiciary, is politically accountable as well as uniquely knowledgeable, and its accountability argues for deference to its judgments about how to assess the competing facts and values.” (footnote omitted)).

  240. [240]. See, e.g., Jonathan R. Siegel, The Constitutional Case for Chevron Deference, 71 Vand. L. Rev. 937, 946–55 (2018) (cataloguing the growing chorus of attacks on Chevron from members of the judiciary, Congress, and academia).

  241. [241]. Compare, e.g., Scialabba v. Cuellar de Osorio, 573 U.S. 41, 75 (2014) (plurality opinion) (concluding that Chevron deference is meant precisely to ensure that courts defer to agencies’ reasonable constructions of self-contradictory provisions in a complex statutory scheme), with id. at 76 (Roberts, J., concurring) (“Direct conflict is not ambiguity, and the resolution of such a conflict is not statutory construction but legislative choice. Chevron is not a license for an agency to repair a statute that does not make sense.” (footnote omitted)).

  242. [242]. See, e.g., Engine Mfrs. Ass’n v. EPA, 88 F.3d 1075, 1104 (D.C. Cir. 1996) (Tatel, J., dissenting in part) (arguing that “[b]ecause the court’s reading of a portion of the statute establishes a bizarre and unprecedented regulatory regime that conflicts with other provisions of the text, undermines the statute’s purpose and structure, finds no support in the legislative history, and produces a result that serves no apparent legislative purpose, and because a perfectly plausible alternative explanation exists,” the court should have found enough ambiguity to trigger deference to the agency’s logical and simple alternative).

  243. [243]. See, e.g., Maureen B. Cavanaugh, Order in Multiplicity: Aristotle on Text, Context, and the Rule of Law, 79 N.C. L. Rev. 577, 585 n.25 (2001) (“The literature attempting to explicate theories of statutory interpretation is, however, voluminous.”); Peter L. Strauss, “Deference” Is Too Confusing
    —Let’s Call Them “
    Chevron Space” and “Skidmore Weight, 112 Colum. L. Rev. 1143, 1144 (2012) (“Administrative law scholars have leveled a forest of trees exploring the mysteries of the Chevron approach contemporary judges take to reviewing law-related aspects of administrative action.” (footnote omitted)).

  244. [244]. See supra text accompanying notes 194–97.

  245. [245]. See Finance Committee Republicans, supra note 160.

  246. [246]. See, e.g., Richman, Practitioners, supra note 187, at 1618–20 (cataloguing various practitioners’ suggestions, including having central IRS agents tell the IRS examiners not to raise the issues on exam).

  247. [247]. See, e.g., Lisa Schultz Bressman, Judicial Review of Agency Inaction: An Arbitrariness Approach, 79 N.Y.U. L. Rev. 1657, 1691 (2004) (discussing problems with informal or internal agency policies of nonenforcement); Gillian E. Metzger, The Constitutional Duty to Supervise, 124 Yale L.J. 1836, 1919–20 (2015) (same); Osofsky, supra note 127, at 102–03 (same); Peter L. Strauss, Publication Rules in the Rulemaking Spectrum: Assuring Proper Respect for an Essential Element, 53 Admin. L. Rev. 803, 808 (2001) (same).

  248. [248]. See supra text accompanying notes 126–30.

  249. [249]. See, e.g., Tom C. Clark, U.S. Dep’t of Justice, Attorney General’s Manual on the Administrative Procedure Act 26 (1947), available at
    downloads/AttorneyGeneralsManual.pdf [] (explaining that the purpose of the rulemaking procedures “is to guarantee to the public an opportunity to participate”); 1 Richard J. Pierce, Jr., Administrative Law Treatise § 6.8 (4th ed. 2002) (describing the benefits of administrative rulemaking including, among others, political accountability).

  250. [250]. See, e.g., Hoctor v. U.S. Dep’t of Agric., 82 F.3d 165, 171 (7th Cir. 1996) (discussing the need for rulemaking procedures to legitimize agency legislative decisions).

  251. [251]. Robert A. Anthony, Interpretive Rules, Policy Statements, Guidances, Manuals, and the Like
    —Should Federal Agencies Use Them to Bind the Public?
    , 41 Duke L.J. 1311, 1317–18 (1992) (discussing the values lost when an agency uses informal guidance to actually or practically bind the public). But see, e.g., Metzger, supra note 247, at 1919–20 (explaining that informal “guidance is a crucial part of agency efforts to fulfill their internal oversight responsibilities and curtail lower-level discretion” (footnote omitted)).

  252. [252]. See supra Part II.

  253. [253]. This was true of the critical drafting mistake that threatened to imperil the ACA. By the time members of the public had identified the mistake, the new Republican control of the House of Representatives seemingly foreclosed any paths to fixing the mistake. See infra text accompanying note 286.

  254. [254]. See supra Section IV.B; cf. Hershey Jr., supra note 50 (pointing out that, after the passage of the last major tax reform, the Tax Reform Act of 1986, there was also a significant delay in passing many of the needed technical corrections because of concerns that attempting to pass them would re-enmesh Congress in substantive tax policy debates).

  255. [255]. See, e.g., Asha Glover, Senate, House Agree on Need for Technical Corrections, 163 Tax Notes 1580, 1580 (2019) (describing this dynamic in the post-TCJA context). Indeed, the much-delayed passage of a qualified improvement property fix as part of the emergency CARES Act legislation illustrates this dynamic particularly well. See supra note 172 for a discussion of the CARES Act.

  256. [256]. Faler, supra note 8.

  257. [257]. See, e.g., Bob Bryan, Republicans Made a Huge Mistake in Their Tax Bill—And They’re Scrambling to Fix It, Bus. Insider (Mar. 21, 2018, 11:44 AM),
    grain-glitch-tax-bill-mistake-omnibus-bill-fix-2018-3 [] (explaining the problem).

  258. [258]. Consolidated Appropriations Act of 2018, Pub. L. No. 115-141, § 101, 132 Stat. 348; see Saturno, Heniff Jr. & Lynch, supra note 71 (describing the appropriations process generally); Gluck, O’Connell & Po, supra note 44, at 1832 (discussing how Congress sometimes adds substantive legislation as riders on must-pass appropriations bills, to make it harder to defeat the riders).

  259. [259]. Andrew Khouri, Boost to Affordable Housing is Part of New $1.3-Trillion Federal Budget, L.A. Times, [] (last updated Mar. 23, 2018, 2:30 PM).

  260. [260]. See, e.g., Ryan D. Doerfler, High-Stakes Interpretation, 116 Mich. L. Rev. 523, 534 (2018) (exploring the well-known proposition that legislation depends on compromise).

  261. [261]. Jad Chamseddine, Bill to Fix Retail Glitch in TCJA Gets House Companion, 163 Tax Notes 147, 147 (2019).

  262. [262]. Richman, GOP Taxwriters, supra note 187, at 1161.

  263. [263]. 5 U.S.C. § 553 (2012).

  264. [264]. See, e.g., Sprint Corp. v. FCC, 315 F.3d 369, 373–74 (D.C. Cir. 2003) (noting instrumental function of notice-and-comment procedures in the rulemaking context).

  265. [265]. Gavin Young, Statutory Review Program, Admin. Conf. U.S.,
    research-projects/statutory-review-program [].

  266. [266]. Fleischer, supra note 47, at 20 (articulating the playbook of keeping partisan legislation secret as long as possible and then passing it as quickly as possible to beat out the opposition); Oei & Osofsky, supra note 26, at 1326 (discussing the need to keep bills secret from lobbyists to ensure passage); see also, e.g., Schick, supra note 39, at 165–66 (describing the House Ways and Means Committee’s practice of striking legislative deals in private prior to publicly convening to approve them to stave off lobbyist interference).

  267. [267]. For one assertion that this was the case with the TCJA, see Catherine Rampell, Why Are Republicans in Such a Rush to Pass Tax Reform? To Outrun the Truth., Wash. Post (Nov. 27, 2017),
    term=.b2e2ae97c852 [].

  268. [268]. See, e.g., Shobe, supra note 26, at 840 (describing the careful back and forth between legislative counsel and American Law Division attorneys to ensure that provisions are drafted correctly).

  269. [269]. Eric A. Posner & Adrian Vermeule, Inside or Outside the System?, 80 U. Chi. L. Rev. 1743, 1745 (2013).

  270. [270]. See, e.g., Richard S. Markovits, The Case for “Business as Usual” in Law-and-Economics Land: A Critical Comment, 78 Iowa L. Rev. 387, 387 (1993) (providing an alternative explanation of the theory of the second best).

  271. [271]. See supra Part II.

  272. [272]. See supra notes 75–76, 81–83 and accompanying text.

  273. [273]. See supra Section II.C.

  274. [274]. See, e.g., Alan J. Auerbach, US Experience with Federal Budget Rules, 7 CESifo DICE Rep.: J. for Institutional Comparisons 41, 42–43 (2009) (describing briefly the federal budget rules over time and how and why they have shifted); Rebecca M. Kysar, Lasting Legislation, 159 U. Pa. L. Rev. 1007, 1023–24 (2011) (describing generally the endogeneity of legislative rules).

  275. [275]. See, e.g., Daniel N. Shaviro, Taxes, Spending, and the U.S. Government’s March Toward Bankruptcy 159 (2007) (“The choice of a baseline is inevitably arbitrary, or at least subject to differing interpretations.”).

  276. [276]. See supra Section II.C.

  277. [277]. Aprill & Hemel, supra note 1, at 104–08, 122–25, 131–32, 133–35 (discussing opacity and subjectivity of the application of the Byrd rule); David Kamin & Rebecca Kysar, Temporary Tax Laws and the Budget Baseline, 157 Tax Notes 125, 126 (2017); Yin, supra note 65, at 221.

  278. [278]. For scoring purposes, Congress looks to CBO estimates, which rely on the data prepared by the JCT. Schick, supra note 39, at 170. Although it would seem that the Senate Parliamentarian would have to be the one to make the change, Congress has shown itself capable of getting what it wants when its view conflicts with the Senate Parliamentarian’s. See id. at 148
    –49 (describing firing of Senate Parliamentarian as a result of dispute over reconciliation procedures); Newton-Small, supra note 62 (detailing back and forth between different parliamentarians to allow Congress to get what it wanted).

  279. [279]. See, e.g., Jeff Davis, The Rule that Broke the Senate, Politico Mag. (Oct. 15, 2017), https:// [] (forcefully making this argument). However, as a contrary point of view, some have argued that the limitations on the ability to use reconciliation are arbitrary and can have deleterious effects on legislation. See, e.g., 139 Cong. Rec. 19325 (Remarks of Rep. Rostenkowski) (“As a result of [the Byrd Rule], policies that would have significantly improved the Medicare Program could not even be considered.... Even more absurd is the fact that most of the items stripped were minor and technical provisions that received bipartisan support when they passed both the House and the Senate last year.”).

  280. [280]. See, e.g., David Kamin, Basing Budget Baselines, 57 Wm. & Mary L. Rev. 143, 153–60 (2015) (summarizing the history of budgeting baselines and concluding that “[i]t is a history not just of widespread use but also of confusion and abuse”); cf. Aprill & Hemel, supra note 1, at 102 (arguing that “the Byrd rule preserves a role for the minority party in the Senate, potentially promoting a more consensus-oriented approach to lawmaking”).

  281. [281]. See, e.g., Yin, supra note 65, at 225–26 (lamenting lack of transparency into Byrd rule determinations). But see, e.g., Elizabeth Garrett & Adrian Vermeule, Transparency in the U.S. Budget Process, in Fiscal Challenges: An Interdisciplinary Approach to Budget Policy 75–80 (Elizabeth Garrett, Elizabeth A. Graddy & Howell E. Jackson eds., 2008) (positing that opacity can have some benefits such as stymieing interest group influence).

  282. [282]. Most notably, some scholars have argued that shifting baselines allow Congress to pass permanent tax cuts without ever fully taking into account the cost of such cuts. Kamin & Kysar, supra note 277, at 130–32; Kysar, supra note 274, at 1028–33. Congress does so by first passing temporary tax cuts, which expire at the end of the budget window. By making the tax cuts temporary, Congress is able to argue initially that the tax cuts will not have a revenue cost outside of the budget window. Then, when the tax cuts are about to expire, proponents of the tax cuts argue that, since the tax cuts have been in effect for quite some time, the scoring baseline should assume that the tax cuts are permanent. The effect is that shifting baselines allows Congress to pass tax cuts it never would have been able to in the first instance, had it been forced to fully take into account the cost of such cuts. Kamin & Kysar, supra note 277, at 130–32; Kysar, supra note 274, at 1028–33.

  283. [283]. To be sure, Congress is likely to find some unanticipated ways to abuse any budgeting rule. However, the mere possibility of abuse should not prevent us from moving toward the most sensible rule in a given context.

  284. [284]. The Budget Enforcement Act, which controls at least some aspects of budgeting procedures, states that the baseline shall be calculated under the assumption that “[l]aws providing or creating direct spending and receipts are assumed to operate in the manner specified in those laws for each such year.” Omnibus Budget Reconciliation Act of 1990, Pub. L. No.
    101-508, § 257, 104 Stat. 1388–591 (emphasis added). This baseline assumption would seem to call for applying the law as enacted, not the law as (incorrectly) scored. See Kamin, supra note 280, at 188–91 (explaining how the assumptions underlying the Budget Enforcement Act control in certain circumstances, but how CBO and OMB have also deviated from them in various ways); see also, e.g., Schick, supra note 39, at 61 (explaining how, generally, the baseline is “the amount that would be spent if current law continues in effect without change.”); Yin, supra note 65, at 185–87 (discussing the current law baseline).

  285. [285]. Cf. Samarth Desai, The Senator’s Dilemma: Game Theory, Gorsuch, and the Nuclear Option, Harv. Pol. Rev. (Apr. 8, 2017), [] (explaining future cost to Republicans of having eliminated the filibuster for Supreme Court confirmations).

  286. [286]. Of course, the change would eliminate the ability of the minority party to extract high concessions in exchange for agreeing to fixes. However, which way this cuts is not clear, as, in the abstract, each party could one day imagine being in the position of the minority party.

  287. [287]. Election 2010: House Map, N.Y. Times,
    results/house.html []; U.S. House Election Results 2018, N.Y. Times (May 15, 2019, 2:08 PM),
    results-house-elections.html [].

  288. [288]. See, e.g., Megan S. Lynch & James V. Saturno, Cong. Research Serv., R44058,
    The Budget Reconciliation Process: Stages of Consideration 2 (2017), available at [] (explaining steps in process); David Reich & Richard Kogan, Introduction to Budget “Reconciliation,Ctr. on Budget & Pol’y Priorities, [] (last updated Nov. 9, 2016) (explaining that the Senate can only consider one reconciliation bill a year with respect to each of the three basic reconciliation subjects (spending, revenues, and debt limit), unless Congress passes a new budget resolution); Yin, supra note 65, at 219 (pointing to difficulty in passing a budget as a potential limiting factor on use of reconciliation).

  289. [289]. Oei & Osofsky, supra note 26, at 1322–23 (citing concerns of Congress members that technical corrections legislation would get hijacked to serve other purposes).

  290. [290]. Cf. Letter from Businesses, supra note 82 (trying to sell Congress on passing a qualified improvement property fix by explaining that there would be no revenue cost, since the desired fix was scored in the original legislation). The reconciliation legislation must follow the reconciliation instructions, in terms of whether the legislation will increase or decrease revenue and what limitations apply to any such increases or decreases. See 2 U.S.C. § 641(c)(1)(A)–(B) (2012). However, only the legislation, as a whole, needs to increase or decrease revenue in accordance with reconciliation instructions. This leaves the possibility that a particular fix could decrease revenue even if the reconciliation instructions required an increase. See Lynch & Saturno, supra note 288, at 3.

  291. [291]. See supra text accompanying notes 66–68.

  292. [292]. See supra text accompanying notes 66–68.

  293. [293]. Gerson, supra note 81, at 931.

  294. [294]. Bernstein, supra note 84 (describing technical corrections as previously a “routine” part of the process).

  295. [295]. See, e.g., Oei & Osofsky, supra note 26, at 1322–23 (citing this concern); Stamper, supra note 81, at 716 (citing concern of Senate Finance Committee Chair Chuck Grassley that technical corrections can become a “magnet for every change that people want to make in the tax code”).

  296. [296]. See supra Section III.C.3.

  297. [297]. See supra text accompanying notes 249–50.

  298. [298]. Freeman & Spence, supra note 25, at 17–62 (examining how agencies have to address outdated statutes in the face of congressional inaction); Zachary S. Price, Enforcement Discretion and Executive Duty, 67 Vand. L. Rev. 671, 676 (2014) (“Given the breadth of modern statutory prohibitions and the limitations on available resources for enforcement, federal officials must necessarily leave many statutory violations unpunished.”).

  299. [299]. See, e.g., Adam B. Cox & Cristina M. Rodríguez, The President and Immigration Law Redux, 125 Yale L.J. 104, 174–207 (2015) (praising the Obama administration’s institutionalization of immigration enforcement discretion on rule of law and other grounds); Metzger, supra note 247, at 1928–29 (addressing President Obama’s nonenforcement policies); Gillian E. Metzger & Kevin M. Stack, Internal Administrative Law, 115 Mich. L. Rev. 1239, 1258 (2017) (identifying “authorization, notice, justification, coherence, and procedural fairness” as rule-of-law values promoted by internal administrative measures (footnote omitted)); Osofsky, supra note 127, at 87–112 (examining the impact of categorical nonenforcement on agency legitimacy).

  300. [300]. Beermann, supra note 123, at 122–30 (exploring Congress’ involvement in the administration of the law); Elena Kagan, Presidential Administration, 114 Harv. L. Rev. 2245, 2380 (2001) (setting forth a model of presidential control of agencies); Glen Staszewski, Political Reasons, Deliberative Democracy, and Administrative Law, 97 Iowa L. Rev. 849, 861–66 (2012) (arguing that there are limits on political control and exploring a deliberative approach).

  301. [301]. See, e.g., John M. de Figueiredo & Elizabeth Garrett, Paying for Politics, 78 S. Cal. L. Rev. 591, 609–11 (2005) (exploring what money can and cannot buy in the political process).

  302. [302]. Cf. Melnick, supra note 45, at 784 (discussing executive action as a prod to get Congress to act).

  303. [303]. Cf. Lawrence Zelenak, The Uses and Abuses of Simplexity, 66 Emory L.J. Online 2011, 2020–22 (2017) (describing how the IRS dealt with a legislative drafting mistake by issuing what it believed to be the correct law in a nonprecedential publication, so as to get the correct result without having to make the statement in a more formal source of law).

  304. [304]. Cf., e.g., Final Bulletin for Agency Good Guidance Practices, 72 Fed. Reg. 3432, 3432
    –40 (Jan. 25, 2007) (developing policies and procedures for use of significant guidance by agencies).

  305. [305]. This is akin, in some ways, to statutes that explicitly provide agencies the authority to forbear applying certain statutory provisions. See, e.g., Zelenak, supra note 127, at 851–52 (contemplating a broad legislative grant to narrow the application of certain provisions when necessary for administrative reasons). See generally David J. Barron & Todd D. Rakoff, In Defense of Big Waiver, 113 Colum. L. Rev. 265 (2013) (examining the phenomenon); Deacon, supra note 25 (examining the phenomenon).

  306. [306]. See, e.g., Deacon, supra note 25, at 1560–64 (setting forth potential constitutional problems with Congress granting agencies explicit statutory authority to override statutes).


Associate Dean for Faculty Development and William D. Spry III Family Distinguished Professor of Law, UNC School of Law.

Thank you to UNC School of Law for providing funding for this project through the Extraordinary Scholarly Contributions Awards Program. Many thanks for helpful comments received at presentations at Duke University School of Law, University of Minnesota Law School, the 2019 Association of American Law Schools Annual Meeting in New Orleans, and the National Tax Association Annual Meeting in Tampa. Special thanks to Alice Abreu, Jake Brooks, John Coyle, Kathleen DeLaney Thomas, Brandon Garrett, Andy Grewal, Carissa Hessick, Kristin Hickman, Stephanie Hoffer, Eisha Jain, Rebecca Kysar, Bill Marshall, Ajay Mehrotra, Shuyi Oei, Nina Olson, Zach Osofsky, Guy-Uriel Charles, Chris Walker, Clint Wallace, and Larry Zelenak for comments about this Article. Finally, thank you to Grant Figari, Mara Howard-Williams, and Melissa Hyland for excellent research assistance.