Printer Friendly

LEGISLATIVE ENTRENCHMENT AND FEDERAL FISCAL POLICY.

I

INTRODUCTION

Amid concerns about government debt, the 112th Congress passed the Budget Control Act of 2011 to reduce the federal government's budget deficits. (1) The law set out caps on discretionary spending for each fiscal year over the following decade and required sequestration of discretionary spending to enforce those caps. (2) It thus attached specific policy consequences to specific actions (or inaction) by future Congresses. If, for example, Congress in 2016 exceeded the spending caps for that year, the sequestration prescribed five years earlier would take hold. But succeeding Congresses chafed under the strictures. With considerable self-congratulation, the 113th Congress relaxed the spending caps for fiscal years 2014 and 2015 in the Bipartisan Budget Act of 2013, the 114th Congress relaxed the spending caps for fiscal years 2016 and 2017 in the Bipartisan Budget Act of 2015, and the 115th Congress did the same for fiscal years 2018 and 2019 in the Bipartisan Budget Act of 2018. (3) Those developments surprised few close observers of the federal budget process. The idea that a later Congress would adhere to significant spending caps enacted by an earlier Congress seemed fanciful from the start.

But what if the 112th Congress had not simply set out specific policy consequences that a later Congress could reverse but instead had purported to bind its successors? What if the Budget Control Act of 2011 had provided that no later Congress could repeal or modify the spending-cap and sequestration provisions? The possibility of such legislative entrenchment presents an enduring puzzle. For decades, legal scholars have debated whether legislative entrenchment is possible under the Constitution and desirable as a matter of policy. That debate has produced surprisingly limited insights. The aim here is to present a fresh analysis of the puzzle through a particular focus on federal fiscal policy.

In general terms, legislative entrenchment is legislative action that prevents or hinders action by a simple majority in a subsequent legislature. The dominant position among legal scholars--including Charles Black, Aaron-Andrew Bruhl, Erwin Chemerinsky, David Dana, Julian Eule, Catherine Fisk, Paul Kahn, Michael Klarman, Susan Koniak, John McGinnis, Michael Rappaport, John Roberts, Stewart Sterk, and Laurence Tribe--holds that legislative entrenchment is unwise, uncommon, and unconstitutional. (4) In a prominent criticism of that position, Eric Posner and Adrian Vermeule concede that legislative entrenchment is rare but argue that it is nonetheless sound as a matter of both policy and constitutional law. (5)

The two sides distinguish between what can be called "hard entrenchment" and what can be called "soft entrenchment." "Hard entrenchment" is legislative action that strictly binds a simple majority in a subsequent legislature. (6) A federal statute requiring Congress to enact a balanced budget and also prohibiting Congress from repealing or modifying the requirement, whether absolutely or by simple majority, would be a case of hard entrenchment (assuming the prohibition were effective). "Soft entrenchment," by contrast, is legislative action that impedes (but does not strictly bind) a simple majority in a subsequent legislature, thereby making a change to the policy status quo by simple majority more difficult or less likely than it otherwise would be. Soft entrenchment covers a broad spectrum of possible legislative action. At one end, every statute entrenches a policy outcome simply by setting that outcome as the status quo. A budget passed by Congress for a subsequent fiscal year is at least modestly entrenched because any change must overcome a legislative bias favoring the status quo. At the other end, the Senate filibuster and the Senate rule that nominally protects the filibuster entrench the status quo to a much greater extent; defeating a filibuster requires either a supermajority coalition to invoke cloture or a majority agreement to amend the standing rules of the Senate. Soft entrenchment might require a subsequent majority to navigate additional obstacles, but it ultimately does not prevent a subsequent majority from acting.

Note that, as the term is used here, legislative entrenchment is legislative action that either strictly binds or impedes a simple majority in a subsequent legislature. This implies three important points. First, legislative entrenchment requires legislative action and therefore excludes structural features of the legislative process imposed by the Constitution. (7) The veto power and the requirement of bicameralism and presentment, for example, undoubtedly entrench the status quo, (8) but they do not constitute legislative entrenchment because they are not endogenous to the legislature. Congress cannot change those features of the legislative process. Second, legislative entrenchment binds or impedes a simple majority in a subsequent legislature. Every legislature must start somewhere, and the starting point, in the absence of a contrary constitutional mandate, is simple majoritarianism. A legislature may adopt supermajority requirements as part of its internal structure, such as the suspension-of-the-rules procedure in the House (which requires a two-thirds majority) or the filibuster in the Senate (which requires a three-fifths majority). But such structures are themselves subject to revision, and the bedrock principle for institutional change is simple majoritarianism. Third, legislative entrenchment affects a simple majority in a subsequent legislature. Action by any Congress to bind or impede itself during its two-year constitutional term may present interesting questions, but it does not present the problem of legislative entrenchment.

The conventional academic analysis focuses almost exclusively on hard entrenchment; scholars generally consider soft entrenchment unimportant and uninteresting. That approach leads to analytic dead ends. The two sides debate whether hard entrenchment is constitutional and, separately, whether it makes for good policy. Those controversies are not easily resolved. The constitutional analysis is problematic, with superficially persuasive arguments for and against hard entrenchment falling apart on close examination. The policy analysis forces a seemingly unavoidable choice between flexibility and commitment, both of which can be attractive attributes of legislative action. But there is also a certain pointlessness to those controversies. Hard entrenchment in federal statutory law may not exist at all, and it probably is not even structurally possible. With so little at stake, the conventional analysis of hard entrenchment quickly devolves into an academic parlor game.

Soft entrenchment, however, presents entirely different considerations. Although largely ignored by the conventional analysis, soft entrenchment is commonplace. Even if an earlier Congress does not strictly bind a simple majority in a later Congress, the earlier Congress can make it more or less difficult for a simple majority in a later Congress to change policy outcomes. The earlier Congress might embed in a particular statute a provision that purports to insulate the statute from repeal or modification, thereby making a later Congress at the least override the anti-repeal or anti-modification provision. The earlier Congress might also design its internal structure and procedures in a way that makes it more difficult to change the policy status quo, such as by setting up committees that are not representative of the full House or Senate. Such legislative actions have entrenching effects, and once one turns to these, the entrenchment landscape looks very different. What formerly seemed uncommon, unwise, and unconstitutional reveals itself to be universal, unavoidable, and constitutionally unremarkable.

Soft entrenchment is of two general types. In the first, which can be called "deliberate soft entrenchment," Congress enacts statutes and adopts rules for the specific purpose of entrenching the status quo. The Budget Control Act of 2011 is a case of deliberate soft entrenchment. By establishing caps on discretionary spending for future years and setting sequestration as the enforcement mechanism for those caps, the 112th Congress did not actually bind its successors; it did not even purport to bind its successors. But the 112th Congress intended to impede majorities in future Congresses by requiring that those majorities affirmatively suspend sequestration in order to spend in excess of the caps. Ultimately, deliberate soft entrenchment binds a later Congress only to the extent that the later Congress chooses to treat itself as bound. In the case of the Budget Control Act of 2011, later Congresses have chosen to treat themselves as bound very loosely, if at all, by the 112th Congress.

In the second type of soft entrenchment, which can be called "incidental soft entrenchment," Congress enacts statutes and adopts rules for the general purpose of organizing itself and its activities. The House and the Senate have numerous internal rules, norms, and practices that systematically, although generally unintentionally, entrench the status quo. The distribution of agenda-setting power among chamber leaders and standing committees, the establishment of numerous veto gates, the formal and informal barriers to floor amendments, the rules of debate, and the mechanisms for resolving interchamber differences all anchor existing policy to some degree. These procedures are strictly necessary for the legislature to function. Institutions based on majority rule, such as Congress, require organizational structures to stabilize policy outcomes, and soft entrenchment is an inevitable consequence of these structures. Because it specifically authorizes the House and the Senate to determine their own rules of proceeding, the Constitution cannot be read to bar such entrenchment.

The incidental soft entrenchment attributable to legislative organizational structures is significant in the fiscal-policy setting. The annual federal budget process, which culminates in the enactment of twelve appropriations bills and numerous authorization bills, is thick with institutional obstacles that frustrate changes to the policy status quo. Often, the outcome is the enactment of one or more continuing resolutions that carry forward existing budget policy with little or no change. Additionally, large portions of fiscal policy--including major social-welfare entitlements and tax expenditures--are beyond the reach of the annual budget process. Revisions to social-welfare entitlements and tax expenditures must overcome all the internal institutional obstacles that generally hinder affirmative legislation action. Not surprisingly, entitlement and tax reform occur very rarely.

Those who prefer greater flexibility in federal fiscal policy--potentially giving Congress more latitude to reduce deficit spending, reform entitlements, and repeal tax preferences--must contend with the problem of incidental soft entrenchment. Fiscal policy is entrenched in part because of legislative organizational structures. But weakening those structures in order to disentrench fiscal policy would be costly. Specifically, it would undermine legislative stability, degrade the quality of legislative information, and frustrate legislative dealmaking. As with so many policy problems, there are trade-offs to be made.

The argument proceeds as follows. Part II examines the existing academic debate about hard entrenchment and argues that hard entrenchment is highly improbable. Part III examines soft entrenchment and explains further the distinction between deliberate soft entrenchment and incidental soft entrenchment. Part IV examines the place of incidental soft entrenchment in federal fiscal policy. The part argues that, for better or worse, incidental soft entrenchment brings meaningful stability to budget policy, social-welfare entitlements, and tax expenditures. It also argues that introducing greater flexibility by reducing incidental soft entrenchment would involve trade-offs on institutional design and could compromise the quality of fiscal policy legislation.

II

HARD ENTRENCHMENT

The conventional analysis of legislative entrenchment breaks into two sharply divided positions. The longstanding consensus, articulated more than fifty years ago by Charles Black, holds that legislative entrenchment is something that, "on the most familiar and fundamental principles, so obvious as rarely to be stated, no Congress for the time being can do." (9) The claimed obviousness notwithstanding, many legal scholars have detailed their objections to entrenchment. (10) In 2002, Posner and Vermeule challenged the consensus, arguing that "legislatures should be allowed to bind their successors" and that, in any event, "entrenchment is ... constitutionally permissible." (11)

Both sides distinguish between hard entrenchment and soft entrenchment, although they do not use those terms. For the most part, scholars consider only the former to constitute actual entrenchment. Posner and Vermeule define "entrenchment" as "the enactment of either statutes or internal legislative rules that are binding against subsequent legislative action in the same form." (12) Fisk and Chemerinsky also argue that a statute or legislative rule constitutes entrenchment if it "binds" future legislatures. (13) Dana and Koniak define "legislative entrenchment" as "a legal hierarchy in which the will of a past legislature trumps the will of a present legislature." (14) And Eule describes entrenchment as "[a] legislature ... inalterably dictating] the future." (15) Both sides recognize that every legislative act has future consequences, and both sides recognize that, for reasons internal and external to Congress, it often is difficult for a legislative majority to work its will. But the conventional analysis considers those points to be qualitatively different. (16) When they talk about entrenchment, legal scholars generally refer to hard entrenchment. (17)

Scholars debate hard entrenchment, but that does not mean that it deserves the attention. The concept itself is problematic. H.L.A. Hart, in considering the consensus position that one Parliament cannot strictly bind a successor Parliament, said that "no necessity of logic, still less of nature, dictates" that outcome; rather, "it is only one arrangement among others, equally conceivable, which has come to be accepted with us as the criterion of legal validity." (18) Hart exactly captures the threshold inquiry. A legal system may allow hard entrenchment, but it is a separate question whether any particular legal system does allow it.

In the United States, the constitutional framework for federal legislation effectively precludes the possibility of hard entrenchment. The Rules of Proceedings Clause, which provides that "[e]ach House may determine the rules of its Proceedings," makes Congress the final arbiter of its own procedures. (19) Congress--and only Congress--determines whether a legislative measure has been validly enacted. (20) As a practical matter, a later Congress cannot be bound by an ostensibly hard-entrenching statute or rule from an earlier Congress because any legislative action by the later Congress inconsistent with that statute or rule would be self-validating under the Rules of Proceedings Clause. In other words, the decision of the later Congress to ignore the ostensibly hard-entrenching statue or rule from the earlier Congress would have the force of law as long as the later Congress says that it has the force of law. Any "binding" effect of what purports to be hard entrenchment derives solely from a decision of a later Congress to treat itself as though it were bound by the action of the earlier Congress rather than from the action of the earlier Congress.

Consistent with that, scholars have struggled to identify meaningful examples of hard entrenchment in federal legislation. Eule notes that instances of entrenchment "are rare indeed." (21) Posner and Vermeule agree. (22) Eule cites only two examples: an "attempt by [the] Ohio legislature to permanently establish the county seat of Mahoning County at Canfield" and an authorization by the 67th Congress that a "committee investigation ... continue until the end of the 68th Congress." (23) Posner and Vermeule also cite the Mahoning County example and "the federal statute at issue in Reichelderfer v. Quinn, which 'perpetually dedicated' certain public lands in the capital for use as Rock Creek Park." (24) One of these involves a state statute, and the other has not been tested by a subsequent congressional majority. Fisk, Chemerinsky, Posner, and Vermeule consider the two-thirds supermajority requirement for changing the Senate cloture rule to constitute entrenchment. (25) But that view is not correct. (26) At the very least, then, congressional hard entrenchment is extraordinarily rare; the better view is that it simply does not exist.

The dubious status of hard entrenchment in federal legislation raises the possibility that the conventional analysis, however interesting and sophisticated, is pointless. Although theoretically possible, hard entrenchment in federal legislation is highly improbable as a matter of process and structure. One might suppose that the conventional analysis nonetheless yields meaningful insights. Perhaps by mooting what amounts to an elaborate hypothetical question about hard entrenchment, the conventional analysis reaches constitutional and policy answers about soft entrenchment. But that optimism proves misplaced. On both constitutional and policy considerations, the focus on hard entrenchment leads to blind alleys.

A. Constitutional Analysis

Opponents and proponents of hard entrenchment have debated its constitutional status at length, but their analyses provide little meaningful guidance. Consider first the textual arguments against entrenchment. (27) No provision in the Constitution expressly permits or prohibits legislative entrenchment, so both sides have searched for indirect textual support. Roberts and Chemerinsky argue that the Rules of Proceedings Clause provides "the best constitutional basis for arguing that legislative entrenchment is not permitted by our Constitution." (28) The clause, they correctly note, makes the House and the Senate largely sovereign over their internal affairs. (29) Roberts and Chemerinsky then assert that the clause "guaranteejs] that each Congress over time will exercise equal plenary authority over its enactment process." (30) If legislative entrenchment were permissible, they argue, an earlier Congress (which they call "Congress One") would effectively "overturn" the Rules of Proceedings Clause with respect to a later Congress (which they call "Congress Two"). (31) They conclude that the Rules of Proceedings Clause "guarantees that both Congress One and Congress Two will have the full authority to structure their proceedings as they see fit." (32)

Their argument works only by assuming the point to be proven. (33) The Rules of Proceedings Clause simply does not say how it applies over time. Consider two readings of the clause: first, each chamber of Congress can adopt rules of proceeding that bind that chamber only for one Congress; second, each chamber of Congress can adopt rules of proceeding that bind that chamber for more than one Congress. (34) On the first reading, it would violate the Rules of Proceedings Clause if the House or the Senate were to impose a supermajority requirement for legislative action by a subsequent House or Senate. On the second reading, it would violate the Rules of Proceedings Clause if the House or the Senate were unable to impose a supermajority requirement on a subsequent House or Senate. The question is whether the rulemaking authority of an earlier Congress is limited by the rulemaking authority of a later Congress (the first reading), or whether the rulemaking authority of the later Congress is limited by the rulemaking authority of the earlier Congress (the second reading).

Nothing in the Rules of Proceedings Clause favors one reading over the other; both are fully consistent with the text. Roberts and Chemerinsky arbitrarily choose the first. They assume that every later Congress must have the same power under the Rules of Proceedings Clause that every earlier Congress has. From there, they reason that, unless the first reading is correct, it would not be the case that reach Congress over time will exercise plenary authority over its enactment process." (35) But whether "each Congress over time" does or does not "exercise plenary authority over its enactment process" is exactly the question that Roberts and Chemerinsky were supposed to answer. Once the circularity is set aside, the analysis returns to the starting point: The Rules of Proceedings Clause does not say whether each chamber may adopt rules that bind that chamber across time or whether each chamber may bind itself only for a single Congress. The clause poses the question, but it does not provide the answer. (36)

The non-textual arguments against legislative entrenchment are also unsatisfactory. (37) The most prominent of these maintains that legislative entrenchment violates the constitutional principle of majoritarianism. Different scholars make the argument in different terms, but the basic structure is the same: Entrenchment by one Congress, which is elected to represent a particular majority of voters, prohibits the exercise of equal sovereignty by a later majority of voters. (38) Paul Kahn puts the point this way: "Legislatures ... may not try directly to control future legislatures. To do so is to assert authority where there is none. It is to transgress on the shadowy concept of popular sovereignty which remains always inalienable and complete." (39) In other words, legislative entrenchment intrudes on the sovereignty of popular majorities across time in the same way that (per Roberts and Chemerinsky) it intrudes on the rulemaking power of Congress across time.

Again, the argument is not persuasive. First, the general appeal to majoritarianism is not dispositive because the Constitution promotes majoritarianism at certain points, such as representation in the House, and flatly impedes it at other points, such as representation in the Senate. The incomplete and imperfect commitment to majoritarianism provides no guidance on questions that the constitutional text itself does not address. Second, the argument has the question-begging problem seen in the Roberts and Chemerinsky argument under the Rules of Proceedings Clause. The complete authority that Kahn and others argue is held by each temporal majority is necessarily ambiguous. If an earlier majority binds a later majority, it is clear that the authority of the later majority is not "inalienable and complete" (to use Kahn's phrase). But if the earlier majority is unable to bind the later majority, it is clear that the authority of the earlier majority is not "inalienable and complete" either. The question is whether the principle of majoritarianism privileges the earlier majority or the later majority. Simply pointing out that legislative entrenchment allows an earlier majority to trump a later majority identifies the problem, but it does not answer it. (40)

Posner and Vermeule correctly point out that no provision in the Constitution expressly prohibits legislative entrenchment. Yet they can identify no provision that expressly permits it. Posner and Vermeule suggest that the power to entrench may be part of the Vesting Clause, under which "[a]ll legislative Powers herein granted" are vested in Congress. They reason that the Constitution does not specifically forbid entrenchment--in contrast, for example, to the specific prohibition on enacting an ex post facto law. That argument is weak. For the argument to succeed, "[a]ll legislative Powers herein granted" would have to encompass all possible legislative power. In other words, Article I would have to begin with comprehensive legislative power and then specifically exclude certain specific powers, such as the power to enact an ex post facto law. That position is inconsistent with the enumeration of congressional powers set forth in Article I, Section 8. (41)

Posner and Vermeule also argue that the constitutional entrenchment effected by Article V supports legislative entrenchment by analogy. (42) Article V imposes a supermajority requirement on Congress and the states for all constitutional amendments and a unanimity requirement on the states for any constitutional amendment changing representation in the Senate. Certainly, as Posner and Vermeule argue, the constitutional entrenchment in Article V is a plain indication that the Constitution tolerates entrenchment in specific cases; and, certainly, as they argue, the constitutional entrenchment in Article V is a plain indication that the Founders thought about entrenchment. But the provision for constitutional entrenchment under Article V indicates nothing about the status of legislative entrenchment. (43)

B. Policy Analysis

The policy analysis of hard entrenchment has not been much more productive than the constitutional analysis. The most compelling arguments made for and against hard entrenchment pit the desirability of governmental flexibility against the desirability of governmental pre-commitment. (44) Opponents argue that hard entrenchment necessarily reduces policy flexibility, thereby obstructing the capacity of future legislators to work their will. Eule argues that, by transferring authority from later to earlier legislative majorities, entrenchment "prevents those with the greatest knowledge of societal needs from acting." (45) Roberts and Chemerinsky find legislative entrenchment "dangerous" because it magnifies the effects of "temporary radical majorities" in the legislature, diminishes legislative responsiveness to "changing national consensus," and inhibits legal reform for "changing social and economic conditions." (46) "Good government," they argue, "requires that each legislature and each public majority reassess the need for new policies." (47)

Proponents reply that legislative entrenchment facilitates long-term governmental commitments and precludes the possibility of post hoc opportunism by future legislative majorities. Posner and Vermeule argue that "[e]ntrenchment enables a government to make a credible pre-commitment that it will not hold up a person (or firm or institution or country) from whom it seeks certain actions." (48) This, they point out, reduces the costs of governmental action. (49) Additionally, the proponents argue, entrenchment helps legislators commit among themselves to resist interest-group pressure. McGinnis and Rappaport point to the House supermajority requirement for tax-rate increases as "a modest precommitment by the majority not to go down a road that will make everyone worse off in the end as concentrated interest groups demand expenditures that beggar the nation as a whole." (50) In effect, entrenchment proponents maintain that the interests of present and future majorities can be served by a present majority's decision to alienate legislative policy flexibility.

The preoccupation with hard entrenchment has led scholars to frame the policy analysis as a sharp choice between governmental policy flexibility and governmental pre-commitment. Hard entrenchment involves policy outcomes enacted at one time that cannot be readily changed at a later time, and it therefore suggests, as a normative matter, a dichotomy between the flexibility to revise those outcomes and the capacity of government credibly to bind itself against the possibility of such revision. But here the conventional analysis effectively backs itself into a corner. Obviously, there are advantages both to governmental policy flexibility and to governmental pre-commitment. Neither is always superior or preferable to the other, and it is not necessary that Congress pursue one to the exclusion of the other. Whether flexibility trumps precommitment or pre-commitment trumps flexibility depends on the specific considerations presented by a particular question. There is much to be said for pre-commitment when the government sells its debt in the bond markets; a credible pledge against repudiation of the debt reduces the government's borrowing cost. By contrast, policy flexibility seems particularly desirable when government attempts to act in areas that are sensitive to changing conditions, such as national security. The conventional analysis thus fails to justify a compelling policy position on legislative entrenchment. Although flexibility and pre-commitment are important bases for assessing the stability of legislative outcomes, the sharp dichotomy between them is neither necessary nor desirable.

III

SOFT ENTRENCHMENT

Hard entrenchment is highly unlikely, perhaps even structurally impossible. But soft entrenchment is commonplace and, in fact, unavoidable. Congress occasionally enacts statutes and adopts legislative rules for their entrenching effects. Although they fall short of hard entrenchment, these statutes and rules anchor the status quo to varying degrees. But still more interesting and more important are the incidental entrenching effects of the basic structures and processes of legislative organization. Every legislative body requires organizational structures and processes to function, and they erect obstacles to changing the policy status quo. The soft but ubiquitous entrenching effects of legislative organization matter considerably more than the hard but hypothetical entrenching effects that dominate the academic debate. In short, legal scholars have missed the real significance of legislative entrenchment.

A. Deliberate Soft Entrenchment

The conventional analysis is not wrong to assume that one Congress may try to prevent a future Congress from repealing or modifying a statute. But entrenchment in practice is best understood as a continuum. Few scholars recognize this point. Daryl Levinson argues that "formal, legal" entrenchment "is clearly a matter of degree." (51) Posner and Vermeule note that different federal statutes and legislative rules may effect different degrees of entrenchment, although they argue that statutes and rules not strictly binding on a successor Congress do not really count as entrenchment. (52) The relevant question is not whether a statute or legislative rule entrenches the policy status quo but how much it does so.

The answer depends on the deference shown by a later Congress. In theory, the majority in every new Congress could sweep aside all the statutes and legislative rules put in place by earlier Congresses; in practice, new majorities leave almost all such statutes and rules in place. The reasons for doing so usually have nothing to do with attempts by earlier Congresses to entrench their work. Each new Congress has little time to pursue its legislative agenda, and the status quo normally provides a tolerable basis from which to make the desired policy interventions. But even when an earlier Congress has purported to prescribe an outcome for a later Congress, the "bindingess" of the earlier action depends only on the extent to which the later Congress treats itself as bound by that action.

Consider several statutes and rules put in place for the purpose of favoring the status quo. In 1871, Congress ended the practice of making treaties with Native American tribes; it did so by means of a statute that purported to bind future Congresses. Section 1 of the Appropriations Act of March 3, 1871 provides that "hereafter no Indian nation or tribe within the territory of the United States shall be acknowledged or recognized as an independent nation, tribe, or power with whom the United States may contract by treaty." (53) The reason for the prohibition was the House's resentment of the Senate's outsized role in setting Native American policy through its exclusive power to ratify treaties. (54) Although the prohibition undoubtedly is not binding, each subsequent Congress has honored it. (55) The entrenchment here derives from the willingness of later Congresses to treat themselves as though they were bound by the earlier Congress.

In other cases, Congress attempts to anchor the status quo by stipulating particular consequences for the action or the inaction of a subsequent Congress. Consider the Defense Base Closure and Realignment Act of 1990 (commonly known as the "Base Closure and Realignment Act"). (56) Near the end of the Cold War, Congress determined that the United States had too many domestic military facilities but found that closing specific facilities was politically difficult. (57) The Base Closure and Realignment Act establishes a process under which the Defense Base Closure and Realignment Commission reviews recommendations of the Defense Secretary for facility closures and then submits its own recommendations to the President. (58) The President, upon approving the recommendations, presents them to Congress. (59) Unless Congress formally rejects them within 45 days, the recommendations have the force of law. (60) The Base Closure and Realignment Act was passed by the 101st Congress but prescribed policy outcomes for the 102nd, 103rd, and 104th Congresses. Although those subsequent Congresses could have voted to override the statute, it exerted a clear pull on policymaking. The 102nd, 103rd, and 104th Congresses dutifully followed the procedures laid down by the 101st Congress. Again, it is the deference of the later Congresses that effects entrenchment.

Senate Rule XXII.2 also has entrenching effects. That rule imposes a three-fifths supermajority requirement to invoke cloture for any pending measure, other than a measure to amend the Senate Rules, and imposes a two-thirds supermajority requirement to invoke cloture for any measure to amend the Senate Rules. On its face, Senate Rule XXII.2 entrenches the three-fifths supermajority requirement for cloture by setting a two-thirds supermajority requirement to bring any change to the three-fifths requirement to a vote. For this reason, legal scholars have cited it as the pre-eminent example of hard entrenchment. (61) But the two-thirds requirement in fact binds a current Senate majority only to the extent that the current majority treats itself as bound. During the second half of the twentieth century, the president of the Senate opined more than once that a simple majority can set aside the two-thirds requirement, and the Senate in 1975 actually (although briefly) upheld a parliamentary ruling to that effect by a simple majority vote. (62) The Senate removed all doubt on the issue when, twice in the last five years, simple majorities set aside the two-thirds requirement in order to amend the rules for debate on presidential nominations. (63) Thus, adherence to Senate Rule XXII.2 is a function of deference, rather than the actual bindingness of the two-thirds requirement.

Legislation aimed at controlling federal budget deficits has been less entrenching. More than thirty years ago, Congress passed the Balanced Budget and Emergency Deficit Control Act of 1985 (generally known as the "Gramm-Rudman-Hollings Act"). (64) The Gramm-Rudman-Hollings Act set out specific deficit-reduction targets and provided for automatic spending cuts if Congress failed to meet those targets. (65) Five years later, the Budget Enforcement Act of 1990 amended the Gramm-Rudman-Hollings Act to set limits on discretionary spending and to impose "pay-as-you-go" (PAYGO) requirements, under which any new spending or tax decreases had to be offset in full. (66) More recently, the Budget Control Act of 2011 resurrected the Gramm-Rudman-Hollings approach of imposing automatic spending cuts if discretionary federal spending exceeded specific spending caps. (67)

All three statutes were intended to have entrenching effects. In each case, the enacting Congress provided specific consequences if its successors enacted (or failed to enact) federal budget legislation. But, again, the entrenchment is soft because the mandates of all three statutes are subject to override by majority vote. Congress set aside the strictures of the Gramm-Rudman-Hollings Act and the Budget Enforcement Act of 1990 several times during the 1980s and the 1990s. And the Budget Control Act of 2011 certainly erected no meaningful obstacles in subsequent Congresses. For fiscal years 2014, 2015, 2016, and 2017, Congress relaxed the spending caps, first through the Bipartisan Budget Act of 2013 (68) and then through the Bipartisan Budget Act of 2015. (69) In February of 2018, Congress relaxed the spending caps for fiscal years 2018 and 2019 as well. (70)

Even Congressional action purporting to provide for hard entrenchment would bind a later Congress only to the extent that the later Congress treated itself as bound. Consider again the hypothetical example of the 112th Congress, in passing the Budget Control Act of 2011, purporting to prohibit any later Congress from repealing or modifying the spending-cap and sequestration provisions. As shown in Part II, that ostensible hard entrenchment would come up short. The constitutional processes for federal legislation give Congress the ultimate authority to determine what constitutes enactment of a law. Thus, a later statute that repealed or modified the spending-cap and sequestration provisions would be valid as long as the enacting Congress declared it valid. That said, the prohibition from the earlier Congress could have soft-entrenching effects on later Congresses.

The claim here is limited but important: in practice, actual legislative entrenchment is soft entrenchment, and soft entrenchment is a matter of degree. The hard entrenchment that preoccupies the conventional analysis is more theoretical than real. But even if a legislative majority cannot be strictly bound by its predecessor, different exercises of legislative authority can have different entrenching effects. Congress has been more willing to work around the Gramm-Rudman-Hollings Act and the Budget Control Act of 2011 than the prohibition on Native American treaties, the Base Closure and Realignment, or (until recently) Senate Rule XXII.2. These statutes and legislative rules represent different points on a continuum of soft entrenchment. (71)

Because fiscal-policy entrenchment--particularly regarding the budget deficit--has been extraordinarily weak, it follows that Congress generally must not want to consider itself bound by earlier Congresses in this area. This is not for lack of effort by the earlier Congresses. In 1985, 1990, and 2011, Congress enacted legislation to reduce long-term budget deficits by setting out specific consequences in the event of certain action or inaction by a later Congress. In all three cases, the legislation was the product of intense and prolonged negotiations involving both political parties, both legislative chambers, and both political branches. The critical failure each time was the unwillingness of the later Congresses to live with the fiscal-policy outcomes provided by the earlier Congress.

B. Incidental Soft Entrenchment

The academic debate over hard entrenchment is a sideshow; the real action is with soft entrenchment. As shown above, statutes and legislative rules intended to effect entrenchment work only to the extent that a later Congress chooses to consider itself bound by an earlier Congress. Consequently, the entrenching effects of such statutes and rules vary widely, with legislation to control federal budget deficits having proved especially weak. But more meaningful entrenchment often results not from deliberate efforts to entrench a specific policy outcome but from efforts simply to organize the legislative process. Although ignored by the conventional analysis, the basic rules and procedures under which Congress organizes itself and conducts its business constitute pervasive, enduring, and significant mechanisms for anchoring the policy status quo. The entrenching effects of these legislative organizational structures is both general and largely incidental. Even so, they are of paramount importance, especially for fiscal policy. (72)

The Constitution provides little detail for the organization of the legislative branch, so the complex internal structure of Congress is largely determined by Congress itself. After vesting "[a]ll legislative Powers" in Congress and specifying that Congress comprises the House and the Senate, the Constitution sets the "Quorum to do Business" in the House and the Senate at a simple majority, requires that the House choose a "Speaker and other Officers," designates the vice president as the "President of the Senate," requires that the Senate choose "other Officers, including a President pro tempore," and stipulates that each chamber must "keep a Journal of its Proceedings" and must record "the Yeas and Nays ... on any question ... at the Desire of one fifth of those Present." (73) Beyond these and similar minima, the Constitution simply states that "[e]ach House may determine the Rules of its Proceedings." (74)

The internal rules, norms, and customs adopted under the Rules of Proceedings Clause fill in the constitutional outline. In broad terms, the organizational structure is the same in both chambers, with three major divisions that track the flow of legislation: committees, chamber floors, and conference committees. The House and the Senate both use standing committees as the principal vehicle for producing legislation. (75) On the floor, measures are debated, amended, and either passed or rejected by formal action. And ad hoc conference committees facilitate the resolution of differences between the two chambers. (76) These structures entrench the policy status quo to varying degrees. Most importantly, these structures confer agenda control on different legislative actors, thereby establishing veto gates in the legislative process. Because any measure must successfully navigate all the veto gates, the dispersal of proposal power and gate-keeping power among different legislators and different groups of legislators increases the chance that any particular measure either will not be enacted or, if enacted, will be modified to a position closer to the status quo.

Assume, for example, that a member of the House introduces a bill to change the status quo on a specific issue. Success is not simply a matter of securing the support of a majority in the House, a majority in the Senate, and the President. First, the House committee of jurisdiction must report the bill favorably to the floor. (77) There are veto gates here: The committee chair could keep the bill off the committee's agenda, and a majority of the committee members could vote against the bill. (78) If more than one committee has jurisdiction over the bill or if subcommittees have jurisdiction, the number of veto gates increases. Once favorably reported, the bill must be brought to the floor. There are veto gates there as well. The chair of the Rules Committee or a majority of the Rules Committee could refuse to report a rule for the bill, and the majority leadership could refuse to allow the bill to come to the floor. (79) Once brought to the floor, both the rule and the bill itself must win majority support.

Then, of course, the process begins again in the Senate, where there are parallel veto gates. The bill must pass through the committees and subcommittees of jurisdiction; it must be brought to the floor by the majority leader, often with the agreement of the minority leader (but without the need to pass through a rules committee); and it must win three-fifths supermajority support to overcome a filibuster. From there, any differences in the House and Senate versions must be resolved either through a conference committee or through successive chamber floor votes. In either case, the inter-chamber resolution process provides new opportunities for the conference committee, the House Rules Committee, the House and Senate leadership, and the House and Senate floors to block the bill.

The complex, interlocking, and overlapping structures of legislative organization thus make it difficult to change existing policy. Consider just two examples--the filibuster and the committee system. The entrenching effects of the former are probably more salient. Under Senate rules, any senator can object to ending floor debate on a pending measure; in that case, debate must continue unless three-fifths of the senators vote for cloture. The filibuster ensures that a simple chamber majority cannot bring a measure to a floor vote; instead, with limited exceptions, (80) passage of legislation in the Senate requires a supermajority. Although filibusters were relatively uncommon until the late 1960s, today it is generally assumed that a measure must have the support of 60 senators to come up for a vote on the Senate floor.

The filibuster obviously entrenches the policy status quo. (81) The point of the filibuster is to allow a minority of senators to block enactment of a measure favored by a simple majority and, thus, to impose a de facto supermajority requirement to change existing policy. Routine use of the filibuster in the contemporary Senate directly blocks measures that otherwise would be passed. But the filibuster has subtler entrenching effects as well. The threat of a potential filibuster induces sponsors to seek compromise on pending measures, thereby moving those measures closer to the policy status quo. And even when the filibuster does not actually prevent the passage of a measure, the filibuster slows the Senate's progress through its legislative calendar. Floor debate and cloture votes take time, and, by expanding the consideration of any one measure, the filibuster prevents the Senate from taking up other measures. (82)

In what may be the most infamous case of entrenchment through legislative organizational structure, the filibuster preserved the policy status quo on racial discrimination for decades. Beginning with an anti-lynching measure in 1922 and ending with the Civil Rights Act of 1964 and the Voting Rights Act of 1965, filibustering senators from southern states blocked ten civil rights measures and forced substantial compromise on five others. (83) The southern senators agreed to the relatively weak Civil Rights Act of 1957, the first federal civil rights measure since Reconstruction, primarily out of concern that failure to pass it would precipitate cloture reform and facilitate more robust civil rights legislation. (84) Even so, the full chamber had to endure Senator Strom Thurmond's filibuster of twenty-four hours and eighteen minutes, the longest one-man filibuster in chamber history. (85) Seven years later, the Senate finally broke the southern filibusters of civil rights legislation when it successfully invoked cloture after fiftyseven days of debate over the Civil Rights Act of 1964. (86)

The congressional committee system also has entrenching effects. The House and the Senate both use standing committees as the principal vehicle for producing legislation. (87) Committees historically have been stronger in the House than in the Senate, (88) and, within the House, institutional reforms in the early 1970s transferred significant authority from full committees to leadership, subcommittees, and the rank and file. (89) Shorter-lived reforms of the middle 1990s further consolidated the power of House leadership at the expense of committees. (90) And each committee always remains subject to the important but rarely used power of the floor to strip the committee of jurisdiction or to disband the committee entirely. (91) But even at the lowest point of their institutional influence, the standing committees still exercise primary authority over legislation. (92) The rules of the House and the Senate provide that each bill or resolution must be referred to the standing committee of jurisdiction. (93) That committee then has the exclusive authority to report or not to report the measure for consideration on the floor. (94) Although there are exceptions, (95) normal legislative process makes each committee effectively sovereign over matters within its jurisdiction.

The entrenching effects of the committee system are substantial. (96) The standing committees normally determine what measures advance to the floor, a power that encompasses two types of agenda control: proposal power or "positive agenda control"--the exclusive ability to propose a change on any issue within the committee's jurisdiction; and gate-keeping power or "negative agenda control"--the ability to block change on any such issue. Because each standing committee is responsible to a greater or lesser extent for the status quo within its jurisdiction, the preferred policy positions of standing committees generally are closer to the status quo than are the preferred policy positions of the floor medians. (97) Thus, relative to non-members, committee members have a greater preference for no change or incremental change rather than broader reform.

Both types of agenda control bias legislative outcomes toward the status quo. By monopolizing positive agenda control, each committee pre-empts all other legislators from advancing legislation on issues within the committee's jurisdiction. (98) No matter how interested a legislator may be in making a policy change, the legislator simply cannot bring a measure directly to the floor; instead, the legislator must go through the committee of jurisdiction. If the legislator is not assigned to that committee, she has particularly bad prospects for moving the measure forward. Committees are jealous of their jurisdiction and do not readily surrender policy entrepreneurship.

The chairs of the House Ways and Means Committee and the House Energy and Commerce Committee demonstrated these entrenching effects when they defeated the so-called "Pepper Bill" on the floor in June of 1988." The prior year, Representative Pepper introduced a bill on long-term home health care. (100) The bill was referred to Ways and Means and to Energy and Commerce, the two committees of jurisdiction, but neither took immediate action on it. (101) In early 1988, Representative Pepper decided to end run the committees by having the Rules Committee, of which he was chair, attach his bill to a separate measure reported by the Education and Labor Committee. (102) This angered Representative Rostenkowski and Representative Dingell, the respective chairs of the Ways and Means Committee and the Energy and Commerce Committee. They recognized the move as an attempt to usurp their proposal power, and they led the floor opposition to the special rule for the Education and Labor bill. (103) Although the Pepper Bill was popular with the Democratic majority, Representatives Rostenkowski and Dingell blocked it from the floor, and the policy status quo was preserved. (104)

The negative agenda control held by committees often has even stronger entrenching effects. (105) The chair of a standing committee can kill a measure by refusing to bring it to the full committee, and the full committee can kill a measure by refusing to report it. In the House, where non-germane floor amendments are not permitted, failure to win committee approval is almost always the end of the matter. In the Senate, where non-germane amendments are permitted, a legislator may attempt to end run the committee by offering a measure as an amendment to an unrelated bill or resolution. But even then, the committee system exercises an entrenching pull. Committees enjoy considerable deference on the floor, and the committee chair's stated opposition to the amendment normally carries substantial weight. Additionally, committee members are disproportionately represented on inter-chamber conference committees, making it more likely that a policy change approved over the standing committee's objections will not return to the floor for final passage.

The Senate Finance Committee demonstrated this entrenching capacity, again in the fiscal-policy setting, when it considered President Clinton's first budget proposal. In an effort both to reduce the federal budget deficit and to improve the environment, President Clinton included in the proposal a broad-based energy tax, known as the "BTU tax." (106) As a revenue measure, the BTU tax fell within the jurisdiction of the Ways and Means Committee and the Finance Committee. After making certain modifications, the Ways and Means Committee reported the BTU tax favorably, and the full House passed it. (107) But the Finance Committee, at the insistence of a single senator, killed the tax. Senator Boren, from energy-producing Oklahoma, opposed the BTU tax; as a Democratic member of the committee, his vote was necessary to report the budget's revenue provisions to the full Senate. (108) The Finance Committee approved the budget proposal (by a vote of eleven to nine) only after replacing the BTU tax with a motor-fuels tax. (109) The Finance Committee used its negative agenda control to move a pending measure closer to the policy status quo. (110)

IV

FEDERAL FISCAL POLICY

Federal fiscal policy tends to be stable over time. Spending programs remain in place over long periods; tax revenues remain relatively constant as a percentage of gross domestic product even as marginal tax rates change; and many tax preferences endure for decades. Perhaps most notably, determined efforts to reduce budget deficits over the long run--such as the Gramm-Rudman-Hollings Act and the Budget Control Act of 2011--dissipate in fairly short order. Why should fiscal policy be so entrenched, particularly against deficit reduction? Part of the answer lies in genuine policy preferences and political calculation. It must be more rewarding as a legislator to spend money than to reduce expenditures or increase revenues. But there is more to fiscal-policy entrenchment than that. Part of the answer lies in the incidental entrenching effects of legislative organizational structures, and that presents a conundrum for those who prefer a more flexible fiscal policy. The same organizational structures that tend to entrench fiscal policy also protect Congress from the chaos of pure majoritarian rule, promote the cultivation of policy expertise, and facilitate legislative dealmaking. In other words, there are trade-offs to consider. Eliminating or weakening legislative organizational structures to induce greater fiscal-policy flexibility would have broader institutional and policy consequences.

A. Incidental Soft Entrenchment and Fiscal Policy

The process set out in the Congressional Budget and Impoundment Control Act of 1974 for enacting the annual federal budget is thick with institutional obstacles that frustrate efforts to change the policy status quo, and the default response to a failure of that process is to carry forward the prior budget. Additionally, a substantial portion of federal spending is not even subject to the budget process. Standing appropriations cover social-welfare entitlements, such as Social Security and Medicare, and standing rules in the tax code cover tax expenditures, such as the deduction for home-mortgage interest and the exclusion for employer-provided health insurance. Changing the status quo for social-welfare entitlements and tax expenditures requires affirmative action by Congress, but the anchoring effect of legislative organizational structures generally helps to keep existing entitlement and tax policies in place.

1. The Federal Budget Process

Begin with the budget process. The Congressional Budget and Impoundment Control Act of 1974 sets out a multi-stage procedure for enacting a federal budget for each fiscal year (running from October 1 to September 30). The President submits a proposed budget to Congress. (111) The House and the Senate then pass a budget resolution, which sets revenue and spending levels for the next five years or more. (112) The budget resolution is not presented to the President; it does not have the force of law, although it does have binding effect for the House and the Senate. (113) Acting within the constraints of the budget resolution, each chamber passes twelve individual appropriations bills covering various policy areas, such as national defense, international aid, and general governmental operations. (114) The appropriations bills are presented to the President and, once enacted, have the force of law. Additionally, each chamber passes authorization bills to provide statutory authority for the specific federal programs covered by the appropriations bills. (115) As with the appropriations bills, the authorization bills are constrained by the budget resolution, are presented to the President, and have the force of law. In theory, this process adheres to a tight timeline: the President's budget proposal is due on the first Monday of the February preceding the start of the fiscal year; the budget resolution is to be passed by April 15; and the twelve appropriations bills are to be completed by June 30. (116) But apart from the submission of the President's budget proposal, these actions are rarely completed by the statutory deadline--if they are completed at all.

The budget process is fraught with institutional obstacles that, however unintentionally, favor the policy status quo. Foremost among these is the fragmentation of the budget into different components that, collectively, must run through virtually every congressional committee and win approval in multiple votes on the House and Senate floors. The budget resolution requires action by both the House Budget Committee and the Senate Budget Committee and a majority vote on the floor of each chamber. (117) There are numerous veto gates here. The budget resolution must secure support from the chair of the House Budget Committee, the chair of the Senate Budget Committee, a majority of the House Budget Committee members, a majority of the Senate Budget Committee members, the chair of the House Rules Committee, a majority of the House Rules Committee members, the Speaker, the House majority leader, a majority of the full House, the Senate majority leader, and a majority of the full Senate. Additionally, if the House and the Senate pass different budget resolutions, the competing resolutions may go to a conference committee--thus requiring agreement of a majority of the conference committee members and, once again, agreement of the House and Senate leadership and a majority of members on the floor of each chamber.

The problems are similar but often more intractable for the appropriations and authorization bills. Each of the twelve appropriations bills originates with a subcommittee of the House Appropriations Committee and a subcommittee of the Senate Appropriations Committee; each bill must pass those subcommittees, the full committees, and the chamber floors (including the possibility of a conference committee and renewed consideration on the chamber floors). The authorization bills originate in still more committees and subcommittees. Nearly all of the twenty-one standing committees in the House and the twenty standing committees in the Senate have authority over federal programs that require authorization legislation.

Navigating these obstacles is difficult, and in recent years Congress has failed more often than not. For fiscal years 1999, 2003, 2005, 2007, 2011, 2012, 2013, 2014, 2015, and 2017 Congress did not pass a budget resolution. (118) For fifteen of the thirty-five fiscal years beginning with 1978 and ending with 2013, Congress was unable to pass all the appropriations bills, and for virtually every one of those thirty-five years, Congress passed one or more of the appropriations bills after the start of the fiscal year. Congress has passed all appropriations bills before the start of a fiscal year only four times since fiscal year 1977. (119) Of course, there often are substantive policy disputes behind such failures, but the effects of those disputes are magnified both by the jurisdictional fragmentation among the budget, appropriations, and legislative committees and by the veto power of individual legislators and coalitions of legislators at each stage of the process.

The result is a bias for the fiscal-policy status quo. In theory, the failure to enact a budget and to pass appropriations bills forces a radical departure from the status quo--a shutdown of government operations. (120) But shutdowns, despite their political salience, rarely happen. Since the enactment of the Congressional Budget and Impoundment Act of 1974, there have been about eighteen full or partial shutdowns of the federal government, most of them lasting fewer than ten days. (121) By far the more common response to a breakdown in the budget process has been the passage of continuing resolutions that generally provide for ongoing appropriations at or near the levels set under the most recent budget and that generally prohibit use of appropriated funds for new programs or activities. (122) Continuing resolutions, by design, carry forward the policy status quo. (123) During the forty fiscal years starting with 1977 and ending with 2016, Congress passed 175 continuing resolutions that fully or partially funded government operations for a total of 8,064 days--the equivalent of more than twenty-two fiscal years. (124)

In short, the legislative organizational structures that shape the federal budget process tend to entrench the status quo. The budget resolution, the appropriations bills, and the authorization bills that collectively make up the federal budget must pass through virtually every committee in Congress--the same committees that put the status quo in place. Any departure from the status quo must navigate numerous veto gates. And when the process breaks down, as it often does, the preferred strategy is to enact a continuing resolution so that the budget status quo is held in place until the process successfully produces a new budget.

2. Social-Welfare Entitlements and Tax Expenditures

Although it often defaults to continuing resolutions, the budget process in theory requires Congress to revisit certain segments of federal spending every year. By contrast, other segments of fiscal policy--specifically, those providing funding for major social-welfare entitlements and tax expenditures--are not part of the regular budget process and are beyond the reach of annual appropriations. (125) Any change to the status quo for social-welfare entitlements and tax expenditures requires affirmative congressional action that must negotiate the institutional obstacles created by legislative organizational structures.

Consider the two largest social-welfare entitlement programs, Social Security and Medicare. The federal government paid out $911.4 billion in Social Security benefits during 2016. (126) Those benefits were not subject to annual appropriation; instead, they were funded through a broad-based wage tax, a modest benefits tax, and interest earned by the program's two trust funds. (127) The federal government paid out $669.5 billion in benefits under Medicare Parts A, B, and D during 2016. (128) Less than half that amount, $319.2 billion, was funded through the annual appropriations process; the remainder was funded through a broad-based wage tax, a modest benefits tax, premiums charged to beneficiaries, modest transfers from the states, and interest on the program's trust funds. (129) Thus, $1.2617 trillion in spending under these two programs was not subject to review and approval by Congress through the annual budget process. Instead, the payments were made under laws enacted many years ago.

Similarly, the federal government spends large amounts each year through tax expenditures, rules in the tax code providing preferential treatment for particular taxpayers or for taxpayers engaged in particular activities. Like much of the social-welfare spending under Social Security and Medicare, tax expenditures fall outside the appropriations process and, thus, are not subject to annual review and approval by Congress. (130) The total spent each year through tax expenditures is hard to determine. One estimate puts the amount for 2015 at $1,399 trillion, (131) but this figure is problematic. Tax expenditures have incentive effects that change behavior, and they interact with other provisions in the tax code. (132) Thus, enacting or repealing a $1 billion tax expenditure would not necessarily decrease or increase federal revenues by $1 billion, once all the relevant behavioral changes and interactions were accounted for. But certainly tax expenditures are sizable. For 2016, the cost of excluding employer-provided health insurance was $155.3 billion, the cost of taxing long-term capital gains and corporate dividends at a reduced rate was $130.9 billion, the cost of deferring non-U.S. earnings by U.S. corporations was $102.7 billion, and the cost of deducting state and local property taxes was $31.2 billion. (133)

Federal spending through social-welfare entitlements and tax expenditures represents an entrenched segment of federal fiscal policy. It is not that Congress cannot change these spending programs but that changing these spending programs is institutionally difficult. And changing them is difficult not because Congress has erected barriers around the programs with the specific intention of entrenching them but simply because Congress has organized itself in such a way as to inhibit affirmative legislative action. The structures of legislative organization result in an unintentional yet still meaningful bias for the status quo. The last major revision to Social Security occurred thirty-five years ago, in 1983, when the program was about to default on benefit payments. (134) The last major revision to Medicare occurred fifteen years ago, in 2003, and that revision had the effect of increasing annual spending through the addition of a prescription drug benefit. (135) And prior to the enactment of the Tax Cut and Jobs Act in December 2017, (136) it had been thirty-one years since the last major revision to the tax code to repeal or modify tax expenditures on a substantial scale. (137)

B. The Costs of Dis-Entrenching Fiscal Policy

There are cogent arguments for different policy positions on the federal budget deficit, social-welfare entitlements, and tax expenditures. The point here is not to argue that Congress should or should not reduce the deficit, that Congress should or should not reform entitlements, or that Congress should or should not repeal tax expenditures. But assume for the purpose of analysis that one simply wanted Congress to have greater latitude to change these elements of fiscal policy. Successful dis-entrenchment of deficit spending, entitlements, and tax preferences would require weakening important legislative organizational structures such as the committee system, internal managerial hierarchies, and voting procedures. That would be costly. Legislative organization has a life of its own--desirable in some respects and undesirable in others--that is completely independent of its entrenching effects. These structures provide institutional stability against position cycling, encourage specialization and the development of policy expertise, and facilitate durable agreements; consequently, weakening them would increase chaos on the chamber floors, degrade the quality of legislative information, and frustrate dealmaking among legislators.

1. Internal Legislative Stability

The structures of legislative organization induce stability in Congress. Begin with the possibility of instability. Arrow's Theorem holds that, under straightforward conditions for democratic action, majority voting readily yields open-ended cycling among any three or more policy options. (138) The theorem implies that, except in the uncommon case of a Condorcet winner, legislators must compromise pure majoritarian rule in order to avoid legislative chaos. (139) Following Arrow, William Riker demonstrated the instability of political coalitions and offered compelling historical evidence of manipulated position cycling among such coalitions. (140) Charles Plott showed that, in the absence of a Condorcet winner, voting yields non-cyclical results only when the individual preferences of legislators are perfectly symmetrical, (141) an extreme rarity in multidimensional policy space. (142) Richard McKelvey demonstrated that, if legislators vote sincerely, an agenda setter can construct a series of pairwise votes leading to any outcome, thereby using the voting agenda to control policy. (143) Subsequent analysis showed that even sophisticated voting imposes only loose constraints on the agenda setter's ability to manipulate legislative results. (144) Thus, the implementation of pure majority rule generally fails to produce durable equilibria. (145)

But why, if the outcomes of social choice are inherently unstable, do legislative bodies such as Congress demonstrate high levels of stability? (146) Kenneth Shepsle argues that specific institutional structures anchor collective decision-making. (147) He shows that the committee system and restrictiveamendment procedures combine with heterogeneous preferences to produce "structure-induced equilibria." (148) In effect, Shepsle domesticates the chaos implied by rational-choice theory; he demonstrates the possibility of legislative stability through institutional arrangements that restrict the domain of available policy options. (149) Shepsle's argument has led political scientists to re-examine the structural features of Congress that ground stability and that affect the substantive content of policy outcomes. (150)

Institutional analysis therefore implies that some level of incidental soft entrenchment is unavoidable. Congress must have organizational structures in order to function. The same institutional structures that incidentally perpetuate budget deficits, entitlements, and tax preferences tame the chaos of pure majoritarianism; weakening those organizational structures in pursuit of greater fiscal-policy flexibility would undermine institutional stability more broadly. At the extreme, maximum fiscal-policy flexibility could sharply reduce Congress's ability to act at all.

2. Quality of Legislative Information

Legislative organizational structures also improve the quality of legislative information. Keith Krehbiel's informational theory maintains that such structures provide legislators the opportunity and the incentives to invest in policy expertise. (151) Krehbiel argues that legislators act under conditions of uncertainty about the instrumental relationships between legislative policies and non-legislative ("real world") outcomes. (152) The legislature therefore establishes certain organizational structures, such as the committee system, to mitigate the uncertainty. The reasoning is straightforward. Congress as a whole benefits from "informational efficiency." (153) But the pursuit of informational efficiency encounters the usual collective-action problem. Policy expertise is costly to acquire and, once acquired by any one legislator, is potentially available to all legislators. Thus, the "benefits of policy expertise will be realized only if institutional arrangements are such that some legislators have strong incentives to specialize and to share their expertise with their fellow legislators."154

By assigning individual legislators to specific committees with independent jurisdictions, Congress provides legislators the opportunity to develop expertise in particular policy areas, (155) and by establishing and maintaining committees that cover the entire policy spectrum, the House and the Senate create institutional reservoirs of specialized knowledge. (156) An individual legislator, however, rationally may not want to reveal policy expertise to the full chamber, anticipating that the chamber might exploit the expertise to the legislator's disadvantage. (157) Congress therefore institutionalizes certain parliamentary prerogatives, such as restrictive floor-voting procedures, that allow committees to secure outsized gains in final policy outcomes. (158) These prerogatives increase the prospects that the committee will "credibly transmit[] private information to get a [floor] majority to do what is in the majority's interest." (159)

Again, there is a trade-off to be made in the dis-entrenchment of fiscal policy. Weakening legislative organizational structures, such as the committee system and related parliamentary prerogatives, may introduce greater flexibility into federal fiscal policy. But it may also erode incentives for individual legislators to invest in the acquisition and maintenance of policy expertise and, accordingly, degrade the quality of legislative information. The result could be a more flexible but less informed fiscal policy.

3. Legislative Dealmaking

Legislative organizational structures also facilitate stable dealmaking by legislators. Distributive theory argues that legislators organize Congress to produce mutual benefits such as "pork barrel projects, ... expenditures targeted to their districts, and policy outcomes desired by favored constituents." (160) The theory assumes that legislators have heterogeneous policy preferences, that few such preferences are held by a majority of legislators, and that legislators may secure gains from trade by exchanging support across different measures. (161) Because the "spot market" for trading votes presents problems of enforceability and long-term stability, (162) legislators establish institutional structures to facilitate both the formation and the enforcement of deals that deliver benefits to their constituents. (163)

Congress addresses the enforcement problems and the transaction costs of vote-trading through the committee system, which allows members to sort themselves according to the interests of their constituents. (164) Within any single committee, legislators can make enforceable deals without the need to trade votes across separate measures. (165) Additionally, each committee holds almost unchecked agenda control over policy matters within its jurisdiction. Distributive theory thus regards the committee system as "the formal expression of a comprehensive logrolling arrangement." (166) The committees' parliamentary prerogatives strengthen this arrangement. Once a measure has left a committee's direct control, floor amendments could undo intra-committee deals and resurrect the enforceability problems of measure-by-measure vote trades. But the parliamentary prerogatives stabilize legislative outcomes in favor of committee preferences. (167)

Weakening internal organizational structures, therefore, may undermine legislative dealmaking. This may be desirable in part. Those interested in reforming fiscal policy may think that undermining pork-barrel spending is exactly the right path to pursue, and they may be correct. It is also possible, however, that such legislative deals are needed to help Congress function--that pork-barrel spending provides the grease that facilitates action in other policy areas. If legislators cannot form stable logrolling arrangements on spending, they may not be able to reach agreement on other matters. Again, there is a trade-off to be made in dis-entrenching fiscal policy.

V CONCLUSION

The conventional academic debate about legislative entrenchment centers on hard entrenchment, but hard entrenchment is highly improbable--perhaps simply impossible--under the constitutional framework for federal legislation. By contrast, soft entrenchment is far more common. Of particular importance for fiscal policy is the incidental soft entrenchment induced by the structures of legislative organization, such as the congressional committee system and institutional voting procedures. Such structures bring stability to the federal budget process, to social-welfare entitlements, and to tax expenditures; weakening those structures in order to make fiscal policy more flexible may compromise internal legislative stability, legislative information, and legislative dealmaking.

Copyright [C] 2018 by Michael Doran.

This article is also available online at http://lcp.law.duke.edu/.

Michael Doran, Professor, University of Virginia School of Law. Thanks to James Ryan, Louis Michael Seidman, George Yin, and Lawrence Zelenak for comments on earlier drafts. Feci quod potui, faciant meliora potentes.

(1.) Budget Control Act of 2011, Pub. L. No. 112-25,125 Stat. 240 (2011).

(2.) Budget Control Act of 2011, [section][section] 101, 302, 125. See also BILL Heniff, Jr. et al., Cong. Research Serv., R41965, The Budget Control Act of 2011, at 11-14,27-32(2011).

(3.) Bipartisan Budget Act of 2018, Pub. L. No. 115-123,132 Stat. 64 (2018); Bipartisan Budget Act of 2015, Pub. L. No. 114-74,129 Stat. 584 (2015); Bipartisan Budget Act of 2013, Pub. L. No. 113-67,127 Stat. 1165 (2013); Grant A. Driessen & Marc Labonte, Cong. Research Serv., R42506, The Budget Control Act of 2011 as Amended: Budgetary Effects, at 6-7 (2015).

(4.) See generally Laurence h. Tribe, American Constitutional Law [section] 2-3 125 n.1 (2000); Charles L. Black, Jr., Amending the Constitution: A Letter to a Congressman, 82 Yale L.J. 189 (1972); Aaron-Andrew P. Bruhl, Using Statutes to Set Legislative Rules: Entrenchment, Separation of Powers, and the Rules of Proceedings Clause, 19 J.L. & pol. 345 (2003); David Dana & Susan P. Koniak, Bargaining in the Shadow of Democracy, 148 u. Pa. L. Rev. 473 (1999); Julian E. Eule, Temporal Limits on the Legislative Mandate: Entrenchment and Retroactivity, 1987 AM. B. found. res. J. 379 (1987); Catherine Fisk & Erwin Chemerinsky, The Filibuster, 49 stan. L. rev. 181 (1997); Paul W. Kahn, Gramm-Rudman and the Capacity of Congress to Control the Future, 13 hastings const. L. Q. 185 (1986); Michael J. Klarman, Majoritarian Judicial Review: The Entrenchment Problem, 85 geo. L.J. 491 (1997); John O. McGinnis & Michael B. Rappaport, Symmetric Entrenchment: A Constitutional and Normative Theory, 89 va. L. Rev. 385 (2003) [hereinafter McGinnis & Rappaport I]; John O. McGinnis & Michael B. Rappaport, The Constitutionality of Legislative Supermajority Requirements: A Defense, 105 Yale L.J. 483 (1995) [hereinafter McGinnis & Rappaport II]; John c. Roberts & Erwin Chemerinsky, Entrenchment of Ordinary Legislation: A Reply to Professors Posner and Vermeule, 91 cal. L. Rev. 1773 (2003) [hereinafter Roberts & Chemerinsky]; Stewart E. Sterk, Retrenchment on Entrenchment, 71 Geo. Wash. L. Rev. 231 (2003).

(5.) Eric A. Posner & Adrian Vermeule, Legislative Entrenchment: A Reappraisal, 111 Yale L.J. 1665,1666 (2002).

(6.) Admittedly, bindingness is an ambiguous concept, but as H.L.A. Hart maintained, it is "familiar to lawyers and tolerably clear in meaning." H.L.A. Hart, The Concept of Law 216 (2d ed. 1994).

(7.) It also excludes governmental action through contracting and property conveyancing. Cf. Christopher Serkin, Public Entrenchment through Private Law: Binding Local Governments, 78 U. CHI. L. Rev. 879,889-892 (2011) (analyzing how governments avoid anti-entrenchment rules by using private law mechanisms to make binding precommitments).

(8.) See generally Michael D. Gilbert, Entrenchment, Incrementalism, and Constitutional Collapse, 103 va. L. Rev. 631 (2017); Michael D. Gilbert, Optimal Entrenchment of Legal Rules (Univ. of Va., Law & Economics Research Paper No. 2017-10, 2017, Law & Legal Theory Research Paper No. 201725, 2017), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2970164 [https://perma.cc/AT6LDHZU],

(9.) Black, supra note 4, at 191.

(10.) See supra note 4.

(11.) Posner & Vermeule, supra note 5, at 1666.

(12.) Posner & Vermeule, supra note 5, at 1667 (emphasis added). See also Bruhl, supra note 4, at 373; Roberts & Chemerinsky, supra note 4, at 1777-78.

(13.) Fisk & Chemerinsky, supra note 4, at 250, 253.

(14.) Dana & Koniak, supra note 4, at 529 (emphasis added).

(15.) Eule, supra note 4, at 381 (emphasis added). Although Eule distinguishes between what he calls "absolute entrenchment" and other forms of entrenchment, all his entrenchment forms assume that a legislature prohibits certain actions by a simple majority in a subsequent legislature. Specifically, Eule categorizes entrenchment as "absolute," "procedural," "transitory," and "preconditional." He defines "absolute entrenchment" as entrenchment that denies a subsequent legislature the power to repeal a statute "for all time, under any conditions, and by whatever procedure." He says that procedural entrenchment "entails an attempt not to bind the future irrevocably, but to prescribe the 'manner and form' by which the promulgated directives can be changed," that transitory entrenchment "seeks to prevent alteration for a specified period of time only," and that preconditional entrenchment "purports to permit change only on the occurrence of a preordained event." Id. at 384-85. Thus, even the three non-absolute forms of entrenchment do more than simply impede action by a legislative majority: within their terms, they purport to prohibit majority action. See id. at 384 n.14.

(16.) See, e.g., Klarman, supra note 4, at 504-05; Roberts & Chemerinsky, supra note 4, at 1778, 1814-18; Sterk, supra note 4, at 232. Dana and Koniak flatly state that such effects of legislative acts should not be considered entrenchment. Dana & Koniak, supra note 4, at 530-31. But see Posner & Vermeule, supra note 5, at 1767 (arguing that legislative "policy choices become entrenched de facto through path dependence and inertia") (emphasis added). Even Posner and Vermeule, however, do not consider legislative inertia to constitute entrenchment. Id. at 1696-97.

(17.) Klarman uses the term "legislative entrenchment" to refer to "the so-called agency problem of representative government--elected representatives discounting their constituents' preferences in furtherance of their own perpetuation in office." Klarman, supra note 4, at 502. By contrast, he uses the term "cross-temporal entrenchment" to refer to the problem of a legislature binding the majority of a subsequent legislature. Id. at 504.

(18.) Hart, supra note 6, at 149. Cf. lon L. Fuller, THE Law in Quest of itself 35-36 (1940) (considering various arrangements of limitation on sovereign authority).

(19.) John C. Roberts, Are Congressional Committees Constitutional? Radical Textualism, Separation of Powers, and the Enactment Process, 52 CASE W. RES. L. REV. 489, 530-42 (2001). See also Roberts & Chemerinsky, supra note 4, at 1793.

(20.) Roberts, supra note 19, at 542. In United States v. Ballin, 144 U.S. 1, 5 (1892), the Supreme Court indicated that, subject to three specific limitations, the power of each chamber of Congress under the Rules of Proceedings Clause is "absolute and beyond the challenge of any other body or tribunal." The limitations are: (1) that the House or the Senate may not use the Rules of Proceedings Clause to "ignore constitutional restraints" (such as the constitutional requirement of a two-thirds vote in the Senate to convict the president in impeachment proceedings); (2) that the House or the Senate may not use the Rules of Proceedings Clause to "violate fundamental rights"; and (3) that any rule of the House or the Senate must have a "reasonable relation" to the "result which is sought to be attained." Although those limitations might invalidate a congressional rule in a specific case (for example, if the House adopted a rule requiring that all legislation promote the establishment of the Episcopal Church as the official religion of the United States), none of those limitations categorically implicates the exclusive authority of Congress to determine what legislative action has the force of law.

(21.) Eule, supra note 4, at 406.

(22.) Posner & Vermeule, supra note 5, at 1693.

(23.) Eule, supra note 4, at 406 n.122.

(24.) Posner & Vermeule, supra note 5, at 1667. See also Bruhl, supra note 4, at 374.

(25.) Fisk & Chemerinsky, supra note 4, at 245-52; Posner & Vermeule, supra note 5, at 1694-95. But see Roberts & Chemerinsky, supra note 4, at 1780 n.20.

(26.) See infra Part III. A. In a final effort to identify entrenchment examples, Posner and Vermeule point to statutory rules of statutory interpretation as "partially entrenched"--that is, as '"[!Intermediate between ... genuine entrenchment ... and ... pseudo-entrenchment." Posner & Vermeule, supra note 5, at 1697-98.

(27.) Although not exhaustive, the arguments considered here are the more plausible ones put forth in the conventional analysis.

(28.) Roberts & Chemerinsky, supra note 4, at 1789.

(29.) Id. at 1793.

(30.) Id. at 1794 (emphasis in original).

(31.) Id.

(32.) Id. at 1795 (emphasis in original).

(33.) It is important to distinguish the argument that Roberts and Chemerinsky make from the argument above (at pages 33-34) about the Rules of Proceedings Clause. Roberts and Chemerinsky argue that the clause prohibits hard entrenchment--that hard entrenchment is unconstitutional because it violates the clause. The argument above is that, whether or not hard entrenchment is constitutionally permissible, the exclusive authority of Congress under the clause to determine whether an act of Congress has the force of law precludes hard entrenchment as a practical matter. The difference between the two arguments is the difference between saying that hard entrenchment is invalid (the Roberts and Chemerinsky argument) and saying that hard entrenchment is without effect (the argument above).

(34.) A third reading--that neither chamber of Congress can adopt rules of proceeding that bind that chamber even for one Congress--does not engage the question of legislative entrenchment. As such, it is not distinct from the first reading for these purposes.

(35.) Roberts & Chemerinsky, supra note 4, at 1794 (emphasis in original). Roberts and Chemerinsky claim support for their assumption from a passage in United States v. Ballin, 144 U.S. 1,5 (1892), in which the Supreme Court refused to overturn a determination by the House that its quorum requirement had been satisfied. The Court, in describing the Rules of Proceedings Clause, said that "[t]he power to make rules is not one which once exercised is exhausted. It is a continuous power, always subject to be exercised by the house, and within the limitations suggested, absolute and beyond the challenge of any other body or tribunal." Id. See also Roberts & Chemerinsky, supra note 4, at 1794. Roberts and Chemerinsky read too much into this. The passage says that an exercise of rulemaking power by Congress does not forfeit the rulemaking power of a future Congress. It does not say that an exercise of rulemaking power by Congress may not bind a future Congress. Reading the passage as making the second point renders it incoherent. If the rules of an earlier Congress can bind a future Congress, the rulemaking power is not "absolute" at all times. But if the rules of an earlier Congress cannot bind a future Congress, the rulemaking power still is not "absolute" at all times. Again, either the rulemaking power of the earlier Congress must yield to the rulemaking power of the later Congress or the rulemaking power of the later Congress must yield to the rulemaking power of the earlier Congress. The Ballin passage does not answer that question.

(36.) See also McGinnis & Rappaport II, supra note 4, at 504. Chemerinsky in an earlier article had taken the position that the Rules of Proceedings Clause "is silent as to timing" and, thus, could not be used as textual support for the unconstitutionality of legislative entrenchment. See Fisk & Chemerinsky, supra note 4, at 246.

(37.) As with the textual arguments, the argument set forth in this section is not exhaustive of the non-textual arguments made in the conventional analysis. For example, Fisk and Chemerinsky cite a handful of Supreme Court decisions to conclude that "entrenchment ... violates a fundamental constitutional principle: One legislature cannot bind subsequent legislatures." Fisk & Chemerinsky. supra note 4, at 247-418. Of course, their argument then stands or falls with the correctness of the cases, but they generally do not go behind the decisions. See also Virginia A. Seitz & Joseph R. Guerra, A Constitutional Defense of "Entrenched" Senate Rules Governing Debate, 20 J.L. & POL. 1, 22-24 (2004).

(38.) See, e.g., Dana & Koniak, supra note 4, at 531-36; Eule, supra note 4, at 394-406; Klarman, supra note 4, at 499-501; McGinnis & Rappaport II, supra note 4, at 505-07.

(39.) Kahn, supra note 4, at 231.

(40.) Cf. Hart, supra note 6, at 149-50 ("These two conceptions of [Parliamentary] omnipotence have their parallel in two conceptions of an omnipotent God: on the one hand, a God who at every moment of his existence enjoys the same powers and so is incapable of cutting down those powers, and, on the other, a God whose powers include the power to destroy for the future his omnipotence.").

(41.) Posner and Vermeule themselves do not appear to put much stock in the argument. They note that the Vesting Clause does not address "how the legislative power, whatever that power encompasses, is allocated over time to successive Congresses." Posner & Vermeule, supra note 5, at 1674.

(42.) Id. at 1681-82.

(43.) Cf. McGinnis & Rappaport I, supra note 4, at 395 ("[T]he distinction between constitutional and ordinary legislation is fundamental in our system, and entrenchment flouts that distinction."). For the same reason, the Article V argument does not support the conclusion that legislative entrenchment is unconstitutional.

(44.) There are other normative arguments for and against legislative entrenchment. See, e.g., Eule, supra note 4, at 387-88; Sterk, supra note 4, at 237-40.

(45.) Eule, supra note 4, at 387. See also Sterk, supra note 4, at 240-44.

(46.) Roberts & Chemerinsky, supra note 4, at 1809-12.

(47.) Id. at 1813.

(48.) Posner & Vermeule, supra note 5, at 1670.

(49.) Id.

(50.) McGinnis & Rappaport II, supra note 4, at 510. See also Posner & Vermeule, supra note 5, at 1671.

(51.) Daryl J. Levinson, Parchment and Politics: The Positive Puzzle of Constitutional Commitment, 124 Harv. L. Rev. 657, 697 n.128 (2011). See also Daryl J. Levinson & Benjamin I. Sachs, Political Entrenchment and Public Law, 125 yale L.J. 400, 407-08 (2015).

(52.) Posner & Vermeule, supra note 5, at 1671, 1695, 1705.

(53.) Indian Appropriations Act of 1871, ch. 120, [section]1,16 Stat. 544,566.

(54.) See FEUX S. cohen's Handbook on Federal indian LAW 74-75 (Newton et al. eds., 2005).

(55.) Id. at 75.

(56.) National Defense Authorization Act for Fiscal Year 1991,10 U.S.C. [section][section] 2901-26 (1990).

(57.) Edwin R. Render, The Privatization of a Military Installation: A Misapplication of the Base Closure and Realignment Act, 44 naval L. Rev. 245, 245, 250-51 (1997).

(58.) 10 U.S.C [section][section] 2903(c)-(d), 2914(a)-(d).

(59.) Id. [section][section] 2903(e), 2904, 2908, 2914(e).

(60.) Id. [section][section] 2904, 2908, 2914(e).

(61.) See, e.g., Eule, supra note 4, at 410.

(62.) John C. Roberts, Majority Voting in Congress: Further Notes on the Constitutionality of the Senate Cloture Rule, 20 J.L. & POL. 505, 514-18 (2004).

(63.) Ed O'Keefe & Sean Sullivan, Senate Republicans Go "Nuclear, " Pave the Way for Gorsuch Confirmation to Supreme Court, wash. post, Apr. 7,2017; Jeremy w. Peters, In Landmark Vote, Senate Limits Use of the Filibuster, N. Y. TIMES, Nov. 22, 2013.

(64.) Balanced Budget and Emergency Deficit Control Act of 1985, Pub. L. No. 99-177,99 Stat. 1037 (1985). The U.S. Supreme Court declared the original version of the Gramm-Rudman-Hollings Act unconstitutional for reasons unrelated to legislative entrenchment. Bowsher v. Synar, 478 U.S. 714, 736 (1986). Congress re-enacted a modified version of the statute through the Balanced Budget and Emergency Deficit Control Reaffirmation Act of 1987. Pub. L. No. 100-119,101 Stat. 754 (1987).

(65.) Megan Suzanne Lynch, Cong. Research Serv., R41965 Statutory Budget Controls in Effect Between 1985 and 2002, at 1-8 (2011).

(66.) Omnibus Budget Reconciliation Act of 1990, Pub. L. No. 101-508, 104 Stat. 1388 (1990). See LYNCH, supra note 65, at 8-12.

(67.) Budget Control Act of 2011, Pub. L. No. 112-25, [section]251 A, 125 Stat. 240, 256 (2011).

(68.) Bipartisan Budget Act of 2013, Pub. L. No. 113-67,127 Stat. 1165 (2013).

(69.) Bipartisan Budget Act of 2015, Pub. L. No. 114-74,129 Stat. 584 (2015).

(70.) Bipartisan Budget Act of 2018, Pub. L. No. 115-123,132 Stat. 64 (2018).

(71.) Although he does not distinguish between hard entrenchment and soft entrenchment, the examples discussed by Amandeep Grewal involve soft entrenchment (or no entrenchment at all). See generally Amandeep S. Grewal, Legislative Entrenchment Rules in the Tax Law, 62 admin. L. Rev. 1011 (2010). By contrast, Jonathan Choi mistakenly reasons that soft-entrenchment devices (such as supermajority voting requirements) are binding once a chamber has adopted its legislative rules for a particular Congress. Jonathan H. Choi, Tax Commitment Devices, 15 J. Bus. & SEC. L. 1,20-21 (2014).

(72.) Daryl Levinson and Benjamin Sachs analyze a different--and broader--type of soft entrenchment ("functional entrenchment," to use their term) effected through political mechanisms. See Levinson & Sachs, supra note 51, 426-56. They briefly refer to entrenchment through the structures of legislative organization. Id. at 474, 479.

(73.) U.S. Const, art I, [section] 5, cl. 3.

(74.) Id. [section] 5, cl. 2.

(75.) See generally CHRISTOPHER J. DEERING & STEVEN S. SMITH, COMMITTEES IN CONGRESS (3d ed. 1997).

(76.) See generally Lawrence D. Longley & Walter J. Oleszek, Bicameral politics: Conference Committees in Congress (1989).

(77.) Leadership sometimes brings a measure directly to the floor without committee consideration. Barbara Sinclair, Unorthodox Lawmaking: New Legislative Processes in the U.S. Congress 17-20 (3d ed. 2007).

(78.) These blocking powers are not absolute. The committee could overrule the chair's efforts to keep the bill off the committee's agenda, and the Rules Committee could extract the bill from the committee. But those mechanisms are not commonly used.

(79.) A majority of the full House can discharge a bill from the Rules Committee and bring it directly to the floor. But successful use of the discharge petition is extremely rare. Also, certain bills do not require a rule from the Rules Committee. But those bills, typically brought up under the unanimousconsent or suspension-of-the-rules procedure, are almost always non-controversial measures that command unanimous or supermajority support.

(80.) One important exception is that, under the Congressional Budget and Impoundment Control Act of 1974, a budget-reconciliation bill cannot be filibustered.

(81.) Fisk & Chemerinsky, supra note 4, at 221.

(82.) But see id. at 203.

(83.) Gregory Koger, Filibustering: A Political History of Obstruction in the House and Senate 117 (2010).

(84.) Id. at 120-21.

(85.) Id. at 121.

(86.) id. at 122-24.

(87.) See generally DEERING & SMITH, supra note 75.

(88.) Richard F. Fenno Jr., Congressmen in Committees 147 (1973).

(89.) Deering & Smith, supra note 75, at 33-39.

(90.) Id. at 47-52.

(91.) Id. at 10.

(92.) Id. at 6-10.

(93.) house rules, h.r. doc. NO. 114-192, X.1, X.11, XII.2 at 441-91, 539-54, 623-29 (2017); Senate Rules, S. Doc. No. 113-18, XVII.1-XVII.3, XXV.I at 12,19 (2013).

(94.) Deering & Smith, supra note 75, at 6-10.

(95.) House or Senate leaders sometimes bring a measure directly to the floor without committee consideration. SINCLAIR, supra note 77, at 17-20. Leaders occasionally change the substantive content of a measure after it has been reported out of committee. Id. at 20-23. And, in the House, the Rules Committee may extract a measure from another committee, or a floor majority may discharge a measure from committee by petition. DEERING & SMITH, supra note 75,7-8.

(96.) Although they do not cast the point in terms of entrenchment, McGinnis and Rappaport recognize the capacity of the committee system to frustrate chamber majorities. See McGinnis & Rappaport II, supra note 4, at 497-99.

(97.) Murray J. Horn & Kenneth A. Shepsle, Commentary on "Administrative Arrangements and the Political Control of Agencies": Administrative Process and Organizational Form as Legislative Response to Agency Costs, 75 VA. L. REV. 499,503 (1989); Steven S. Smith, An Essay on Sequence, Position, Goals, and Committee Power, 13 LeGIS. Stud. Q. 151, 157 (1988). See also Dana & Koniak, supra note 4, at 482-83.

(98.) That said, positive agenda control is more entrenching in the House, where non-germane amendments are not permitted on the floor, than in the Senate, where such amendments are permitted.

(99.) Stanley Bach & Steven S. Smith, Managing Uncertainty in the House of Representatives 131-35 (1988); Julie Rovner, Long-Term Care Bill Derailed - For Now, cong. q. wkly., June 11, 1988, at 1604 [hereinafter Rovner I],

(100.) Bach & Smith, supra note 99, at 131; Julie Rovner, "Pepper Bill" Pits Politics Against Process, cong. Q. Wkly., June 4, 1988, at 1491 [hereinafter Rovner II],

(101.) Bach & Smith, supra note 99, at 131; Rovner II, supra note 100.

(102.) Bach & Smith, supra note 99, at 131-32; Rovner II, supra note 100.

(103.) Bach & Smith, supra note 99, at 132-34; Rovner II, supra note 100.

(104.) Rovner I, supra note 99.

(105.) Fisk & Chemerinsky, supra note 4, at 219 n.210.

(106.) William N. Eskridge, Jr. et al., Legislation and Statutory Interpretation 109-10 (2d ed. 2006).

(107.) Id. at 110-11; White House Looks for 'Magic Formula' in House-Senate Conference on Budget, BNA daily tax rep., June 18, 1993, at G-4-G-5.

(108.) Moderates Tell Mitchell to Kill Energy Tax; Revenue Offsets Remain Open Issue, BNA Daily Tax Rep., July 20, 1993, at G-3-G-4.

(109.) Senate Expected to Begin Consideration of Deficit-Reduction Package June 23, bna daily tax Rep., June 21, 1993, at g-7-g-10.

(110.) See eskridge, jr. et al., supra note 106, at 112.

(111.) Congressional Budget and Impoundment Control Act of 1974, Pub. L. No. 93-344, [section] 300,88 Stat. 297, 306 (1974).

(112.) Id. [section][section] 300-01; james v. saturno et al., cong. research serv., R42388, the Congressional Appropriations Process: An Introduction 3 (2016).

(113.) Congressional Budget and Impoundment Control Act of 1974, Pub. L. No. 93-344, [section][section] 302(f), 311, 312, 88 Stat. 297 (1974); SATURNO ET AL., supra note 112, at 3-4.

(114.) Congressional Budget and Impoundment Control Act of 1974, Pub. L. No. 93-344, [section] 300,88 Stat. 297, 306 (1974); Saturno et al., supra note 112, at 2-3, 12; James V. Saturno & jessica Tollestrup, Cong. Research Serv., R42647, Continuing Resolutions: Overview of Components and Recent Practices 1 (2016).

(115.) See generally James V. Saturno & Brian T. Yeh, Cong. Research Serv., R42098, Authorizations and Appropriations: Procedural and Legal Issues (2016).

(116.) Congressional Budget and Impoundment Control Act of 1974, Pub. L. No. 93-344, [section] 300,88 Stat. 297, 306 (1974).

(117.) The budget resolution cannot be filibustered in the Senate. Id. [section] 305(b).

(118.) Saturno et al., supra note 112, at 4. One or both chambers typically mitigate the failure to pass a budget resolution by deeming other legislation to serve the function of a budget resolution. See generally megan s. lynch, cong. research serv., R44296, deeming resolutions: Budget enforcement in the Absence of a Budget Resolution (2015).

(119.) SATURNO, ET AL., supra note 112, at 13; Saturno & Tollestrup, supra note 114, at 1.

(120.) Clinton t. Brass et al., Cong. Research Serv., RL34680, Shutdown of the Federal Government: Causes, Processes, and effects 3 (2017).

(121.) See generally id.

(122.) Saturno et al., supra note 112, at 13-14; saturno & Tollestrup, supra note 114, at 3-7.

(123.) But see SATURNO & TOLLESTRUP, supra note 114, at 8-9.

(124.) Id. at 19-30.

(125.) In total, more than half of annual federal spending is outside the regular budget process. Cra. on Budget and Policy Priorities, Policy Basics: Introduction to the Federal Budget Process 3 (2017); Saturno et al., supra note 112.

(126.) The 2017 Annual Report of the Board of Trustees of the Federal Old-Age and Survivors Insurance and Federal Disability Insurance Trust Funds 7 (2017).

(127.) Id. at 7-8. For 2016, there was a reimbursement of just under $100 million made to Social Security from general federal revenues. Id.

(128.) 2017 Annual Report of the Boards of trustees of the Federal Hospital Insurance and Federal Supplementary Medical Insurance Trust Funds 10.

(129.) Id. at 10-11.

(130.) Certain tax expenditures have statutory expiration (or "sunset") dates, and renewal of those tax expenditures requires affirmative legislative action. See generally Joint Comm. of Taxation, List of Expiring Federal Tax Provisions 2016-2026 (2017). For a broader analysis, see generally George K. Yin, Temporary-Effect Legislation, Political Accountability, and Fiscal Restraint, 84 N.Y.U. L. Rev. 174 (2009).

(131.) Alan Cole, Corporate and Individual Tax Expenditures 5 (Melodie Bowler ed., The Tax Found. 2015).

(132.) U.S. Treas. Dep't Off. of Tax Analysis, Tax Expenditures 1-2 (2015).

(133.) joint Comm. of Taxation, Estimates of Federal Tax Expenditures for Fiscal Years 2016-2020, at 28,32,33,37 (2017).

(134.) Social Security Amendments of 1983, Pub. L. No. 98-21, 97 Stat. 65 (1983).

(135.) Medicare Prescription Drug, Improvement, and Modernization Act of 2003, Pub. L. No. 108173,117 Stat. 2066 (2003).

(136.) An Act to 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).

(137.) Tax Reform Act of 1986, Pub. L. No. 99-514,100 Stat. 2085 (1986).

(138.) See also Peter C. Ordeshook, Game Theory and Political Theory: An Introduction 56-65 (1986). See generally Kenneth J. Arrow, Social Choice and Individual values (1951).

(139.) Ordeshook, supra note 138, at 62-65.

(140.) William H. Riker, Liberalism Against Populism: a Confrontation Between the Theory of Democracy and the Theory of Social Choice 214-53 (1982) [hereinafter Riker I].

(141.) Charles R. Plott, A Notion of Equilibrium and Its Possibility under Majority Rule, 57 Am. econ. Rev. 787 (1967).

(142.) William H. Riker, Implications from the Disequilibrium of Majority Rule for the Study of Institutions, 74 AM. pol. SCI. Rev. 432, 441 (1980) [hereinafter Riker II]; Kenneth A. Shepsle & Barry A. Weingast, Structure-Induced Equilibrium and Legislative Choice, 37 pub. choice 503, 504 (1981) [hereinafter Shepsle & Weingast I].

(143.) Richard D. McKelvey, Intransitivities in Multidimensional Voting Models and Some Implications for Agenda Control, 12 J. econ. theory 472, 481 (1976). See also Richard D. McKelvey, General Conditions for Global Intransitivities in Formal Voting Models, 47 econometrica 1085 (1979); Norman Schofield, Instability of Simple Dynamic Games, 45 rev. econ. STUD. 575 (1978).

(144.) Kenneth A. Shepsle & Barry R. Weingast, Uncovered Sets and Sophisticated Voting Outcomes with Implications for Agenda Institutions, 28 Am. J. pol. SCI. 49,58-68 (1984). See also Morris P. Fiorina & Kenneth A. Shepsle, Formal Theories of Leadership: Agents, Agenda Setters, and Entrepreneurs, in Leadership and Politics: New Perspectives in Political Science 30 (Bryan D. Jones ed., 1989).

(145.) Shepsle & Weingast I, supra note 142, at 504-07. See also Kenneth A. Shepsle & Barry R. Weingast, Institutionalizing Majority Rule: A Social Choice Theory with Policy Implications, 72 Am. econ. Rev. 367,368 (1982) [hereinafter Shepsle & Weingast II].

(146.) See William H. Panning, Formal Models of Legislative Processes, 8 legis. stud. Q. 427, 438 (1983).

(147.) Kenneth A. Shepsle, Institutional Arrangements and Equilibrium in Multidimensional Voting Models, 23 Am. J. Pol. SCI. 27,27 (1979).

(148.) Id. at 47-55. A "structure-induced equilibrium" is a point in multidimensional policy space with respect to which the only other points that could defeat that point "either can be proposed only by those who do not prefer to do so, or cannot be proposed at all." Shepsle & Weingast I, supra note 142, at 512. The committee system and restrictive-amendment procedures are not the only institutional arrangements that can produce structure-induced equilibria. Id. at 513-14.

(149.) Shepsle, supra note 147, at 47-55. See also Riker II, supra note 142, at 443-44; Shepsle & Weingast I, supra note 142, at 507-11; Shepsle & Weingast II, supra note 145, at 368.

(150.) Eric Shickler, Disjointed Pluralism: Institutional Innovation and the Development of the U.S. Congress 5-12 (2001); Gary W. Cox & Matthew D. McCubbins, Theories of Legislative Organization, 15 APSA-CP newsl. 1, 9 (2004); Kenneth A. Shepsle & Barry R. Weingast, Positive Theories of Congressional Institutions, in positive theories of congressional institutions 10-21 (Kenneth A. Shepsle & Barry R. Weingast eds., 1995) [hereinafter Shepsle & Weingast III]. Some object that Shepsle's thesis simply pushes the problem of instability back one stage. Riker posed the objection as one of inherited instability. Riker II, supra note 142, at 444-45. If legislative organizational structures, rather than policy preferences, induce equilibrium outcomes in a majoritarian institution, Riker argued, the heterogeneity of policy preferences ought to create instability when the institution establishes its organizational structures. Id. Levinson puts Riker's point more bluntly: "Without some explanation of what stabilizes the supposedly stabilizing institutional structures, structure-induced equilibrium is a deus ex machina." Levinson, supra note 51, at 682. But this problem is easily overstated. The underlying conundrum is the possibility of disequilibrium when a majoritarian institution acts in multidimensional policy space. Arrow's Theorem does not predict cycling for decisionmaking in unidimensional policy space when legislators have single-peaked preferences; under those conditions, a Condorcet winner regularly emerges. riker I, supra note 140, at 126; Riker II, supra note 142, at 436-37. Those conditions seem the better fit for modeling the choice about whether to adopt a particular set of legislative organizational structures and, subsequently, about whether to retain that set of structures. Most legislators, even those in the minority, have more to lose from legislative chaos than from an organizational status quo. At a minimum, cycling and chaos jeopardize each legislator's ability to participate in the pork-barrel politics that protect incumbency. Thus, when confronted with the straightforward choice between organizational structure and cycling, most legislators should have a clear preference for organizational structure. Here, the policy space is unidimensional, and structure is a Condorcet winner. This helps to explain the inter-temporal stability of legislative organizational structures, which even Riker conceded. Id. at 445.

(151.) Cf. David P. Baron, Legislative Organization with Informational Committees, 44 Am. J. pol. sci. 485, 502 (2000) (explaining a theory of legislative organization incentivizing specialization by committees). See generally keith krehbiel, information and legislative organization (John E. Jackson & Christopher H. Achen eds., Univ. of Mich. Press 1992).

(152.) Krehbiel, supra note 151, at 20.

(153.) Krehbiel defines "informational efficiency" as "the reduction of uncertainty in the course of the choice process." Id. at 74.

(154.) Id. at 64.

(155.) Id. at 68-69. See also H.R. REP. No. 93-916, pt. II (1974).

(156.) KREHBIEL, supra note 151, at 66-81.

(157.) Id. at 69.

(158.) Id. at 90-92, 109 n.1. See also Daniel Diermieier, Commitment, Deference, and Legislative Institutions, 89 AM. POL. SCI. Rev. 344 (1995).

(159.) KREHBIEL, supra note 151, at 76.

(160.) Shepsle & Weingast III, supra note 150, at 7-8. See also C. Lawrence Evans, Legislative Structure: Rules, Precedents, and Jurisdictions, 24 LEG. STUD. Q. 605, 608-09 (1999); Barry R. Weingast & William J. Marshall, The Industrial Organization of Congress; or, Why Legislatures, Like Firms, Are Not Organized as Markets, 96 J. POL. ECON. 132, 133 (1988).

(161.) Weingast & Marshall, supra note 160, at 133,136-37.

(162.) Id. at 135, 138-42; Shepsle & Weingast III, supra note 150, at 11.

(163.) Shepsle & Weingast iii, supra note 150, at 12-13; Weingast & Marshall, supra note 160, at 142-43. See generally Barry R. Weingast, A Rational Choice Perspective on Congressional Norms, 23 Am. J. POL. SCI. 245 (1979).

(164.) Weingast & Marshall, supra note 160, at 143-48.

(165.) Id.

(166.) Shepsle & Weingast iii, supra note 150, at 11-12 (quoting Morris p. Fiorina, Alternative Rationales for Restrictive Procedures, 3 J.l. econ. & Org. 337, 338 (1987)). See also John Ferejohn, Logrolling in an Institutional Context: A Case Study of Food Stamp Legislation, in congress and policy change 223, 224-25 (Gerald c. Wright, Jr. et al. eds., 1986).

(167.) David P. Baron & John Ferejohn, The Power to Propose, in Models of Strategic Choice in Politics 343, at 353-65 (Peter C. Ordeshook ed., 1989). See also David P. Baron, A Noncooperative Theory of Legislative Coalitions, 33 Am. J. POL. Sci. 1048,1081-82 (1989).
COPYRIGHT 2018 Duke University, School of Law
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2018 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:The Past, Present, and Future of the Federal Tax Legislative Process
Author:Doran, Michael
Publication:Law and Contemporary Problems
Date:Mar 22, 2018
Words:16574
Previous Article:THE STATE OF AMERICA'S TAX INSTITUTIONS.
Next Article:TAX LAW AND THE ERODING BUDGET PROCESS.
Topics:

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |