The Administrative Procedures Act prescribes procedural requirements that govern the rulemaking activities of administrative agencies, including the Internal Revenue Service. It imposes, inter alia, notice and comment requirements which an agency may only bypass for good cause. While the Treasury Department’s assertion of the good cause exception when promulgating regulations to prevent tax abuse is a plausible application of the exception, it is distinguishable from typical assertions of good cause in one critical respect. Unlike many other agencies claiming good cause to bypass the notice and comment procedure, the Internal Revenue Service’s organic statute provides a viable alternative by allowing for the retroactive application of regulations issued to prevent abuse. As a result, two functionally equivalent approaches to combat tax abuse emerge, differentiated only by the presence or absence of public participation in rulemaking processes. By comparing the public policy ramifications of these different rulemaking approaches, I contribute to key policy debates centering on administrative rulemaking procedures. I find that issuing regulations under the good cause exception provides efficiency gains, but does so at the expense of increased litigation risk and a failure to mitigate the principal-agent problem animating the rulemaking requirements in the Administrative Procedures Act. On the other hand, regulations issued retroactively reduce agency costs, provide modest efficiency gains and partially outsource the burden of compliance to interested constituents who participate in rulemaking processes. On these grounds, I conclude that Treasury’s assertion of good cause to promulgate regulations to prevent abuse is unconvincing given the viable alternative of retroactive application of finalized regulations. These conclusions are specific to the trade-offs inherent when Treasury promulgates regulations to prevent tax abuse, but also speak to IRS compliance with the Administrative Procedures Act more generally.
- I. Introduction
- II. Survey of the Relevant Law
- III. Policy Analysis
- IV. Limitations
- V. Summary and Conclusion
The Administrative Procedures Act (“APA”) prescribes procedural requirements that govern the rulemaking activities of administrative agencies, including the Internal Revenue Service (“IRS”). It imposes, inter alia, notice and comment requirements which the IRS, like any other agency, may only bypass for good cause. Although scholars debate the scope of the good cause exception, the literature suggests several generally accepted applications of the exception. While the Treasury Department’s assertion of the good cause exception when promulgating regulations to prevent tax abuse is a plausible application of the good cause exception, it differs from these applications in one critical respect. Unlike many other agencies claiming good cause to bypass notice and comment procedures, the IRS’s organic statute provides a viable alternative by allowing for the retroactive application of regulations issued to prevent abuse. Because regulations issued to combat tax abuse enjoy retroactivity under the Internal Revenue Code (“IRC”), they would capture the same set of transactions whether or not they undergo notice and comment. As a result, two functionally equivalent approaches to combat tax abuse emerge, differentiated only by the presence or absence of public participation in the rulemaking process.
In this note, I examine the effect on administrative rulemaking procedures by comparing the public policy ramifications of Treasury either asserting the good cause exception or applying regulations retroactively when issuing regulations to prevent tax abuse. As part of this analysis, I examine the trend in jurisprudence and literature that tax exceptionalism is dead and that the IRS operates under administrative law as do other agencies. By scrutinizing the rulemaking activity of the IRS through the lens of administrative law norms, I supplement the scholarly debate over IRS compliance with APA rulemaking procedures with a policy analysis. More generally, I provide a case study of the difficulty found in balancing the tension between an agency’s justifiable request for expediency in rulemaking and the APA’s call for public participation in the rulemaking process. Comparing the two rulemaking approaches also provides fertile ground for exploring other issues, such as the political accountability of agencies which, if left unfettered in their rulemaking, might transform into a cancerous fourth branch of government.
The note proceeds as follows. To set the stage, Part II surveys the provisions of the APA and the IRC which govern the assertion of good cause and the retroactive application of regulations issued to prevent abuse. Then, Part III engages in a policy analysis of the two rulemaking approaches. I touch first on pragmatic considerations, including litigation risk and the costs associated with APA compliance. Next, I address efficiency arguments in favor of each rulemaking approach. This discussion explores the extent to which regulations issued for good cause and retroactive regulations reduce abusive tax sheltering transactions which misallocate taxpayer resources and cause tax revenue losses. Lastly, I look to the democratic purposes of the APA. This analysis adopts a principal-agent framework and assesses the effectiveness of either approach in mitigating agency costs. Part IV qualifies the foregoing analysis with a discussion of the limitations inherent in certain assumptions made by this note.
The APA contains a general presumption against the retroactive application of agency rulemaking. Section 553(d) of the APA sets out a general rule requiring a delay of at least thirty days before a rule may become effective. In Bowen v. Georgetown University Hospital, the Supreme Court reiterated that “[r]etroactivity is not favored in the law,” but provided a means to overcome this presumption. The Bowen Court held that “a statutory grant of legislative rulemaking authority will not, as a general matter, be understood to encompass the power to promulgate retroactive rules unless that power is conveyed by Congress in express terms.” Thus, the general presumption against retroactive application embedded in the APA may be overcome if Congress provides so expressly.
Section 7805(b) of the IRC provides explicit Congressional authorization to apply regulations retroactively to taxable periods after the earliest of three enumerated events. Under § 7805(b), regulations may apply retroactively to the date when the regulation is filed in the Federal Register, when the proposed or temporary form of a regulation later finalized was first issued, or when a notice with a substantive description of a rule later consummated as a regulation first appeared. These three events constitute bounded retroactivity; while they do not reach infinitely into the past, they do apply “new rules to transactions that have already been consummated.”
In § 7805(b)(3), however, Congress expressly provided for unbounded retroactive application of regulations issued “to prevent abuse.” Importantly, this provision allows for retroactive application to occur beyond the statutory constraints of § 7805(b)(1). Thus, Treasury may reach tax abuse in years that predate any rulemaking activities whatsoever, provided the statute of limitations on assessment and collection has not run. In this light, § 7805(b)(3) represents express Congressional authorization within the language of Bowen for the IRS to depart sharply from the APA’s general presumption against retroactivity. For purposes of this note, “retroactivity” will always refer to the unfettered retroactive application of regulations to prevent abuse under § 7805(b)(3).
Treasury regulations are subject to the procedural requirements of the APA. The IRS is within the purview of the APA because it is an “agency” within the language of § 551. Once subject to the procedural requirements of the APA as an “agency,” the IRS is not excluded by virtue of any of the exceptions enumerated in § 551(1). The statutory language is unambiguous on this score: the IRS is an agency whose action is governed by the APA. The IRS recognizes that the APA governs its rulemakings activities and provides guidance for compliance in Section 184.108.40.206.7.5.1 of the Internal Revenue Manual.
The APA places several procedural requirements on IRS rulemaking, including a notice and comment period. Section 553(b) requires that agencies engaging in rulemaking publish a notice of proposed rulemaking in the Federal Register. In addition to a notice of proposed rulemaking, agencies must allow public participation by “giv[ing] interested persons an opportunity to participate in the rule making through submission of written data, views, or arguments.” After due consideration of the evidence and arguments submitted by the public, agencies “shall incorporate in the rules adopted a concise general statement of their basis and purpose.” This notice and comment process, a kind of informal rulemaking, forms the core of the APA procedural requirements imposed on agency rulemaking activity.
When exigent circumstances arise, agencies may forgo these notice and comment procedural requirements by asserting good cause under § 553(b) of the APA. The language of the good cause exception is unhelpfully barren; it provides only that the notice and comment requirement does not apply to agency rulemaking “when the agency for good cause finds (and incorporates the finding and a brief statement of reasons therefor in the rules issued) that notice and public procedure thereon are impracticable, unnecessary, or contrary to the public interest.” Congress defined neither good cause nor any of the subsequent conditions necessary for such a finding (i.e., impracticality, lack of necessity, or contrary to public interest) in the statute. However, courts and scholars have drawn on legislative history in a cautious attempt to fill this lacuna with narrow definitions. Under these definitions, the IRS could not rely on impracticability or lack of necessity to promulgate regulations in final form without notice and comment. Ultimately, the public interest prong of the exception provides the only hook on which the IRS may hang an argument for promulgating regulations to combat abuse without notice and comment.
Treasury may argue that subjecting such regulations to notice and comment procedures is contrary to the public interest because this method of rulemaking does less to deter sham transactions and corresponding tax revenue losses. The Attorney General’s Manual on the APA illustrates the application of the public interest prong of the exception using the example of an agency considering the issuance of price controls. In this example, advance notice of the price controls would undermine the objectives of the agency by creating incentives for regulated firms to alter the timing of certain market activities. Because advance notice of price controls causes market distortion, it defeats the public interest and may be promulgated without notice and comment. In a similar fashion, advance notice of proposed regulations targeting tax abuse may defeat the public interest. As discussed more fully in Section III.B below, prosecuting a proposed regulation to its final form can take many years. Although the two rulemaking approaches are functionally equivalent in theory, the statutes of limitations terminate their equivalence after three or six years by tying the IRS’s hands. This introduces a meaningful possibility that the first wave of taxpayers employing an abusive structure will evade penalties. Like the announcement of price controls, notice of proposed rulemaking targeting an abusive transaction may create perverse incentives for taxpayers to expeditiously consummate the targeted transaction and to devise new abusive structures. These structures waste taxpayer resources on transactions without economic substance and lead to tax revenue losses, which may be contrary to the public interest and good cause for bypassing notice and comment procedures.
The policy analysis below centers on two extreme rulemaking alternatives available to Treasury when it issues regulations to combat tax abuse. While intermediate solutions do exist and some are addressed later in this note, focusing primarily on these extreme rulemaking alternatives provides a useful analytical framework. The following example of a regulation promulgated in response to tax abuse clarifies the nature of these extreme alternatives and sets the stage for the policy analysis to come.
Example: Taxpayers develop and begin to employ a new tax sheltering transaction on January 1, 2000. On January 2, 2000, Treasury learns of this transaction and deems it abusive. In seeking to combat this tax abuse, Treasury has two alternatives.
Alternative 1: Treasury issues final regulations on January 2, 2000, effective as of January 2, 2000. To bypass the notice and comment requirements of the APA, Treasury includes an explicit assertion of good cause in the regulation. In addition, Treasury explains in the regulation that compliance with notice and comment procedures is contrary to the public interest.
Alternative 2: Treasury does nothing until July 1, 2000. On July 1, 2000, Treasury issues proposed regulations designed to combat the abusive transaction and publishes a notice of proposed rulemaking in the Federal Register. From July 1 until December 30, 2000, Treasury allows interested taxpayers to participate in the rulemaking process through the submission of arguments and data. During this period, Treasury gives serious consideration to these viewpoints and modifies the proposed regulation to reflect the concerns and suggestions of the public. On December 31, 2000, Treasury issues final regulations combating the abusive transaction. Under § 7805(b)(3), Treasury applies the regulation retroactively to the period beginning on January 2, 2000.
The example above demonstrates the equivalence of the alternative rulemaking processes in terms of their application. Whether Treasury issues a regulation targeting abuse immediately in final form under the good cause exception or applies it retroactively after notice and comment, the transactions to which that regulation applies remain the same. In the example above, the regulation issued immediately in final form and the regulation applied retroactively both applied to the taxable period beginning on January 2, 2000. Thus, the distinctions between the rulemaking procedures hold no bearing whatsoever on the applicability of the regulation.
Although retroactivity renders rulemaking procedures irrelevant for purposes of determining the application of a regulation, these procedures remain relevant with regard to public policy concerns. The discussion below fleshes out public policy distinctions between the two rulemaking approaches and comments more generally on the balance to be struck between expediency and public participation in rulemaking processes. In addition, this policy analysis supplements existing scholarly literature on IRS compliance with APA procedural requirements. The analysis below leads me to conclude that Treasury may secure modest efficiency gains, avoid substantial litigation risks, and best vindicate the democratic purposes underlying the APA by issuing retroactive regulations to combat tax abuse.
If issued without notice and comment, Treasury regulations targeting abuse may increase the costs of rulemaking activity in two main ways. First, asserting good cause to bypass notice and comment will likely lead to increased litigation. This result stems largely from the character of the litigants and the decline of tax exceptionalism in matters of administrative law. For their part, tax abusers are particularly litigious and ready to bring legal challenge to agency action on administrative procedural grounds. Courts, inspired by recent jurisprudential developments toward the integration of IRS rulemaking into the general body of administrative law, may be ready to hear administrative procedural challenges to IRS rulemaking. Second, properly asserting good cause imposes the cost of compliance with the requirements of § 553(b). As discussed in greater detail below, the costs of § 553(b) compliance are minimal but the litigation risk of asserting good cause is substantial.
Several characteristics differentiate tax-abusing litigants from other taxpayers who bring suit and allow them to effectively raise APA procedural challenges to IRS rulemaking. Recent years have seen a surge in the market for tax shelters, costing “Treasury tens of billions of dollars in lost tax revenue a year.” As tax shelters increase in prevalence, so do the volume and intensity of tax litigation. The players involved in tax shelters are sophisticated. Scholarship has documented the “prominent role” that “elite professionals played . . . in the emergence of this market.” Large institutional powerhouses such as “accounting firms, investment banks, and corporate law firms all became involved, designing and marketing hundreds of highly lucrative shelters.” The stakes are large for those participating in these schemes. This is particularly true in light of accuracy-related penalties ranging from twenty to forty percent of understated income tax. Players in this market are better organized and more concentrated than ordinary, individual taxpayers. As already noted above, sophisticated institutions play an active role in structuring and executing tax-sheltering transactions. These parties do not suffer the collective action problems that would undermine litigation efforts by diffuse, individual taxpayers to challenge regulations on APA procedural grounds. Lastly, players in the market for tax shelters have greater access to legal counsel than do typical taxpayers. While the tax bar may have pushed for “law reforms intended to rein in the tax shelter market” and put forward “proposals intended to strengthen practice standards,” individual tax attorneys, on the other hand, “were deeply involved in abusive tax shelters.” In sum, deep-pocketed tax abusers and their sophisticated legal counsel have the ability and motivation to bring legal challenge against IRS rulemaking on APA procedural grounds.
At the same time, courts have been more willing to entertain APA procedural challenges to IRS rulemaking activity in a recent trend away from tax exceptionalism. As an example, the line of cases on the Son of BOSS tax shelter structure demonstrates the willingness of taxpayers to raise and the readiness of courts to recognize challenges to Treasury regulations on APA procedural grounds. The Bond Option Sales Strategy (commonly known as the “Son of BOSS” shelter) gained popularity in the 1990s, allowing “wealthy investors [to] shield billions of dollars from taxes.” The complexity of the transactions involved naturally restricted participation to highly sophisticated players with access to legal counsel. When the IRS took action to restrict use of the Son of BOSS transaction, it became embroiled in a brutal legal battle that spanned over a decade. Intermountain v. C.I.R. contained one of the most interesting twists in this saga, when the IRS moved to vacate a prior adverse decision in light of recently promulgated Treasury Decision 9466. Although the majority of the Tax Court rejected the motion on substantive grounds, Judges Halpern and Holmes concurred in result, reasoning instead that the “regulations [were] procedurally invalid under the Administrative Procedure Act.” The concurrence rooted the procedural infirmity of Treasury Decision 9466 in a failure to undergo notice and comment without claiming a legitimate exception under § 553. Although the good cause exception was not directly at issue in Intermountain and Judges Halpern and Holmes spoke only in concurrence, the case demonstrates a readiness on the part of courts to entertain APA procedural challenges to Treasury regulations.
If Treasury promulgates regulations to prevent abuse by asserting good cause, it risks protracted legal conflicts of a nature similar to the Son of BOSS cases. The costs of litigation to the IRS are substantial. In 2010 alone, the IRS Chief Counsel reported receiving 32,596 litigation cases with 31,016 still pending as of September 30, 2010. Given this hostile and litigious environment, the IRS ought to give serious thought to potentially applying regulations promulgated to prevent abuse retroactively. Although tax abusers may still challenge such retroactive regulation, they are far less likely to do so on procedural grounds and undercutting this argument would mitigate legal costs associated with defending the regulation.
Promulgating final regulations pursuant to the good cause exception also imposes compliance costs, but these costs are minimal. In relying on the exception, the IRS must remedy certain issues of form manifested in its past assertions of good cause. In particular, regulations asserting good cause must provide an explicit invocation of the exception and incorporate a succinct rationale in the rule issued. Language akin to the boilerplate statements of the Internal Revenue Manual suffices to expressly assert the exception and satisfy the previous deficiency.  As for the latter issue, a recent empirical study catalogued failures on the part of the IRS to incorporate into the rules that it issued a brief explanation for reliance on the good cause exception. Despite the prevalence of this deficiency, the IRS may correct its past non-compliance without any herculean effort. Rather, moderate inquiry into and reporting of the reasons justifying the assertion of good cause would suffice to cure this procedural infirmity. Indeed, the text of § 553(b)(3)(B) suggests brevity of rationale, a reasonable burden for the IRS to carry. Thus, the compliance costs associated with asserting good cause do not weigh heavily against immediate promulgation of final regulations to prevent abuse.
If, on the other hand, Treasury retroactively applies regulations to prevent abuse, it will incur compliance costs. It can, however, mitigate these costs by relying on public participation in the rulemaking process. Under § 553(b), Treasury must solicit and consider public commentary on its proposed regulations. Given the volume of IRS rulemaking, compliance with § 553(b) presents a meaningful strain on the agency’s limited resources. However, public commentary on proposed rules partially outsources information gathering and analysis to interested parties, which in turn reduces the cost of compliance to the IRS. As discussed further below, public participation in rulemaking activities mitigates compliance costs enough to make retroactive regulations the preferred rulemaking approach from a pragmatic point of view.
Compliance with APA notice and comment procedural requirements imposes non-trivial costs on the IRS. Pursuant to § 553(b), Treasury must publish notice of the proposed rulemaking in the Federal Register, solicit and consider public commentary on the proposal, and then incorporate into the finalized rule an explanation for its basis and purpose in light of the comments received. It is reasonable to conclude that the characteristics that make tax abusers particularly litigious also strongly position them to push back against proposed regulations that target abusive transactions. Given the sophistication of these clients and their legal counsel, they are likely to produce compelling data and arguments against the rules in their proposed form. Treasury, for its part, must entertain all comments and account for them in the basis and purpose statements in final regulations. The resulting morass of public comments could strain resources and act as a drag on IRS rulemaking activity.
Given the volume of IRS rulemaking, these procedural requirements could be onerous. According to a recent empirical study of IRS rulemaking activity, Treasury issued 203 Treasury Decisions and 163 Notices of Proposed Rule Making between January 1, 2003, and December 31, 2005. Information gathered from Treasury officials and tax attorneys verify that these data accurately illustrate the high volume of IRS rulemaking activities. Given the time required to prosecute a major rule from inception to finalized form, this volume presents a meaningful commitment of agency resources.
However, public participation in IRS rulemaking mitigates these compliance costs by allowing interested parties to shoulder the burden of information gathering and analysis. As a general matter, administrative agencies engage in substantial information gathering when promulgating rules in compliance with APA procedural requirements. The limitations of the IRS to inform itself are clear enough, seeing as how the drafters of Treasury regulations can never boast full information. Furthermore, the costs of information gathering are non-trivial in light of the impressive volume of IRS regulatory activity. For the IRS, as for other administrative agencies, public participation in rulemaking effectively outsources the burden of information gathering. Interested parties act as “boots on the ground” by reporting the application of the law to their idiosyncratic circumstances. Additionally, important constituents and their tax attorneys analyze the proposed rule, testing the language of the statute by subjecting it to alternative fact patterns. By performing these analyses, interested parties smoke out the “unexpected and unique applications” of the proposed regulation. In sum, public participation in rulemaking mitigates APA compliance costs by outsourcing information-gathering efforts and legal field testing of the regulation to important constituents, who, in turn, partially bear the expense of these efforts.
On balance, pragmatic considerations weigh in favor of subjecting regulations to notice and comment processes and applying them retroactively to prevent tax abuse. Although issuing regulations for good cause streamlines the rulemaking process, saving limited agency resources, it substantially backloads the cost of rulemaking by increasing litigation risk. Were Treasury to rely heavily on the good cause exception to promulgate regulations targeting abuse, the particularly litigious parties affected by such legislation could ride the wave of recent judicial activism in the arena of IRS compliance with the APA and bring APA procedural challenges. The Chief Counsel’s office already bears an enormous litigation burden without factoring in the sort of grueling legal disputes seen in the Son of BOSS line of cases. It is true that compliance with APA notice and comment requirements is costly for administrative agencies. This is particularly true for Treasury, which issues a substantial volume of regulations. However, public participation in the rulemaking process effectively outsources several components of the rulemaking process, including information collection and legal analysis. All else equal, pragmatic considerations counsel Treasury to combat tax abuse through the retroactive application of final regulations that have undergone notice and comment in compliance with § 553.
Abusive tax shelters are inefficient because they encourage unproductive transactions which misallocate taxpayer resources and reduce tax revenue. Defined generally, tax-sheltering transactions generate artificial losses which shield unrelated income from taxation under the code. Abusive tax shelters, the underlying transactions of which lack economic substance, are inefficient because the transactions serve no purpose other than to trigger certain preferences in the tax code. The transactions only add value to the taxpayer in the form of sheltered income, which, in the absence of the tax code, would represent a complete waste of resources. The lost productive capacity of the resources expended to execute the sham transactions constitutes deadweight loss. In addition, the sheltered income constitutes lost taxable income. Thus, abusive tax shelters waste taxpayer resources on meaningless transactions and reduce tax revenues.
Promulgating final regulations under the good cause exception strongly deters the execution of abusive tax shelters, reducing resources wasted on phony transactions and lowering tax revenue losses. Because the IRS cannot audit every taxpayer to identify who utilized abusive tax-sheltering structures, the deterrent force of regulations issued for good cause is a function of the probability that the IRS will assess a deficiency and the magnitude of the potential penalty to be imposed. A deficiency will likely be assessed for audited taxpayers who engaged in the transactions prohibited by regulations issued for good cause. If the taxpayer appeals the assessment, both the IRS and tax courts will likely take a very unsympathetic and aggressive stance against abusive tax-sheltering transactions. Most importantly, the statute of limitations does little to limit the reach of regulations issued for good cause. As for the magnitude of the penalty, the underpayment penalty of an assessed deficiency is substantial. In cases of underpayment, whether from substantial understatement, valuation misstatement or otherwise, a twenty percent penalty is imposed. In cases of gross valuation misstatements, which are likely the case with abusive tax-sheltering transactions, the penalty is increased to forty percent of the underpayment. Given the likelihood of a deficiency assessment and the substantial penalties associated with it, promulgating regulations for good cause strongly discourages abusive tax-sheltering activities. In particular, it discourages the inception of new transactions, the consummation of pending transactions, and encourages those who already executed such transactions to settle with the IRS. As a result, promulgating regulations to prevent abuse pursuant to the good cause exception reduces the amount of taxpayer resources wasted on phony transactions while recovering lost revenue, plus fines, for the fisc.
Issuing retroactive regulations provides a weaker deterrent to the execution of abusive tax shelters than do regulations issued for good cause, and they thus generate fewer reductions in resources wasted on sham transactions and tax revenue losses. The primary difference between the two rulemaking approaches, which weakens the deterrent effect of retroactive regulations, rests in the probability that the IRS will assess a deficiency. Like regulations issued for good cause, both the IRS and tax courts will likely take an unsympathetic and aggressive stance against abusive tax sheltering transactions prohibited by retroactive regulations. Unlike regulations issued for good cause, however, the ability of retroactive regulations to reach prohibited conduct is substantially limited by the statute of limitations.
Given that promulgation of a major final regulation can take many years, the statute of limitations gives quarter to those who engage in abusive transactions first. This leads to several inefficient outcomes. First, applying regulations to prevent abuse retroactively actually incentivizes the expeditious consummation of pending transactions and the inception of new transactions. Second, because the statute of limitations restricts the retroactive application of regulations, it places a premium on new tax sheltering structures. As Treasury takes years to retool, it is constantly playing catch-up with taxpayers. This creates a “get in first, get off scot-free” scenario where the statute of limitations shields taxpayers who dedicate resources to devising new structures. While applying regulations retroactively does reduce the amount of taxpayer resources wasted on phony transactions to some extent, it nonetheless creates perverse incentives to accelerate existing transactions and devise new ones. The costs incurred by taxpayers to consummate these transactions constitute deadweight loss. Thus, retroactive regulations do less to restrain the inefficiencies inherent in abusive tax shelters than do regulations issued immediately in final form under the good cause exception.
All things equal, efficiency arguments weigh in favor of immediate promulgation of regulations targeting abusive transactions under the good cause exception. As discussed above, abusive tax shelters are inefficient because they encourage unproductive transactions that misallocate taxpayer resources and reduce tax revenues. Issuing regulations for good cause strongly deters the execution of abusive tax shelters, reducing resources wasted on sham transactions as well as lowering tax revenue losses. If Treasury assesses a deficiency pursuant to these regulations, there is little doubt that the taxpayer will suffer harsh underpayment penalties. By contrast, retroactive application of final regulations to prevent tax abuse introduces substantial uncertainty as to whether or not a tax abuser will suffer underpayment penalties at all. To a large extent, this stems from the interplay between the statutes of limitations and the retroactive effect of final regulations, which effectively shields the first tax abusers to engage in tax-sheltering transactions. Consequently, efficiency considerations weigh in favor of promulgating regulations to prevent abuse under the good cause exception.
Broadly speaking, the APA procedures respond to the primary issue of any representative democracy—that is, facilitating policy formation which reflects and directly responds to the needs and interests of the democracy’s citizens. Few of these administrative procedures find actual root in the Constitution because “[t]he U.S. Constitution deals with the electoral side of this problem by constructing institutional safeguards and incentive structures designed to make elected representatives responsive to citizens.” Rather, Congress created certain APA procedures, above and beyond those animated by the Constitution, to ameliorate the principle-agent problem arising from the delegation of legislative authority to unelected bureaucrats. Put differently, “administrative procedures are another mechanism for inducing compliance” while avoiding the costs of directly monitoring and conditioning agent behavior through reward and punishment.
Issuing regulations retroactively after completion of APA rulemaking procedures mitigates the principle-agent problem by shifting the costs of monitoring IRS policy decisions to interested constituents. Section 553 achieves this goal through two primary means. First, the APA forces information flow by requiring the publication of certain information in the Federal Register. This breaks down the asymmetries of information that would otherwise pose a disadvantage to politicians when they interact with the bureaucrats to whom they delegated authority. Publication of certain information in the Federal Register also publicly disseminates information, which facilitates monitoring by taxpayers. Second, the APA rulemaking procedures “enfranchise important constituents in agency decision-making processes, thereby assuring that agencies are responsive to their interests.” The APA enfranchises interested constituents by opening a direct channel of communication between them and administrative rule makers and providing for transparency. When issuing a final rule pursuant to § 553, IRS rulemakers must summarize comments received in regards to the proposed rule. In addition, they must justify the form of the final rule in light of these comments. The IRS’s transparent consideration of public comments enables interested constituents to watch for and identify instances in which the Agency failed to reflect important constituent interests. In sum, public dissemination of the proposed regulation as well as a transparent and open opportunity for public comment on the proposal allow those impacted by the rules to monitor IRS decision-making.
Although taxpayers may not “vote out of office” bureaucrats who fail to properly consider their comments, they may nonetheless indirectly condition IRS behavior through appeals to representatives in the executive and legislative branches. Taxpayers can exert electoral pressure on the political officials who delegated authority to the IRS in the first place. These politicians, in turn, may exercise Congressional or executive oversight over the IRS to resolve policy disputes. This relationship imposes on bureaucrats some political accountability, despite the absence of electoral pressures. If the IRS deviates sharply from the interests of its constituents, indirect pressure from taxpayers may help to realign its policy objectives with the interests of taxpayers. Thus, taxpayers have the ability to mitigate the principal-agent problem when Treasury retroactively applies regulations to prevent abuse. This, in turn, vindicates the democratic ideals underpinning the APA by encouraging IRS policy decision-making to better reflect the interests of individual taxpayers.
Regulations issued for good cause fail to mitigate the principal-agent problem, and thus do little to vindicate the democratic ideals embedded in the APA. Regulations issued immediately in final form do not provide the public with the time or information necessary to monitor IRS policy decisions. Unlike retroactive regulations, regulations issued under the good cause exception provide no advance warning to the public (e.g., notice of proposed rulemaking). They also fail to provide for a period of evaluation. Thus, bypassing the notice and comment procedure precludes the flow of information relating to the regulation at hand. Consequently, interested constituents may only voice their concern, if at all, on the basis of their independently acquired data.
Regulations issued for good cause also undercut the ability of taxpayers to condition IRS rulemaking in several ways. First, time constraints imposed by a regulation issued for good cause reduces the ability of taxpayers to indirectly exert pressure on IRS rulemakers. Once Treasury issues final regulations, the taxpayer, if they are to indirectly condition IRS rulemaking, must simultaneously devote their resources towards compliance and towards efforts to lobby elected officials. This divides taxpayer resources and lessens the force that these taxpayers may exert on elected officials to condition IRS policy decision-makers. Second, the absence of public commentary breaks down potential collective lobbying efforts among similarly situated constituents. The public nature of the notice and comment process allows for taxpayers with similar needs and concerns to band together in lobbying campaigns. In sum, regulations promulgated for good cause substantially limit the ability of the public to monitor and condition IRS rulemaking. This failure to ameliorate the principal-agent problem represents a failure to achieve the democratic purposes that animate the APA.
Thus, Treasury best achieves these purposes by retroactively applying regulations targeting abuse after compliance with notice and comment processes. The APA establishes rulemaking procedures in an effort to provide political accountability for administrative agency policy decisions. These APA rulemaking procedures allow the public to monitor and indirectly condition IRS rulemaking, which mitigates the principal-agent problem. If Treasury issues regulations to prevent abuse under the good cause exception, taxpayers are not in a strong position to monitor or condition IRS policy decision-makers. On these grounds, I conclude that Treasury best vindicates the democratic ideals underpinning the APA by subjecting regulations targeting abuse to notice and comment procedures and applying them retroactively.
Some ambiguity does exist as to the definition of “abuse” within the language of § 7805(b), but that does not undercut this discussion. It is unnecessary to define “abuse” for purposes of this note. It is sufficient to say that abuse does exist in some form and that determining which conduct constitutes abuse is only a second-order inquiry. Ambiguity as to which particular transactions or conduct are covered by this policy discussion does not undermine the validity of its conclusions, for it speaks of tax abuse generally rather than specific transactions. The assumption that abuse does, in fact, exist finds support in the IRC. Why else would § 7805(b) exist, but that Congress recognized the existence of certain transactions which aptly fall under the umbrella term “abuse”?
The assumption that Treasury may either assert good cause or retroactively apply regulations to combat tax abuse overlooks the potential role of temporary regulations in this effort, but this is intentional. At first blush, temporary regulations appear to offer an attractive combination of the desirable attributes of regulations issued in final form and those modified by the incorporation of public comments. Under § 7805(e), all temporary regulations are simultaneously put forward as proposed regulations to solicit public comment. Thus, temporary regulations might provide the efficiency gains of regulations issued for good cause, yet also boast the benefits of engaging the public in rulemaking processes if the finalized form of the rule is applied retroactively. If any concerns exist regarding the use of temporary regulations, the sunset provision in § 7805(e) might seem to put them to rest. However, recent scholarship strongly suggests that temporary regulations are a loose cannon that neither provide the certainty for taxpayers necessary to deter inefficient transactions nor engages them in public rulemaking. Neither can the sunset provision redeem any pernicious qualities of the temporary regulation as its abuse has led to a class of “permanently temporary” regulations. Inclusion of temporary regulations in this policy discussion would unnecessarily muddy the waters of what is otherwise a unique opportunity to explore key policy tensions in administrative rulemaking under the APA. Thus, to facilitate coherent policy analysis, I excluded temporary regulations as a possible means of combating tax abuse.
Similarly, the assumption that Treasury may either assert good cause or retroactively apply regulations to combat tax abuse does not consider certain intermediate solutions but sacrifices little in doing so. For instance, Treasury may issue revenue rulings or notices to provide immediate guidance to taxpayers. Recent literature on Treasury rulemaking under the APA suggests that such intermediate solutions provide meaningful guidance to taxpayers and carry enough weight to encourage taxpayer compliance. However, two reasons justify exclusion of these intermediate solutions from the policy analysis. First, including revenue rulings and notices would weaken the analytical framework unique to the tax abuse context. As informal forms of rulemaking, these intermediate solutions are categorically exempted from the APA rulemaking procedures. By definition, informal rulemaking takes place without public participation whether or not such rulemaking addresses abusive transactions. Thus, including these intermediate solutions would distort the dichotomous rulemaking framework used in this note for analytical clarity. Second and more importantly, revenue rulings and notices do not likely influence tax abusers the same way that they influence other taxpayers. To a large extent, these intermediate solutions encourage compliance by mitigating underpayment penalties. Put somewhat differently, risk-averse taxpayers may simply comply with informal interpretations to avoid enforcement action. Tax abusers have already crossed the Rubicon; such informal and toothless forms of guidance will have little impact on their behavior. For these reasons, excluding intermediate solutions from the policy analysis does not undermine the conclusions that I reach below.
Although an efficiency analysis favors regulations issued for good cause, pragmatic considerations and examination of the democratic purposes of the APA lead me to conclude that retroactivity is the better alternative to prevent tax abuse. The efficiency arguments in favor of regulations issued for good cause are relatively straightforward. When promulgated for good cause, regulations targeting abusive transactions minimize the inefficiencies inherent in abusive tax shelters by deterring sham transactions that misallocate taxpayer resources and lower tax revenue. Regulations issued for good cause provide a greater deterrent than retroactive regulations because they provide a greater likelihood that underpayment penalties will be imposed on tax abusers with fewer restrictions imposed by the statute of limitations. However, certain pragmatic arguments outweigh the efficiency gains of regulations issued for good cause. For instance, issuing regulations for good cause increases litigation risk, which is especially acute in light of the particularly litigious parties affected by such legislation and recent judicial activism in the arena of IRS compliance with the APA. Retroactive regulations, on the other hand, will likely mitigate this litigation risk. The compliance costs associated with retroactive regulations do not weigh heavily against this rulemaking approach as public participation in information gathering and legal analysis efforts substantially shifts this burden to interested constituents. Retroactive regulations also better vindicate the democratic purposes of the APA. Issuing regulations retroactively to prevent tax abuse, unlike regulations issued for good cause, allows the public to monitor and indirectly condition IRS rulemaking. This, in turn, mitigates the principal-agent problem that animates the APA rulemaking procedures.
In sum, I find Treasury’s assertion of good cause to promulgate regulations to prevent abuse unconvincing, given the viable alternative of retroactive application of finalized regulations. The APA rulemaking procedures codified a delicately struck balance between expediency and public participation in agency rulemaking activities. The rulemaking procedures provide political accountability sufficient to mitigate principal-agent issues arising from unelected bureaucrats wielding delegated legislative power, while also providing an exception in cases of exigency. Whenever possible, agencies ought to allow for public participation in rulemaking. In the tax abuse context, retroactivity offers a viable alternative to regulations issued for good cause that allows for public participation while yielding superior outcomes in public policy terms. Other scholars have similarly concluded that Treasury’s assertions of good cause are tenable in light of retroactivity. In sum, Treasury’s assertion of good cause to combat tax abuse is plausible, but is weakened by the unique retroactive flexibility codified in § 7805.
Although the policy arguments above address the trade-offs inherent in the promulgation of regulations to prevent tax abuse, they also shed light more generally on rationales for compliance with the APA. Given the long and troubling history of IRS non-compliance with APA requirements, Treasury ought to reconsider the policies motivating and the benefits flowing from public participation in rulemaking processes. This is particularly true in light of a strengthening consensus among courts and commentators that the IRS does not deserve special treatment under law; its rules are to be promulgated and reviewed according to the norms of administrative law. Upon full integration into the body of administrative law, the IRS stands to gain the tangible and economically significant benefits of public participation in rulemaking processes.
 See, e.g., James Kim, For A Good Cause: Reforming the Good Cause Exception to Notice and Comment Rulemaking Under the Administrative Procedure Act, 18 Geo. Mason L. Rev. 1045 (2011); Juan J. Lavilla, The Good Cause Exemption to Notice and Comment Rulemaking Requirements Under the Administrative Procedure Act, 3 Admin. L.J. 317 (1989); Juan F. Vasquez, Jr. & Peter A. Lowy, Challenging Temporary Treasury Regulations: An Analysis of the Administrative Procedure Act, Legislative Reenactment Doctrine, Deference, and Invalidity, 3 Hous. Bus. & Tax L. J. 248 (2003) (analyzing the good cause exception’s role in promulgating temporary regulations); Michael Asimow, Public Participation in the Adoption of Temporary Tax Regulations, 44 Tax Law 343, 348 (1991) (same).
 Lavilla, supra note 1, at 352–94 (arguing that time and economic costs are sufficient justifications for asserting the exception); Kristin E. Hickman, Coloring Outside the Lines: Examining Treasury’s (Lack of) Compliance with Administrative Procedure Act Rulemaking Requirements, 82 Notre Dame L. Rev. 1727, 1782 (2007) (“[T]he good cause exception exists principally to give agencies flexibility in dealing with emergencies and typographical errors, plus the occasional situation in which advance notice would be counterproductive.”). See also Kim, supra note 1, at 1049–51.
 Although some ambiguity exists as to what conduct constitutes “abuse” within the language of I.R.C. § 7805(b), it is beyond the scope of this paper to define such conduct. Defining “abuse” is a second order inquiry that does not bear on the policy discussion that follows. See infra Part IV for a more complete discussion of this assumption and its limitations.
 Hickman, supra note 2, at 1785.
 Compare, e.g., 16 U.S.C. §§ 1613, 551 (2012) (no explicit grant of authority to Forest Service to retroactively apply regulations), Farm Security and Rural Investment Act of 2002, PL 107–171, 2002, 116 Stat 134 (no explicit grant of authority to Farm Service Agency to retroactively apply regulations), 21 U.S.C. §§ 463, 608, 621, 623, 624, 1043 (2012) (no explicit grant of authority to Food Safety and Inspection Service to retroactively apply regulations), 47 U.S.C. § 904 (2012), (no explicit grant of authority to National Telecommunications and Information Administration to retroactively apply regulations), 42 U.S.C. §§ 7172, 7173(a) (no explicit grant of authority to Federal Energy Regulatory Commission to retroactively apply regulations) (2012), 21 U.S.C. §§ 345, 363, 371(a) (2012) (no explicit grant of authority to U.S. Food and Drug Administration to retroactively apply regulations), with I.R.C. § 7805(b)(3) (2012) (explicit grant of authority to IRS to retroactively apply certain regulations).
 For examples of courts clearing the vestiges of tax exceptionalism, see Mayo Found. for Med. Educ. & Research v. United States, 131 S. Ct. 704, 713 (2011) (holding that “the principles underlying our decision in Chevron apply with full force in the tax context” and commenting that “we see no reason why our review of tax regulations should not be guided by agency expertise pursuant to Chevron to the same extent as our review of other regulations”); Cohen v. United States, 650 F.3d 717, 723 (D.C. Cir. 2011) (holding that “[t]he IRS is not special” and “no exception exists shielding it—unlike the rest of the Federal Government—from suit under the APA”). For examples of scholarship discussing the decline of tax exceptionalism, see Roger Dorsey, Mayo and the End of ‘Tax Exceptionalism’ in Judicial Deference, 87 PRACTXST 63, 63 (2011); Kristin Hickman, Goodbye National Muffler! Hello Administrative Law?, TaxProf Blog (January 11, 2011), http://taxprof.typepad.com/taxprof_blog/2011/01/hickman-.html (“[T]he Supreme Court has finally and decisively rejected the notion of tax exceptionalism in judicial review standards.”).
 Rimma Tsvasman, A Case for the IRS’s Full Compliance with the Administrative Procedure Act, 76 Brook. L. Rev. 837, 841–43 (2011).
 5 U.S.C. § 553(d) (2012).
 Bowen v. Georgetown Univ. Hosp., 488 U.S. 204, 208 (1988).
 In other words, “[t]he APA applies to all federal agencies and acts as a default rule, supplying procedures when organic statutes do not.” Lisa Schultz Bressman, Procedures as Politics in Administrative Law, 107 Colum. L. Rev. 1749, 1755–56 (2007). But see Toni Robinson, Retroactivity: The Case for Better Regulation of Federal Tax Regulators, 48 Ohio St. L.J. 773 (1987).
 I.R.C. § 7805(b)(1)(a) (2012).
 I.R.C. § 7805(b)(1)(b) (2012).
 I.R.C. § 7805(b)(1)(c) (2012).
 Michael J. Graetz, Retroactivity Revisited, 98 Harv. L. Rev. 1820, 1821 (1985). My nomenclature differs from that used by Graetz. In his article, Graetz states that “all changes in tax law—indeed, I think, all changes in economic laws—are inherently retroactive.” Id. at 1822. He defines retroactivity to include even “purportedly prospective changes in the law.” Id. Despite Graetz’s insistence that distinctions between “retroactive” and “prospective” changes in the law are “misleading” and “illusory,” I retain the traditional nomenclature in this article for clarity. Id. at 1822–23. Thus, I use the term “retroactive regulation” in reference to those regulations promulgated under § 7805(b) of the IRC.
 I.R.C. § 7805(b)(3) (2012).
 Generally, the statute of limitations on assessment and collection is three years. I.R.C. § 6501(a) (2012). However, in cases of substantial omissions, the statute of limitations is extended to six years. I.R.C. § 6501(e) (2012).
 The APA defines an agency as “each authority of the Government of the United States, whether or not it is within or subject to review by another agency.” 5 U.S.C. § 551(1) (2012).
 The list of excluded bodies is brief and explicit. 5 U.S.C. § 551(1)(A)–(H) (2012). There is no room for argument that the IRS is removed from the APA under Section 551(1).
 I.R.S., Internal Revenue Manual § 220.127.116.11.7.5.1 (Jan. 12, 2012), available at http://www.irs.gov/irm.
 For the definition of “rule,” “rulemaking,” and other key terminology used in Section 553, see 5 U.S.C. § 551 (2012).
 5 U.S.C. § 553(b) (2012).
 5 U.S.C. § 553(c) (2012).
 For a discussion of informal rulemaking and how it differs from formal rulemaking, see Richard J. Pierce, Jr., Administrative Law Treatise §§ 7.1, 7.2, at 411 (4th ed. 2002). See also Kim, supra note 1, at 1048–49.
 The good cause exception is not the only carve-out from the procedural requirements under § 553. Certain agency rulemaking activity, such as “interpretative rules, general statements of policy, or rules of agency organization, procedure, or practice” are excluded from APA rulemaking procedural requirements as well. 5 U.S.C. § 553(b)(3)(A) (2012). For further discussion of the exception for interpretive rules and IRS rulemaking, see Asimow, supra note 1, at 350–61.
 5 U.S.C. § 553(b)(3)(B) (2012).
 Some scholars have relied on legislative history to divine the meaning of these terms. See, e.g., Lavilla, supra note 1, at 333–35.
 For examples of courts narrowly construing the exception, see Levesque v. Block, 723 F.2d 175, 184 (1st Cir. 1983) (stating that the First Circuit narrowly construes the good cause exception), Tenn. Gas Pipeline Co. v. F.E.R.C., 969 F.2d 1141, 1144 (D.C. Cir. 1992) (“Despite the broad nature of this language, our cases make clear that the good cause exception is to be narrowly construed and only reluctantly countenanced.”) (internal citation omitted). For an example of scholarship relying on legislative history to reach the same conclusion, see Lavilla, supra note 1, at 342.
 Notice and comment procedures are “unnecessary” when correcting such things as typographical, “obvious grammatical[,] or cross-referencing errors.” Hickman, supra note 2, at 1790. Nobody argues that regulations targeting abuse should forego notice and comment for their insignificance. To the contrary, any argument for notice and comment rests on their significance. Notice and comment periods are impracticable when “an agency finds that due and timely execution of its functions would be impeded by the notice otherwise required.” U.S. Dep’t of Justice, Attorney General’s Manual on the Administrative Procedure Act 30 (1947). In the tax abuse context, there are no deadlines imposed by Congress nor threats to public safety that would support a “timeliness” claim under the impracticability prong. Hickman, supra note 2, at 1783; U.S. Dep’t of Justice, Attorney General’s Manual on the Administrative Procedure Act 30–31 (1947).
 U.S. Dep’t of Justice, Attorney General’s Manual on the Administrative Procedure Act 31 (1947).
 Hickman, supra note 2, at 1785.
 See infra note 69.
 See supra note 17.
 See infra Section III.B.
 See infra Part IV for a more detailed discussion of this approach and its limitations.
 See infra notes 40–58 and accompanying text.
 See infra notes 47–58 and accompanying text.
 Tanina Rostain, Sheltering Lawyers: The Organized Tax Bar and the Tax Shelter Industry, 23 Yale J. on Reg. 77, 77 (2006). See also Garrison Grawoig DeLee, Note, Tax Shelters: Will the Latest Tools Really Help?, 57 S. Cal. L. Rev. 431, 431 (1984) (estimating annual revenue losses attributable to abusive tax shelters to be $3.5 billion).
 For example, “[m]any taxpayers who did not accept the settlement offer [for engaging in Son of BOSS transactions] . . . challenged the IRS’s deficiency notices in court.” Derek B. Wagner, Note, Who’s the (Son of) BOSS?: The Struggle Between the Federal Circuit and Treasury to Define “Omits from Gross Income” in Son of BOSS Tax Shelters and Other Overstatement-of-Basis Tax Cases, 21 Fed. Circuit B.J. 45, 48 (2011).
 Rostain, supra note 40, at 78.
 For instance, “[b]etween 1996 and 2000 more than 60% of corporations avoided paying any taxes.” Id. at n. 23 (citing U.S. Gen. Accounting Office, Comparison of the Reported Tax Liabilities of Foreign and U.S. Controlled Corporations, 1996–2000 (2004)). Rostain further explains that, “[w]hile corporations were able to reduce some taxes by legally permissible means, a large part of the offset was likely the result of abusive tax shelters.”
 In cases of underpayment, whether from substantial understatement, valuation misstatement or otherwise, a twenty percent penalty is imposed on the underpayment. I.R.C. § 6662(a). In cases of gross valuation misstatements, the penalty is increased to forty percent of the underpayment. I.R.C. § 6662(h).
 Rostain, supra note 40, at 77. Rostain further points out that “[l]awyers who eschewed direct involvement in development and sales were still able to profit handsomely from the tax shelter boom by writing opinion letters for clients,” since “[o]pinion letters were a sine qua non of tax shelter purchases.” Id. at 92.
 For a thorough explanation of the procedural history of and commentary on these cases, see Wagner, supra note 41.
 Id. at 46–47.
 Lynnley Browning, U.S. Judge Backs I.R.S. Ruling Invalidating Tax Shelter, Possibly Aiding U.S. Criminal Case, N.Y. Times (Apr. 22, 2006), http://www.nytimes.com/2006/04/22/business/22shelter.html.
 Wagner, supra note 41, at 46–47 (commenting on the complexity of the transactions involved in the Son of BOSS shelter).
 Id. at 46–48.
 Bakersfield Energy Partners, LP v. Comm’r, 568 F.3d 767 (9th Cir. 2009).
 Definition of Omission From Gross Income, 74 Fed. Reg. 49321-01 (Sept. 28, 2009) (to be codified at 26 C.F.R. pt. 301). Treasury Decision 9466 promulgated temporary regulations §§ 301.6229(c)(2)-1T and 301.6501(e)-1T, which aimed to cure the procedural defects responsible for previous adverse judicial decisions. Wagner, supra note 41, at 50. For further discussion of the practice of employing retroactive interpretive Treasury regulations to counter adverse judicial determinations, see Andrew Pruitt, Judicial Deference to Retroactive Interpretative Treasury Regulations, 79 Geo. Wash. L. Rev. 1558 (2011).
 Intermountain Ins. Serv. of Vail, Ltd. Liab. Co. v. C.I.R., 134 T.C. 211, 220–24 (2010) rev’d sub nom. 650 F.3d 691 (D.C. Cir. 2011), as amended on denial of reh’g (Aug. 18, 2011).
 Id. at 238.
 Id. at 239–42.
 The Secretary conceded in his brief that he was not relying on the good cause exception. Intermountain, 134 T.C. at 211, n.7 (2010) rev’d sub nom. 650 F.3d 691 (D.C. Cir. 2011), as amended on denial of reh’g (Aug. 18, 2011).
 Another example of judicial willingness to enter the fray is seen in Cohen v. United States, 650 F.3d 717, 723 (D.C. Cir. 2011).
 Hickman, supra note 2 at 1799–800 ( “Lawsuits tie up resources that could be better used in other, more productive ways, and if taxpayers succeed in persuading courts to invalidate Treasury regulations for procedural reasons, Treasury and the IRS will be forced to expend even further resources re-promulgating existing regulations rather than addressing newer issues.”).
 Dep’t of the Treas., Internal Revenue Service Data Book 61 (2010), available at http:// www.irs.gov/pub/irs-soi/10databk.pdf.
 Hickman, supra note 2, at 1778 (remarking that the vast majority of Treasury projects reviewed as part of the study which relied on the good cause exception manifested “problems of form, substance, or both”).
 5 U.S.C. § 553(b)(3)(B) (2012).
 But see Hickman, supra note 2, at 1798.
 Id. at 1778.
 See infra notes 68–75 and accompanying text.
 5 U.S.C. §§ 553(b)–(c) (2012).
 Hickman, supra note 2, at 1740–41.
 Id. at 1741.
 According to Richard Pierce, “promulgation of a single major rule often requires five to ten years and tens of thousands of agency staff hours.” Richard J. Pierce, Jr. et al., Administrative Law and Process § 6.4.6b at 336 (4th ed. 2004). Recent literature has adopted this estimate in the tax context. Kristin E. Hickman, A Problem of Remedy: Responding to Treasury’s (Lack of) Compliance with Administrative Procedure Act Rulemaking Requirements, 76 Geo. Wash. L. Rev. 1153, 1202–03 (2008). To view details of costs incurred by the IRS in performing its regulatory function in the years 2009 and 2010, see Dep’t of the Treas., Internal Revenue Service Data Book 65 (2010), available at http://www.irs.gov/pub/irs-soi/11databk.pdf.
 See Pierce, et al., supra note 69, at 336.
 Id. at 366.
 See supra notes 67–69 and accompanying text.
 Richard Pierce comments, “[P]arties have incentives to include in their comments studies and affidavits of experts addressing such issues as (1) the frequency of occurrence of various factual patterns, (2) the likely efficacy of alternative rules in shaping conduct, (3) the cost of compliance with alternative rules, and (4) the practical problems inherent in implementing or enforcing alternative rules in varying factual contexts.” Pierce, supra note 25, § 6.8 at 369.
 Richard Pierce indicates that affected parties have a “natural incentive” to scrutinize proposed rules in this fashion. Id.
 Asimow, supra note 1, at 366; see also Pierce, supra note 25, § 6.8 at 369.
 See supra note 58 and accompanying text.
 Lawrence Zelenak, When Good Preferences Go Bad: A Critical Analysis of the Anti-Tax Shelter Provisions of the Tax Reform Act of 1986, 67 Tex. L. Rev. 499, 524–27 (1989).
 For a discussion of “legitimate” and “abusive” tax shelters, see id.
 For a full discussion of this theoretical framework for analyzing tax compliance, see Michael G. Allingham & Agnar Sandmo, Income Tax Evasion: A Theoretical Analysis, 1 J. Pub. Econ. 323 (1972). See also James Andreoni, Brian Erard & Jonathan Feinstein, Tax Compliance, 36 J. Econ. Literature 818 (1998).
 For an example of the aggression with which the IRS will wage war on abusive tax shelters, see Wagner, supra note 41. This aggression, combined with the likely unsympathetic position of the tax court, provides a counterforce to the natural urge of taxpayers to play the “tax audit lottery.” Michael J. Graetz, Tax Reform Unraveling, 21 J. Econ. Perspectives 69, 83 (2007).
 For a discussion of the relevant statutes of limitations, see supra note 18.
 5 U.S.C. § 6662(a) (2012).
 5 U.S.C. § 6662(h) (2012).
 The uncertainties arising from selective tax audits and tax court rulings remain the same as in the good cause context. The magnitude of the penalties imposed also remains the same. See supra note 75–76.
 See supra note 69.
 The conflict between the statute of limitations and the retroactive application of regulations is not theoretical, but has taken center stage in the recent Son of BOSS litigation. See Wagner, supra note 41.
 See Pierce, et al., supra note 69.
 Mathew D. McCubbins, Roger G. Noll & Barry R. Weingast, Administrative Procedures as Instruments of Political Control, 3 J. L. Econ. & Org. 243, 243 (1987).
 Bressman, supra note 12, at 1767–71. See also McCubbins et al., supra note 88, at 243–44; Asimow, supra note 1, at 366; Kim, supra note 1, at 1049 (“[N]otice and comment procedures give affected parties the ability to participate and influence agency decisionmaking at an early stage, which promotes the political accountability of the agency.”).
 McCubbins et al., supra note 88, at 243–44. Lisa Shultz Bressman sees the principle-agent framework as only a partial explanation of the role the APA plays in governing the relationship between administrative agencies and other branches of the government. Drawing on legal scholarship and positive political theory, she contends that administrative procedures reflect the political agenda of the Court. Bressman, supra note 12, at 1767.
 See, e.g., 5 U.S.C. § 553(b) (2012).
 McCubbins et al., supra note 88, at 243; Bressman, supra note 12, at 1765–71.
 McCubbins et al., supra note 88, at 244.
 5 U.S.C. § 553 (2012).
 Richard Pierce comments that the “notice of proposed rulemaking enables citizens who oppose or support the proposal to alert the President and members of Congress to the existence of the proposal and to express their views of the agency’s proposal to those politically accountable officials.” Pierce, supra note 26, § 6.8 at 369. These officials, in turn, “express to the agency their views concerning the proposed policy decision and, through the process of Executive and congressional oversight . . . affect agency resolutions of policy disputes.” Id.
 5 U.S.C. § 7805(e)(1) (2012).
 Finalized regulations may be applied retroactively to the date on which temporary regulations were filed with the Public Register. I.R.C. § 7805(b)(1)(B) (2012).
 This provision provides that “[a]ny temporary regulation shall expire within 3 years after the date of issuance of such regulation.” I.R.C. § 7805(e)(2) (2012).
 Hickman, supra note 2, at 1801–04.
 Asimow, supra note 1, at 366-67; Hickman, supra note 2, at 1801 (“[T]axpayers and their representatives may be discouraged from making comments by the comparative finality of temporary regulations as opposed to regulations that are merely proposed.”).
 Vasquez et al., supra note 1, at 254 (citing as an example Treasury Regulation § 145.4051-1, which had been in temporary form for over twenty years).
 Hickman, supra note 2, at 1804–05.
 Internal Revenue Serv., Notice 90-20, at 4 (1990). The relevant language of the notice reads: “[f]or purposes of section 6662(d)(2)(B)(i) of the Code, the Service also will treat as authority General Explanations of tax legislation prepared by the Joint Committee on Taxation (the ‘Blue Book‘), proposed regulations, information or press releases, notices, announcements, and any other similar documents published by the Service in the Internal Revenue Bulletin.” Id.
 Hickman, supra note 2, at 1805.
 Lavilla, supra note 1, at 341; Hickman, supra note 2, at 1785.