Reconciliation lets a Senate majority legislate with fifty-one votes, and one section of the Congressional Budget Act decides what that majority may carry. That section is the Byrd rule, a definition of extraneous matter joined to a point of order any senator may raise. A bill traveling under reconciliation receives twenty hours of debate, protection against extended delay, and passage by simple majority. The Byrd rule prices those advantages. A provision that earns a place in the bill through a genuine change in outlays or revenues keeps the expedited timetable, and a provision that borrows the timetable for a purpose the budget cannot measure loses the timetable and faces removal.

The rule in force on March 15, 2020 was section 313 of the Congressional Budget and Impoundment Control Act of 1974, codified at 2 U.S.C. 644, added in temporary form by the Consolidated Omnibus Budget Reconciliation Act of 1985 and made permanent by the Budget Enforcement Act of 1990 as Title XIII of Public Law 101-508. Senator Robert C. Byrd of West Virginia proposed the rule in October 1985 after counting 122 extraneous items in a pending reconciliation bill. The rule applies in the Senate alone, applies only to measures considered under the reconciliation authority of section 310, and operates as a point of order rather than as a prohibition, so extraneous matter may be drafted and reported and reaches removal only when a senator invokes the rule and the presiding officer sustains the objection.

Six tests supply the definition, and each test polices a distinct failure of fit between a provision and the vehicle. A provision is extraneous when the provision produces no change in outlays or revenues, when the provision increases outlays or decreases revenues inside a title whose reporting committee failed to achieve an instruction, when the provision lies outside the jurisdiction of the committee that submitted the provision, when the budgetary effects of the provision are merely incidental to the nonbudgetary components of the provision, when the provision increases the deficit in a fiscal year beyond the years the reconciliation measure covers without an offset in the same title and the same year, or when the provision recommends changes in the Social Security program established under Title II of the Social Security Act. Most of the tests end in a document, an estimate, an instruction, a jurisdictional assignment, or a subject matter identification, and the distribution of difficulty across the six explains the pattern of disputes the rule generates.

One undefined phrase governs American lawmaking. The merely incidental test carries the contested disputes because Congress enacted the phrase without a definition, a ratio, a dollar threshold, or a list of factors, and the accumulated advice of the Senate parliamentarian supplies the reasoning case by case. An effect measured in the hundreds of billions of dollars failed the test in 2015 because a nationwide rule of conduct dwarfed the dollars, and a leasing program passed the test in 2017 because lease sales and royalty shares made revenue the operative mechanism of the provision. The fifth test works quietly, instructing drafters years in advance to print expiration dates into tax law, so that the individual provisions of the tax act of 2017 terminate after December 31, 2025 while the corporate rate reduction, carried by permanent offsets in the same title, carries no termination date at all.

Enforcement completes the design. A point of order is not debatable, a motion to waive requires the affirmative vote of three fifths of the senators duly chosen and sworn, and an appeal of a ruling carries the same threshold, so the escape valve asks a majority to assemble for one provision the supermajority whose absence justified the expedited vehicle. Before any floor stage, committee staff argue provision by provision before the parliamentarian in the pre-floor review called the Byrd bath, and most provisions that fail die in that room through removal or redrafting. The sections that follow take the tests in statutory order, trace the sunset architecture the fifth test built into three tax statutes, and examine the adviser whose formally nonbinding judgment settled the content of reconciliation bills across changes in party control.

Why the Senate Built a Fence Around Reconciliation

The Byrd rule exists because a procedure built for arithmetic acquired a second use as a vehicle. Reconciliation was designed to bring revenue and mandatory spending law into line with numbers Congress had already adopted in a budget resolution. That design gave reconciliation bills a powerful advantage in the Senate, where ordinary legislation faced extended debate and where a determined minority could prevent a final vote. Once committees and majorities learned that a provision placed inside a reconciliation bill traveled under limited debate, the temptation to load the bill with matter that had little connection to the budget became structural rather than occasional. The Byrd rule, named for Senator Robert C. Byrd of West Virginia, answered that temptation with a definition and a remedy. The definition identified matter Congress labeled extraneous. The remedy armed any senator with a point of order against such matter during Senate consideration of a reconciliation bill or a conference report on such a bill. The rule marked a boundary around a privileged procedure and supplied a parliamentary means of enforcing that boundary when a senator chose to invoke it.

The statute behind the procedure was the Congressional Budget and Impoundment Control Act of 1974, enacted as Public Law 93-344. The act created the House and Senate Budget Committees, established the Congressional Budget Office, and provided for budget resolutions through which Congress set aggregate levels for revenues, spending, deficits, and debt. Within that architecture, the reconciliation procedure the act created occupied a deliberately subordinate place. Section 310 of the act authorized a budget resolution to carry reconciliation instructions directing specified committees to report legislation changing laws within their jurisdiction so that revenue and mandatory spending levels conformed to the resolution. Reconciliation therefore began as an implementing device. The budget resolution stated the fiscal policy. Reconciliation changed permanent law to execute that policy. A provision belonged in a reconciliation bill because of the budgetary change the provision made, not because the provision commanded majority support or had failed to advance through ordinary channels.

The original conception was also modest in timing and scale. Under the timetable contemplated in 1974, Congress adopted a budget resolution in the spring, conducted most legislative business during the summer, and then considered a second resolution in the autumn that confirmed or revised the first in light of economic conditions and enacted legislation. Reconciliation belonged near the end of that sequence, as a final alignment exercise that adjusted existing law to decisions already taken. The Congressional Research Service, in the report on the Byrd rule by Bill Heniff Jr., described reconciliation as a process for changing substantive law so that revenue and mandatory spending levels came into line with budget resolution policies, and that description captures the limited ambition of the original device. Nothing in section 310 required Congress to use reconciliation in any session, and six years passed between enactment in 1974 and first use in 1980.

The mechanics of the two-step process explain both the usefulness and the vulnerability of the device. In the first step, a budget resolution instructed one or more committees in each chamber to develop legislation producing specified changes in spending, revenues, or both. An instruction typically identified the committee, the fiscal years covered, and the amount by which the committee had to change outlays or revenues. In the second step, each instructed committee reported recommendations, and the Budget Committee in each chamber assembled the submissions into a single omnibus reconciliation bill. Heniff explained that the Budget Committee incorporated committee submissions without substantive revision, a mechanical feature with large consequences. The Budget Committee acted as an assembler rather than as an editor with authority to rewrite the policy choices of the legislative committees. If an instructed committee reported language that exceeded its instruction, strayed into another committee jurisdiction, or produced no budgetary change at all, that language entered the omnibus bill unless the Senate possessed a procedure for removing it.

Senate consideration supplied the protection that made inclusion so valuable. Under the expedited procedures associated with sections 305 and 310 of the Congressional Budget Act, debate on a reconciliation bill in the Senate was limited to twenty hours, amendments had to be germane, and related motions and appeals operated under compressed time limits. A reconciliation bill therefore could not be filibustered in the manner available against ordinary Senate legislation. A simple majority could bring the bill to passage after the allotted debate expired. In the Senate, where extended debate formed part of the institutional identity Byrd defended throughout a long career, expedition altered the strategic calculation for every provision that could plausibly be attached to a reconciliation vehicle. Placement determined whether a provision needed the sixty votes associated with cloture on ordinary legislation or only the majority needed to pass the omnibus bill.

What problem was the Byrd rule written to solve?

Committees were loading reconciliation bills with provisions that had little or no budgetary purpose, because limited debate made the bills easier to pass than ordinary legislation. The Byrd rule was written to confine reconciliation to genuine budget implementation and to return unrelated policy to regular Senate procedure.

The first reconciliation measures showed the procedure working close to its stated purpose. The Omnibus Reconciliation Act of 1980, Public Law 96-499, signed on December 5, 1980, was the first reconciliation bill to pass the House and the Senate, according to the Congressional Research Service catalogue of reconciliation measures enacted into law. The act reduced the projected deficit for fiscal year 1981 through a combination of outlay reductions and revenue increases spread across subjects that included child nutrition, student loans, federal retirement adjustments, Medicare and Medicaid, unemployment compensation, and revenue provisions affecting items such as mortgage subsidy bonds and telephone excise taxes. Whatever disagreements attended individual components, the components changed spending or revenue law in the direction the budget resolution had instructed, and enactment moved actual law toward the fiscal aggregates the resolution had adopted earlier in the session.

The Omnibus Budget Reconciliation Act of 1981, Public Law 97-35, signed on August 13, 1981, expanded the scale of the experiment without breaking its fiscal logic. President Ronald Reagan used the act, together with separate tax legislation enacted outside reconciliation, to advance a large part of a first year program built around spending restraint. The Congressional Research Service estimated a deficit reduction of 130.6 billion dollars across fiscal years 1982 to 1984, drawn from changes affecting health program block grants, Medicaid, food stamps, dairy price supports, energy assistance, education programs, and student loans. The 1981 act demonstrated that reconciliation could carry a governing program of considerable breadth when that program consisted principally of changes to spending law. Breadth alone therefore did not signal abuse. The distinction that later required a rule lay between provisions that implemented budget resolution policies through changes in outlays or revenues and provisions that used the vehicle while making no such contribution.

Experience between 1980 and 1985 eroded that distinction in the submissions committees sent to the Budget Committees. Heniff summarized the pattern in terms that the statute later adopted almost point for point. Reconciliation submissions included provisions that had no budgetary effect. Submissions included provisions that increased spending or reduced revenues when the governing instructions called for reduced spending or increased revenues, so that a committee could claim the protection of a deficit reduction bill for language that moved the deficit in the opposite direction. Submissions also included provisions that violated another committee jurisdiction, allowing one committee to legislate on subjects assigned to a rival panel under the cover of an omnibus measure that the Budget Committee assembled without substantive revision. Each category represented a different failure of fit between a provision and the purpose of the vehicle. A provision without a budgetary effect failed the basic test of implementation. A provision moving money in the wrong direction failed the instruction that justified expedited treatment. A provision outside the reporting committee jurisdiction failed the allocation of work on which the instruction system depended.

Scale converted those failures from irritations into an institutional problem. As reconciliation became a regular part of the congressional session during the first half of the 1980s, the omnibus bill became one of the few measures certain to reach the Senate floor under a fixed timetable and certain to receive a vote. A provision inserted into reconciliation bypassed the obstacles of regular order at once, needing no separate scheduling decision and no separate cloture vote. Before 1986, Senate practice dealt with the problem through negotiation rather than through law. Heniff and the Senate procedural tradition recorded in Riddick’s Senate Procedure described leadership amendments, negotiated on a bipartisan basis and offered during floor consideration, that struck provisions regarded as extraneous. A negotiated cleanup depended on agreement. A rule promised a standard that a single senator could invoke without first assembling a leadership coalition.

A gap in existing parliamentary law made the negotiated approach necessary. Points of order under the Congressional Budget Act reached many defects in a reconciliation bill, and section 310 itself barred certain amendments and barred reconciliation legislation recommending changes in Social Security. Germaneness requirements constrained amendments offered on the floor. Yet no point of order lay against extraneous language already contained in the bill as reported, as distinct from language a senator attempted to add by amendment. A committee could therefore accomplish through an original submission what a senator acting alone on the floor could not accomplish through an amendment, even though both routes placed the same unrelated policy inside the same protected vehicle. The Byrd rule closed that asymmetry by defining the defect in the underlying text and by attaching a point of order to the text itself.

Senator Byrd brought the objection to the floor on October 24, 1985, during Senate consideration of the Consolidated Omnibus Budget Reconciliation Act of 1985. Byrd offered Amendment No. 878, on behalf of himself and other senators, to S. 1730, and the Senate adopted the amendment by a vote of 96 to 0, as Heniff recorded. Byrd framed the grievance in two connected parts. The first part concerned effectiveness. Extraneous provisions provoked controversy without aiding deficit reduction, consumed the limited debate time reconciliation allotted, and endangered the fiscal agreements the bill existed to implement. The second part concerned the character of the Senate. Matters with no central connection to deficit reduction deserved debate under regular procedures, where extended consideration, open amendment, and the possibility of a filibuster formed part of the legislative test a controversial proposal had to pass. Heniff recorded that Byrd counted 122 items in the pending reconciliation bill as extraneous, a figure that gave the abstract complaint a concrete measure.

The remedy Byrd proposed followed the diagnosis closely. Rather than prohibit reconciliation bills above a certain size, require advance certification of every provision by a central authority, or assign the Budget Committee a substantive veto over committee submissions, the amendment defined categories of extraneous matter and created a point of order that any senator could raise against matter falling within those categories. If the presiding officer sustained the point of order, the offending part of a title or provision was deemed stricken from the bill and could not be offered again as a floor amendment. The design preserved majority control over genuine budget policy while withdrawing expedited protection from matter that failed the definition. A majority determined to enact stricken policy retained every option available under regular order. The rule therefore regulated the route rather than the destination. Content that belonged in reconciliation kept the twenty hour timetable. Content that failed the statutory definition lost the timetable.

Enactment gave that design a temporary and divided legal form that took five years to settle into permanent shape. The Byrd amendment entered law, in modified form, as section 20001 of the Consolidated Omnibus Budget Reconciliation Act of 1985, Public Law 99-272. Although the parent act bore a 1985 title, enactment arrived on April 7, 1986, when the measure was signed into law, and section 20001 took effect on that date. Title XX of the act carried the provision, and subsection (c) of section 20001 set an expiration date of January 2, 1987. The statutory component governed initial Senate consideration of reconciliation bills and resolutions. A second component operated through Senate action alone. On December 19, 1985, the Senate adopted S.Res. 286, sponsored by Senator Alan Simpson and other senators, extending portions of the rule to conference reports and amendments between the two chambers, the stages at which text negotiated after Senate passage could reintroduce matter the Senate had never approved under the definition. Because enactment of the 1985 act had slipped into April 1986, the resolution component took effect before the statutory component. On October 16, 1986, the Senate adopted S.Res. 509, which clarified that the resolution component also reached an amendment between the chambers.

Extensions followed in two steps before permanence replaced the cycle. Section 7006 of the Omnibus Budget Reconciliation Act of 1986, Public Law 99-509, enacted on October 21, 1986, made several minor changes in the rule and extended the expiration date by one year, to January 2, 1988. Section 205 of Public Law 100-119, a measure increasing the statutory limit on the public debt and revising budget process rules, signed on September 29, 1987, added an item to the list of definitions of extraneous matter and extended the expiration date to September 30, 1992. Between April 1986 and November 1990, the Byrd rule therefore operated as an experiment under a deadline, resting in part on Senate resolutions of the 99th Congress whose reach covered conference stages that the statute alone had not addressed. Consolidation required a single enactment that merged the components and removed the expiration date.

The Budget Enforcement Act of 1990 supplied that enactment. The Budget Enforcement Act formed Title XIII of the Omnibus Budget Reconciliation Act of 1990, Public Law 101-508, enacted on November 5, 1990. Section 13214 of that act made significant revisions to the Byrd rule and incorporated the rule into the Congressional Budget Act of 1974 as section 313, in permanent form. The mechanism deserves emphasis because descriptions that date section 313 to 1985 compress two distinct legal events into one. A rule against extraneous matter existed in law from April 7, 1986, under section 20001 of the 1985 reconciliation act and under the accompanying Senate resolutions. Section 313 as a numbered part of the Congressional Budget Act dated from November 5, 1990, when section 13214 merged the statutory and resolution components and placed the unified text inside the act that governed the budget process itself. Seven years after permanence, the Budget Enforcement Act of 1997, enacted as Title X of the Balanced Budget Act of 1997, Public Law 105-33, made minor technical changes to section 313 through section 10113(b)(1) to correct drafting problems in the 1990 act. No further change to the rule followed between the 1997 corrections and March 15, 2020, according to the legislative history Heniff compiled.

The formal identity of the rule in force on March 15, 2020 can therefore be stated with exactness. The Byrd rule was section 313 of the Congressional Budget and Impoundment Control Act of 1974, the formal name of the statute commonly called the Congressional Budget Act of 1974, and section 313 was codified at 2 U.S.C. 644. A complete orientation to the parent statute appears in the Congressional Budget Act pillar guide, which sets out the budget resolution, committee, and enforcement structure into which section 313 was inserted in 1990. The section bore the heading Extraneous Provisions in Reconciliation Legislation. The rule applied only in the Senate. No equivalent provision governed House consideration of reconciliation bills, where the chamber own rules on germaneness and special orders from the Rules Committee performed different work under different institutional conditions. The rule also applied only to reconciliation measures considered under section 310, not to appropriations bills, authorization bills, or other legislation that happened to affect the deficit.

Equally decisive is the distinction between a prohibition and a point of order. Section 313 did not declare extraneous provisions void, did not forbid committees from reporting such provisions, and did not direct any officer to remove such provisions automatically before floor consideration. Extraneous matter could be drafted, reported, assembled into an omnibus bill by a Budget Committee, and printed in the text senators received. Section 313 instead made such matter subject to a point of order in the Senate. Enforcement required action. A senator had to identify the matter, raise the point of order at the proper stage, and obtain a ruling sustaining the objection, after which the matter was stricken or the amendment carrying the matter fell. The Senate could also waive the point of order, and the statute set the waiver threshold at three fifths of senators duly chosen and sworn, a requirement later sections examine in detail. A provision that met a definition of extraneous matter therefore faced a procedural vulnerability rather than an absolute legal bar, and the practical fate of such a provision depended on whether an opponent invoked the rule and whether sixty senators voted to protect the provision when invocation occurred.

Section 313 defined extraneous matter through six tests, and the architecture of those tests maps onto the history traced in this section. One test asked whether a provision produced any change in outlays or revenues at all. Another asked whether a provision moved outlays or revenues against the instruction of a committee that had failed to comply with that instruction. Another asked whether the provision lay outside the jurisdiction of the committee that submitted the provision. Another asked whether the budgetary effects of a provision were merely incidental to the nonbudgetary components of the provision. Another asked whether a provision increased the deficit in a fiscal year beyond the years covered by the reconciliation measure. The final test asked whether a provision recommended changes in the Social Security retirement and disability program established under the Social Security Act of 1935. Later sections take those tests in groups, beginning with the tests that police the relationship between a provision, a committee instruction, and committee jurisdiction, and proceeding to the tests that require judgment about the character and timing of budgetary effects. Every test, whatever the grouping, enforced the single premise on which the Senate acted on October 24, 1985: expedited consideration was a privilege Congress granted for budget implementation, and a provision that could not establish a genuine budgetary claim to that privilege had to travel by the ordinary road.

The First Three Tests: Money, Instructions and Jurisdiction

Section 313(b)(1) of the Congressional Budget Act, codified at 2 U.S.C. 644, defines extraneous matter through six separate tests, and a provision that fails any one of them fails altogether. The first three tests, set out in sections 313(b)(1)(A) to (C), share a quality that separates them from the tests that follow in the statutory sequence. Each asks a question answerable from documents already on the record: a cost estimate from the Congressional Budget Office or the Joint Committee on Taxation, the instructions printed in the budget resolution, and the standing rules assigning subjects among Senate committees. None of the three requires the presiding officer to weigh the purpose of a provision against its fiscal effect or to decide whether a policy change dwarfs the money attached to it. Because the answers were usually knowable in advance, the tests did most of their work before introduction and during committee drafting, when legislative counsel and committee staff removed or rewrote vulnerable language in the pre-floor review conducted with the Senate parliamentarian, a process known in Senate usage as the Byrd bath.

What makes a provision extraneous under the first three tests?

A provision fails the first test when it changes no outlays or revenues, the second when it worsens the budget while its committee missed its instruction, and the third when it lies outside the reporting committee’s jurisdiction, so each defect exposes the provision to a point of order and striking.

The first test, in section 313(b)(1)(A), provides that a provision of a reconciliation bill or reconciliation resolution is extraneous if the provision does not produce a change in outlays or revenues. The statutory sentence carries two qualifications that define its real reach. The first qualification expands the test in the drafter’s favor, because the change in outlays or revenues may be brought about by changes in the terms and conditions under which outlays are made or revenues are required to be collected. Eligibility criteria for a benefit program, the formula by which a payment is calculated, and the definition of the persons or transactions subject to a tax can therefore carry a provision through the first test even where the operative sentence reads as a definition rather than as an appropriation or a rate, provided that the term or condition genuinely determines how money moves. Parliamentarian advice described in the Congressional Research Service explanation of the rule treated qualifying terms and conditions as matter necessary or essential to the determination of how outlays are distributed or revenues are collected, and language whose removal left an estimate untouched performed no budgetary work. The second qualification protects symmetry, because a provision in which outlay decreases or revenue increases exactly offset outlay increases or revenue decreases is not extraneous by virtue of the first test alone.

The mechanism the first test polices is decoration. Reconciliation exists to change the fiscal position recorded in the budget resolution, and a provision that changes no outlay and no revenue line contributes nothing to that purpose, however worthy the statement it makes. The categories of matter that the Congressional Research Service records as treated as extraneous under the first test describe the decorative layer of legislation with precision: sense of the Senate and sense of the Congress language, congressional findings, policy statements, tables of contents, rules of construction and short titles. Inside reconciliation, each such device consumes the privilege of expedited consideration, including the twenty hour limit on debate that protects a reconciliation bill from a filibuster, without moving a single dollar in the direction the budget resolution instructed. The first test withdraws that bargain. Section 313(b)(2) softened the test for a narrow class of Senate originated provisions, and the exception confirms the test’s logic by the conditions it imposes. A provision with no change in outlays or revenues escaped the first test if the chairman and ranking minority member of the Budget Committee and the chairman and ranking minority member of the reporting committee jointly certified that the provision was tied to a fiscal effect, as where the provision mitigated direct effects of another provision changing outlays or revenues, or would produce a substantial fiscal effect in later years, or was likely to produce an effect that could not be reliably estimated. The bipartisan signature requirement imposed an institutional cost that discouraged use of the exception as a routine shelter for decoration.

The December 2017 consideration of the conference report on H.R. 1, the tax reconciliation measure enacted as Public Law 115-97, supplied the first test’s most visible application. The conference report carried a provision, subsection 11000(a), stating the short title by which the act was to be cited, the Tax Cuts and Jobs Act. A short title changes no rate, no deduction, no credit and no payment, and it produces no change in outlays or revenues under any estimate the Joint Committee on Taxation could prepare. In the pre-floor review the Senate parliamentarian advised that the provision was extraneous under the first test, and on December 19, 2017, Senator Bernie Sanders raised points of order against the short title together with two other provisions of the conference report. A motion to waive the points of order was rejected by a vote of 51 to 48, short of the three-fifths vote the Budget Act requires for a waiver, and the points of order were sustained. The short title was struck with the other two provisions, the Senate proceeded to a substitute text without the stricken matter, and the House of Representatives, which had passed the conference report earlier on December 19, voted again on December 20, 2017 on the corrected text. The enacted statute therefore carries no short title at all, and its formal title is the long descriptive sentence that reconciliation bills bear as a matter of course, an act to provide for reconciliation pursuant to titles II and V of the concurrent resolution on the budget for fiscal year 2018. The episode reversed an expectation on which drafters had relied for more than two decades, because advice given in 1993 by an earlier parliamentarian had treated short title provisions as too trifling to fall under the rule, and the mechanical character of the test left no other answer available from the estimate. The first test also operated at smaller scale on March 25, 2010, when Senator Judd Gregg raised two points of order against two brief provisions in the education portion of H.R. 4872, the Health Care and Education Reconciliation Act of 2010, dealing with the Pell grant program. The provisions were judged to produce no changes in outlays or revenues, the points of order were sustained, and the provisions were removed before the measure returned to the House of Representatives. The same sentence of the statute thus removed program language affecting student aid and removed the name of a tax act, examining the estimate and not the importance of the words under challenge.

The second test, in section 313(b)(1)(B), provides that any provision producing an increase in outlays or a decrease in revenues is extraneous if the net effect of the provisions reported by the committee reporting the title containing the provision is that the committee fails to achieve its reconciliation instructions. Both elements must be present before a provision falls. The provision itself must move the budget in the expensive direction, raising outlays or lowering revenues, and the title in which the provision sits must, taken as a whole, miss the instruction that the budget resolution gave to the reporting committee. A committee in full compliance could include provisions that increased outlays or reduced revenues, provided that savings or revenue elsewhere in its title brought the net effect to the instructed figure. A committee that missed its instruction exposed only the expensive provisions of its title, because a provision that reduced outlays or raised revenues moved toward the instruction rather than against it and could not be extraneous under this ground however far the title as a whole had fallen short. The unit of judgment is the title and not the provision, which makes the second test a rule of collective responsibility.

The mechanism the second test polices is discipline, and the discipline runs between the Senate as a whole and its own committees. A budget resolution containing reconciliation instructions assigns a named committee a stated change in outlays or revenues, expressed for stated fiscal years, that the committee is directed to achieve by reporting changes in the laws within its jurisdiction. The second test converts the instruction from an expectation into a condition of privilege. A committee that failed its figure lost the protection of expedited procedure for the very provisions that made its title popular, since each provision that spent money or surrendered revenue could be struck on a point of order, while the savings provisions that no constituency defended would survive untouched. Committees responded to that arithmetic during drafting, which is why the Congressional Research Service record of the second test is dominated by rarity on the floor rather than by a long series of strikes.

The documented floor application of the second test that the Congressional Research Service sets out in detail arose during Senate consideration of S. 1956, the Senate companion measure to the Personal Responsibility and Work Opportunity Reconciliation Act of 1996. The conference report on the budget resolution for that cycle directed the Senate Finance Committee to report changes in laws within its jurisdiction reducing outlays by 260 million dollars in fiscal year 1997, by 98,321 million dollars across the period covering fiscal years 1997 to 2002, and by 36,578 million dollars in fiscal year 2002. The cost estimate printed with the measure in Senate Print 104-59 showed that the Finance Committee title as reported complied in part with the instruction but did not deliver the required outlay savings for fiscal year 2002. That single year shortfall changed the legal position of every expensive provision in the title. On July 18, 1996, as recorded in the Congressional Record at page S8081, Senator Lott used the second test to strike Medicaid spending that section 2923 of the measure had added, on the ground that a provision increasing outlays inside a title whose committee had failed its instruction was extraneous under section 313(b)(1)(B). The Medicaid provision was not challenged as nonbudgetary, since an increase in program spending is budgetary by definition, and it was not challenged as belonging to another committee, since Medicaid sat squarely within the Finance Committee’s jurisdiction. The provision fell because the instruction had been written year by year, so compliance in the aggregate could not cure a miss in a single named year. The Congressional Research Service account describes the 1996 invocation as the most recent recorded since the rule’s adoption, a measure of how completely the deterrent effect operated in committee.

The third test, in section 313(b)(1)(C), provides that a provision is extraneous if the provision is not in the jurisdiction of the committee with jurisdiction over the title or provision in question, a formulation that the Congressional Research Service states in plainer terms as a provision outside the jurisdiction of the committee that submitted the title or provision for inclusion in the reconciliation measure. Jurisdiction in the Senate is assigned by standing Rule XXV, which allocates subjects among the standing committees, and the budget resolution’s reconciliation instructions are addressed to committees by name, directing each committee to report changes in the laws within its jurisdiction. The third test welds those two allocations together. A committee’s reconciliation slot, meaning its share of the one reconciliation bill that the instructions authorize, may be spent only on the subjects that the standing rules assign to that committee, and the parliamentarian determines the boundary by reference to Rule XXV and to the referral history of comparable measures. The test applies to provisions in a reported bill and, with particular force, to floor amendments, because an amendment reaching a subject assigned to a committee that received no instruction in that cycle had no route into the bill.

The mechanism the third test polices is territorial, and the abuse it prevents is the use of one committee’s reconciliation slot as a vehicle for another committee’s legislation. A budget resolution instructs the committees whose programs and revenue sources are needed to reach the resolution’s fiscal totals, and committees outside the instruction receive no share of the expedited procedure for that cycle. A committee legislating beyond its jurisdiction, or a senator amending a reconciliation bill on a subject assigned to an uninstructed committee, would obtain the twenty hour debate limit and the majority vote threshold for a measure the Senate had never agreed to include in the reconciliation enterprise. Section 313(b)(3) states two exceptions that keep the territorial rule from defeating ordinary drafting. A provision reported by a committee is not extraneous under the third test if the provision is an integral part of a provision or title that, introduced as a standalone measure, would be referred to that committee, and the provision sets forth the procedure to carry out or implement substantive provisions that fall within the committee’s jurisdiction. A provision is likewise protected if the provision states an exception to, or a special application of, the general provision or title of which it forms a part, where that general provision or title would itself be referred to the reporting committee.

Documented applications of the third test cluster around floor amendments, and Senate consideration of the Health Care and Education Reconciliation Act of 2010 produced the pattern at scale. During consideration of H.R. 4872 on March 23 to March 25, 2010, nine amendments fell after points of order under the rule were sustained, in each instance after a motion to waive had been rejected, and the Congressional Research Service records that all but one of those points of order rested on the ground that the amendment included provisions outside the jurisdiction of the instructed committees. The cleanest single documented case arose on December 3, 2015, during consideration of H.R. 3762, the Restoring Americans’ Healthcare Freedom Reconciliation Act of 2015. The budget resolution for fiscal year 2016, Senate Concurrent Resolution 11, carried reconciliation instructions for the Finance Committee and the Health, Education, Labor, and Pensions Committee and for no other Senate committee. Senator Joe Manchin offered amendment number 2908, relating to gun control and civil liberties, a subject that Rule XXV assigns to the Judiciary Committee. A point of order under the third test was raised, a motion to waive the point of order was rejected by a vote of 48 to 50, and the amendment fell. No question about the amendment’s budgetary effect required resolution, because the threshold fact that its subject belonged to an uninstructed committee disposed of the matter completely. Section 313(d), which governs conference reports, omits the jurisdiction test from the grounds for a point of order against a conference report, while the parliamentarian advised that a provision outside the jurisdiction of every instructed Senate committee could nonetheless be fatal to the privileged consideration of the measure.

Enforcement of all three tests follows the single procedure that section 313 establishes for extraneous matter. The rule is not self executing. A senator must be recognized on the floor and must state a point of order identifying the provision challenged, which may be designated by title, by section number or by page and line, and section 313(e) permits a single point of order to reach several provisions at once, with the presiding officer ruling on each provision separately and sustaining the point of order as to some provisions and not others. The presiding officer rules with the advice of the Senate parliamentarian. Before the presiding officer rules, any senator may move to waive the point of order as it applies to some or all of the provisions challenged, and a waiver requires the affirmative vote of three-fifths of the senators duly chosen and sworn, sixty votes in a Senate with no more than one vacancy; an appeal of the presiding officer’s ruling carries the same three-fifths requirement. Where no waiver is obtained and the point of order is sustained, section 313(a) provides that the part of the title or provision containing the extraneous material is deemed struck from the bill and may not be offered again as an amendment from the floor. The bill itself continues in consideration, the feature that gives the rule its surgical reputation and that the December 2017 episode illustrated at conference stage. Section 313(c) adds a documentary step addressed to the first two tests directly, directing the Senate Budget Committee, when a reconciliation bill is reported or discharged and again when a conference report is submitted, to place in the record a list of material considered extraneous under sections 313(b)(1)(A) and 313(b)(1)(B), together with the outyear deficit test, while providing that inclusion in or exclusion from the list does not constitute a determination by the presiding officer.

Measured against the fourth test, which treats a provision as extraneous where its changes in outlays or revenues are merely incidental to its nonbudgetary components, the first three tests occupy the low litigation end of the rule. Litigation in this setting means contested argument before the presiding officer over characterization, and the first three tests generate little of it because each test terminates in a document: an estimate showing no change or a change in the wrong direction inside a noncompliant title, an instruction printed in a budget resolution against which a title’s net effect can be scored, and a jurisdictional assignment in Rule XXV supported by referral precedent. The fourth test terminates in no document at all, because the statute supplies no definition of merely incidental and no scale on which a budgetary effect and a policy effect can be weighed against each other, so every application requires the judgment that the first three tests are designed to avoid. Counsel advising a committee on the first three tests could therefore give a firm answer in the great majority of cases and could cure a defect by rewriting a definition, rescoping a title to its instruction or returning a provision to its own committee. The sequence of section 313(b)(1) is a sequence in procedural certainty as much as a list of defects, moving from the presence of money, to obedience to instructions, to respect for jurisdiction, and only then to the question of proportion that the next section takes up.

Merely Incidental: The Test That Does the Work

The fourth definition of extraneous matter in section 313(b)(1)(D) of the Congressional Budget Act, codified at 2 U.S.C. section 644, reaches a provision that produces a change in outlays or revenues when that change is merely incidental to the non-budgetary components of the provision. The Congressional Research Service, in its report The Budget Reconciliation Process: The Senate’s Byrd Rule by Bill Heniff Jr., sets out the six definitions in sequence, and the fourth differs in kind from the five around it. The first definition asks whether a provision changes outlays or revenues at all, a question a score from the Congressional Budget Office or the Joint Committee on Taxation ordinarily answers. The second asks whether a committee met its instruction, an arithmetic comparison. The third asks which committee holds jurisdiction, answered from referral precedent. The fifth asks whether the deficit grows beyond the budget window, another arithmetic comparison. The sixth asks whether a provision recommends changes in the old age, survivors, and disability insurance program, answered by identifying the subject matter. The fourth definition alone asks for a characterization: a judgment about what a provision is for, and about whether the money it moves is the substance of the provision or an accompaniment to substance located elsewhere.

Congress supplied no apparatus for making that judgment. The statute contains no definition of the operative phrase, no ratio between budgetary and non-budgetary effects, no dollar threshold beneath which an effect counts as incidental, and no list of factors to weigh. The text states a conclusion, extraneousness, and leaves the reasoning to be built case by case. The decided material under the fourth test therefore consists of accumulated advice: arguments accepted and rejected in earlier reviews, distinctions drawn between provisions that passed and provisions that fell, and analogies carried from one reconciliation cycle into the next. The Congressional Research Service has described determinations under this definition as lying at the core of many Byrd rule controversies, and has observed that each is necessarily specific to the provision under review, which complicates any effort to state a bright line in advance. A score can be read off a table. A purpose has to be argued.

What does merely incidental mean in practice?

In practice, the test asks whether the money is the point of the provision or only a side effect of the point. A provision passes when changing spending or revenue is the mechanism that does the work, and it fails when the fiscal change only accompanies a decision made for regulatory or social reasons.

The comparison the test demands has two sides of unequal measurability. The budgetary side arrives quantified, because the Congressional Budget Office and the Joint Committee on Taxation estimate the change in outlays or revenues across the budget window. The non-budgetary side arrives unquantified, because no agency scores the size of a regulatory change, the reach of a condition placed on private conduct, or the significance of a change in the legal status of an activity. The advice therefore weighs a number against a description. Arguments in the pre-floor review process known as the Byrd bath take a predictable shape. Defenders emphasize the estimate, the directness of the payment or revenue stream, and the provision’s location inside a spending or tax mechanism. Challengers emphasize the conduct governed, the population reached, and the policy debate that produced the provision, and contend that the estimate measures the shadow of the provision rather than the provision itself. Because the statute fixes no exchange rate between dollars and policy significance, the advice turns on which description of the same text persuades: a budget measure with policy consequences, or a policy measure with budget consequences. Nearly every contested provision can be described in both vocabularies, which is the first reason the fourth test absorbs disputes that the other five cannot settle.

The second reason is structural position. The fourth test is the residual test, available after the mechanical questions have been answered in a provision’s favor. A provision that changes outlays passes the first test, sits inside the reporting committee’s jurisdiction passes the third, and fits inside the window passes the fifth, and a challenger who concedes each point has exactly one ground left. Any provision with a genuine budgetary effect and a contestable purpose is therefore funneled into a single inquiry. Challenges under the other definitions tend to be resolved in drafting, because staff can compute a score, check a referral, or move a date and cure the defect before the bill reaches the floor. A challenge under the fourth test cannot be cured by computation, because the objection is to the character of the provision rather than to any figure in it, and objections of character are the ones majorities and minorities most often insist on pressing. The inquiry attaches to the relationship between money and policy inside a text rather than to the party that drafted the text, and it reaches bills of both parties on the same logic.

Relativity governs the inquiry throughout, and relativity produces results that a dollar threshold never would. The Congressional Research Service has stated the point in both directions: matter with significant fiscal effects but even larger policy effects can be deemed extraneous, and a provision with small budgetary effects and negligible policy effects can be permissible. Size alone settles nothing. A provision moving a modest sum inside a program it also administers may present no difficulty, because the budgetary change and the policy change are the same act described once. A provision moving a very large sum may fail, because the sum purchases a transformation of conduct or status that dwarfs it. The inquiry instead tracks the relationship between the fiscal effect and the provision’s evident object, gathered from its operative terms, its placement, its duration, and the mechanism it uses. A condition written into the definition of an expenditure reads differently from the same condition enacted as a free standing rule of conduct with a funding consequence attached.

The asymmetry the test creates follows from that relationship, and it runs in both directions along the same axis. At one end, a provision whose budgetary effect is large, direct, and produced by its own operative mechanism can survive review even when its policy ambitions are open and substantial, because the fiscal effect is the instrument by which the policy operates, and the statute asks only whether the effect is incidental, not whether the provision also pursues a policy. Tax legislation illustrates the baseline. A change in rates, brackets, credits, or deductions is a policy choice through and through, yet its budgetary effect is its operative content, and reconciliation practice has treated such changes as the core case of permissible matter. At the other end, a provision whose federal cost is small, indirect, or deliberately minimized fails even when every participant agrees that the provision matters a great deal, because agreement about importance is agreement that the provision’s significance is non-budgetary, and a small fiscal shadow cast by a large policy object is the definition of incidental. Between those poles sits the hardest category, the provision with large effects of both kinds, and the 2015 health care reconciliation episode placed that category on the record in unusually explicit terms.

The provision at issue in 2015 was repeal of the individual mandate in section 5000A of the Internal Revenue Code, the requirement that most residents hold qualifying health coverage or pay a penalty. The House passed version of the Restoring Americans’ Healthcare Freedom Reconciliation Act of 2015 would have repealed the mandate outright. Advice from the parliamentarian, reproduced in Congressional Research Service accounts of the rule, held the repeal extraneous under section 313(b)(1)(D). The reasoning conceded the dollars and denied their sufficiency. Repeal carried a net saving estimated at approximately 147 billion dollars over ten years in combination with repeal of the employer mandate. The advice nonetheless characterized the mandate as a massive national policy change reaching approximately 270 million people, designed to change behavior and to move the country toward universal coverage, and held that the dollars were dwarfed by the scope and impact of the requirement itself. The budgetary impact, in the formulation the advice adopted, was merely incidental to the policy that underpinned the provision. The episode fixed the outer marker for the test: an effect measured in the hundreds of billions can fail, provided the non-budgetary component is a nationwide rule of conduct that the fiscal effect merely prices.

The same episode also marked the inner boundary, in the distinction the advice drew between repealing a mandate and adjusting its penalty. The Senate version of the 2015 legislation reduced the penalty to zero rather than repealing the underlying requirement, and advice in the same line of disputes treated that adjustment as permissible. Penalties, the reasoning ran, are inherently budgetary, amounts and formulas within a revenue provision can be adjusted in a reconciliation bill as they had been adjusted in the Health Care and Education Reconciliation Act of 2010, and an adjustment that runs to zero remains an adjustment of a fiscal lever even though the mandate text stands untouched. The distinction is formal, and its formality is the point. The fourth test examines the character of the mechanism a provision employs, and a provision that works by setting the price of noncompliance operates through the budget in a way that a provision abolishing the duty itself does not. The technique of reaching conduct through the amount of an exaction, rather than through the existence of the exaction, became a standard feature of reconciliation drafting in health care and tax bills alike.

Two years later, the test decided the fate of the most contested social policy provisions in the Republican health care bill of 2017, displaying the same logic at the opposite end of the dollar scale. On July 21, 2017, Senate Budget Committee Democrats released the parliamentarian’s guidance on the Better Care Reconciliation Act, and Health Affairs Forefront reported the guidance the same week. Elizabeth MacDonough, the Senate parliamentarian, advised that the provision barring federal payments to Planned Parenthood for one year was subject to a Byrd rule point of order, as were the provisions restricting the use of premium tax credits and small business tax credits to purchase insurance plans that cover abortion. The budgetary effects were real, because a one year exclusion of a provider from Medicaid payment streams changes federal outlays, and conditions on tax credits change the terms on which federal money supports premiums. The advice nonetheless treated those effects as incidental to the provisions’ regulatory and social policy purposes, which concerned abortion related services and coverage and the status of a named provider within federal health programs. A longer pedigree supported the advice. Former parliamentarian Robert Dove, in a 2010 interview reported in health care trade coverage of the 2017 dispute, recalled having ruled an abortion restriction out of order in a 1995 reconciliation bill on the ground that the provision had not been included in order to save money but in order to implement social policy. Advice given in 1995, in 2015, and in 2017 proceeded from the same comparison, and each treated a scoreable saving as the beginning of the inquiry rather than its end.

The tax reconciliation bill of late 2017 then supplied, inside a single process, an exclusion and a survival that together map the test more clearly than either does alone. The exclusion concerned the Johnson Amendment, the provision of the Internal Revenue Code, dating to 1954, that bars organizations exempt under section 501(c)(3) from participating or intervening in political campaigns on behalf of or in opposition to candidates. The House passed tax bill would have created a safe harbor allowing such organizations limited campaign related statements made in the ordinary course of regular activities at no more than de minimis incremental expense. The Senate bill contained no corresponding provision, and the question reached the conference stage. On December 14, 2017, Senator Ron Wyden, the ranking Democrat on the Finance Committee, announced that the parliamentarian had advised that the Johnson Amendment language violated the Byrd rule, and the language was kept out of the final bill, as the Washington Post reported and as the Baptist Joint Committee for Religious Liberty confirmed in a statement dated December 15, 2017. Ellen Aprill and Daniel Hemel, in their study of the 2017 tax process titled The Tax Legislative Process: A Byrd’s Eye View, counted the measure allowing section 501(c)(3) organizations to participate in political campaigns among the provisions the rule eliminated from the final legislation. The structure of the House proposal had made the fiscal side of the comparison exceptionally weak. A safe harbor drafted around de minimis expense is a provision designed to cost almost nothing, and such a provision confronts the fourth test with a budgetary effect approaching zero while its non-budgetary component, the conditions under which tax exempt organizations may address elections, is the entire subject of the provision. The comparison the statute orders was one sided.

The survival concerned the Arctic National Wildlife Refuge. The fiscal year 2018 budget resolution instructed the Senate Energy and Natural Resources Committee to report legislation reducing the deficit by one billion dollars, and the committee, chaired by Senator Lisa Murkowski of Alaska, responded with legislation opening the coastal plain of the refuge to oil and gas leasing. The Congressional Budget Office estimated that the legislation would raise 1.092 billion dollars over the ten year budget window, according to the committee’s release of the chairman’s mark in November 2017, through lease sales, bonus bids, rentals, and royalty shares divided between the federal government and the state of Alaska. The policy stakes were among the largest in the bill. The coastal plain had been closed to development under the Alaska National Interest Lands Conservation Act of 1980 unless Congress authorized development by later statute, and the leasing program reversed that posture for a defined area under a statutory program with mandatory lease sales. An early version drew adverse advice on a different ground, encroachment on the jurisdiction of another committee, and a wording change resolved that defect, as Aprill and Hemel recount. On the merely incidental question, the parliamentarian advised that the measure did not run afoul of the rule, and the provision entered the enacted law as section 20001 of Public Law 115-97. The structure of the provision explains the outcome without recourse to the merits of drilling policy. The provision did not regulate an industry and incidentally raise money. It sold access to a federal asset under a congressionally set revenue instruction, and the revenue was the mechanism, quantified in advance, that the reporting committee had been directed to produce. The policy transformation was vast, and it arrived as a leasing program, and a leasing program is a budgetary instrument in its operative terms even when the land it opens has been contested for decades.

Read together, the refuge survival and the Johnson Amendment exclusion expose the test’s indifference to political weight and its attention to fiscal architecture. The refuge provision carried the heavier consequences and survived, because its budgetary effect was direct and produced by its own sales mechanism. The Johnson Amendment provision failed, because its fiscal effect had been engineered toward zero. A December 2017 episode from the same conference process sharpened the smaller end of the scale. The conference report would have allowed section 529 education savings accounts to fund homeschooling expenses, by treating a homeschool within the definition attached to the benefit. A point of order raised on December 19, 2017 was sustained after a waiver motion failed, and the language was struck. The parliamentarian’s advice, as summarized in the Senate Budget Committee print and reproduced in Congressional Research Service accounts of the rule, treated the homeschool language as a significant policy change attached to a small part of a section whose overall score was modest, with the term itself undefined in federal law. Smallness of score, paired with a definitional change of independent policy significance, was the ground of decision.

Drafting practice around the test is visible in the record of these episodes, as descriptive features of how reconciliation bills have been assembled. One pattern works through the terms and conditions of federal payment. Because the first definition expressly accommodates changes in the terms and conditions under which outlays are made or revenues are collected, provisions that define eligibility, set payment formulas, or condition a federal expenditure operate closer to the budgetary core than provisions that announce a rule of conduct and rely on downstream fiscal consequences for the connection. A second pattern adjusts the fiscal lever instead of the legal duty, reducing a penalty to zero, altering a rate, or revising a formula, on the logic displayed in the mandate disputes. A third pattern builds the revenue mechanism into the provision itself, as the refuge legislation did with scheduled lease sales and royalty shares, so that the effect flows from transactions the provision orders rather than from behavior it predicts. A fourth pattern is subtraction during the Byrd bath itself, where adverse advice leads drafters to narrow a provision, sever a regulatory clause, relocate a policy objective to a regular order vehicle requiring sixty votes, or accept a wording change of the kind that cured the refuge provision’s jurisdictional defect. Each pattern reflects the same recognition: the fourth test rewards provisions whose fiscal effects are produced by their operative terms and penalizes provisions whose fiscal effects are asserted as consequences.

What the accumulated record leaves unresolved is any general formula, and the absence is a feature of the test’s design rather than a gap awaiting a technical fix. A ratio would invite drafting to the ratio, and a threshold would convert an inquiry about character into an inquiry about arithmetic that the other definitions already perform. The comparison instead proceeds by example and analogy, provision against provision, with the parliamentarian’s advice supplying continuity across bills and across changes in party control. That method carries a cost in predictability, because sponsors learn the disposition of a novel provision only when advice issues, and a corresponding discipline, because the advice must be defensible against earlier disputes of the same kind. The fourth test therefore stands between reconciliation and the general legislative agenda, decides most of the disputes that reach decision, and decides them on a phrase Congress enacted without a definition, applied through precedent, to the recurring question of whether the money in a provision is its purpose or its excuse.

The Window, Social Security, and the Sixty-Vote Escape Valve

The fifth test in section 313(b)(1)(E) of the Congressional Budget Act asks a question about time. A provision is extraneous if it would increase net outlays or decrease revenues in a fiscal year beyond those covered by the reconciliation bill, and if those increases or decreases are not offset by other provisions in the same title in that year. The Congressional Research Service, in its report The Budget Reconciliation Process: The Senate’s “Byrd Rule,” paraphrased the test as a bar on increasing the deficit for a fiscal year beyond the budget window covered by the reconciliation measure. Both formulations describe a single hard edge. Inside the period covered by the budget resolution and its directives, a title may raise deficits to the extent the directives allow. Past the final covered year, a deficit increase in even one later year is enough to make offending language extraneous.

The window itself comes from the budget resolution rather than from section 313, and the statute fixes no length for it. The Congressional Research Service explained that a budget resolution was required, beginning in the late 1980s, to cover at minimum the budget year and the four fiscal years following it, and that resolutions sometimes added the current year and as many as five additional years, so that the longest window applied to a budget resolution and its reconciliation legislation covered eleven years including the current year. In practice, windows ordinarily spanned about a decade, and legislative debate treated ten years as the standard frame. Whatever its length in a given cycle, the resolution’s last covered year became the boundary that section 313(b)(1)(E) policed, and provisions with effects running past that boundary were written at the mercy of a point of order.

The measurement requires precision, because the test operates at two levels at once. The statutory text speaks of a provision that increases net outlays or decreases revenues in a year beyond the window without an offset elsewhere in the same title in that year. The Congressional Research Service, in its standing report on the Byrd rule, explained the application in terms of the reported title taken as a whole. If the overall on-budget fiscal effects of a committee’s reconciliation submission would cause the deficit to increase in any year beyond the years covered by the associated budget resolution, and a single such year is sufficient, then any provision within that title that would increase outlays or reduce revenues in an outyear is extraneous. An offset therefore had to sit inside the same title and inside the same outyear to cure the defect; savings in a different title or a different year did not answer the point of order.

Determinations under the test rested on estimates rather than on floor argument. The Congressional Research Service recorded that determinations of budgetary levels for enforcing the Byrd rule were made by the Senate Budget Committee, working from estimates by the Congressional Budget Office and, for revenue provisions, the Joint Committee on Taxation. The scoring agencies ordinarily did not produce point estimates beyond the budget window, and the Congressional Budget Office could instead include a statement in a cost estimate indicating whether a committee’s title would increase the deficit in the outyears. A title therefore passed or failed on the sign of its outyear effect before any precise magnitude, and a committee seeking certainty had to build the offset into the title itself rather than promise favorable effects after the window closed.

The 2003 tax legislation supplied the cleanest illustration of the test in operation against an amendment. During consideration of the Jobs and Growth Tax Relief Reconciliation Act of 2003, Senator Jeff Sessions offered an amendment, designated S.Amdt. 639, that would have sunset permanent revenue increases which offset the cost of the measure’s temporary tax cuts. The budget window in the associated budget resolution, H.Con.Res. 95 of the 108th Congress, covered fiscal year 2003 through fiscal year 2013. The amendment would have ended the revenue offsets in fiscal year 2015, and the Congressional Research Service recorded that agreement to the amendment would have caused the underlying measure to increase budget deficits in one or more years beyond those covered by the budget resolution. The amendment fell on a point of order. The language struck was not an extravagance attached to the bill but a change to the timing of the bill’s own offsets, and the arithmetic of the outyears condemned it regardless of the policy case for letting temporary cuts and permanent offsets expire on different schedules.

That mechanical quality separates the fifth test from the merely incidental test that dominates contested Byrd rule litigation. The fourth test asks for a judgment about the relative weight of a provision’s budgetary and nonbudgetary components, a judgment the statute leaves undefined. The fifth test asks whether a number is positive in a specified year. Argument can shape estimates at the margin, but no speech about the importance of a provision answers a point of order grounded in an outyear deficit increase. Drafters therefore responded to the fifth test in the drafting room rather than on the floor. A provision with unoffset outyear costs could be paired with offsets inside the same title, as the drafters of the Tax Cuts and Jobs Act of 2017 paired the permanent reduction in the corporate income tax rate with base broadening and international tax reforms that the Congressional Research Service described as offsetting its cost. A provision without such offsets could be given an expiration date inside the window, as the mandatory spending for Pell grants in the College Cost Reduction and Access Act of 2007 was made available only for fiscal year 2008 through fiscal year 2017. The sunset, rather than the speech, became the standard answer to the fifth test, and the architecture of expiring provisions that the test produced is examined in the discussion of sunsets that follows in this article.

The sixth test in section 313(b)(1)(F) removes an entire program from the reconciliation vehicle. By its terms, a provision is considered extraneous if it violates section 310(g) of the Congressional Budget Act, and section 310(g) provides that it is not in order in either house to consider a reconciliation bill, an amendment, or a conference report containing recommendations with respect to the old-age, survivors, and disability insurance program established under Title II of the Social Security Act. The Congressional Research Service described the sixth definition as a complement to that freestanding ban, and the pairing matters for enforcement, because a senator confronting Title II language held two different points of order with two very different consequences, a distinction addressed below in connection with the waiver machinery.

Precision about the program’s identity is essential, because the bar reaches one title of one statute and not the statute as a whole. The Social Security Act of 1935 grew into an umbrella carrying many programs under separate titles, and section 310(g) names only the insurance program in Title II. The Congressional Research Service stated the point directly in its Byrd rule report: for purposes of these provisions, Social Security means the Old-Age, Survivors, and Disability Insurance program established under Title II of the Social Security Act, and the term does not include Medicare or other programs established as part of that act. Its accounts of reconciliation practice made the same point by enumeration, naming unemployment insurance under Title III, Medicare under Title XVIII, Medicaid under Title XIX, and the State Children’s Health Insurance Program under Title XXI as programs that had been amended by reconciliation legislation without violating the definition. Medicare provisions therefore moved through reconciliation as a matter of routine across the decades after 1985, while a provision altering Social Security benefits or the payroll taxes that finance Title II could not be carried at all.

The word that does the work in the sixth test is recommendations, and neither section 310 nor section 313(b)(1)(F) defines it. Precedent, as the Congressional Research Service reported, did not confine extraneous matter under the test to provisions affecting the financial operations of the Social Security trust funds. The documented application from 2015 reached administrative language rather than benefit or tax language. In that year, the point of order was used to strike language preemptively, and the language removed would have amended provisions in Title II relating to the use and collection of Social Security numbers and would have eliminated certain reports required under Title II. A provision touching Title II only at the level of numbering privacy and reporting duties was treated as a recommendation with respect to the program, which indicates how broadly the fence was understood to run. The fence had a surveyed boundary on the indirect side. The Congressional Research Service reported that the Senate Parliamentarian had advised that provisions with indirect effects on the Social Security trust funds were not extraneous, giving as an example a change in Social Security revenues attributable to a change in the definition of adjusted gross income, and that whether an effect counted as direct or indirect was decided case by case.

The consequence of the sixth test is visible less in floor defeats than in bills that were never assembled. No reconciliation bill enacted between the adoption of the Byrd rule in the 1980s and the end of 2019 carried a restructuring of Social Security benefits or of the taxes dedicated to Title II, and the absence is structural rather than accidental. A majority intent on restructuring the program faced a vehicle that excluded the subject by definition, so the drafting decision was made before any parliamentarian ruling and usually before any public text existed. The Congressional Research Service’s 2015 example captures the pattern from the other end: the language the test was called upon to remove was peripheral Title II language caught in a preemptive scrub, because core Social Security language was kept out at the drafting stage as a matter of course. For the central subject, the record of the sixth test through 2019 is a record of deterrence, in which the test did its work by making inclusion pointless rather than by producing contested rulings on the Senate floor.

Major Social Security legislation accordingly belongs to the history of regular legislation. The Social Security Amendments of 1983, enacted as Public Law 98-21 and signed on April 20, 1983, predated the Byrd rule, and the amendments serve as the standard illustration of the route a large restructuring takes when reconciliation is unavailable as a matter of law or, after 1985, forbidden as a matter of order. The 1983 package followed the report of the National Commission on Social Security Reform and moved through the ordinary committee and conference process in both houses. After the Byrd rule became permanent in 1990, that ordinary route carried a price reconciliation exists to avoid, because a bill under the regular rules of the Senate is exposed to extended debate and ordinarily requires sixty votes for cloture before a final vote can occur. A Social Security restructuring after 1990 therefore needed a bipartisan coalition broad enough to end debate, and the record through 2019 reflects that threshold rather than any absence of proposals.

How many votes does it take to waive the Byrd rule?

Waiving a Byrd rule point of order takes sixty votes in a full Senate. Section 904 of the Budget Act requires the affirmative vote of three-fifths of the senators duly chosen and sworn, and sustaining an appeal of the presiding officer’s ruling on the point of order requires the same three-fifths majority.

The escape valve is deliberately narrow, and its shape explains much of the rule’s practical force. Section 904 of the Congressional Budget Act supplies the general waiver and appeal procedures for points of order under the act in the Senate, and the Congressional Research Service summarized the application to the Byrd rule in a single sentence: a motion to waive the Byrd rule, or to sustain an appeal of the ruling of the chair on a point of order raised under the rule, requires the affirmative vote of three-fifths of the membership, which is sixty senators when no seat is vacant. The denominator is the full membership duly chosen and sworn rather than the senators present and voting, so absences raise the effective burden on the senators in the chamber.

The two routes differ in what a successful vote concedes. A waiver, as the Congressional Research Service explained in its accounts of the rule, recognizes that the matter violates the Byrd rule and records a decision by the Senate to allow the matter regardless. An appeal asks the Senate to hold that the presiding officer decided the question incorrectly, and a successful appeal establishes new precedent, changing the interpretation of what the Senate considers extraneous. A majority with sixty votes that valued the provision more than the precedent could waive and keep the interpretation intact for the next bill, while a majority that appealed and prevailed rewrote the boundary for every later measure. The mechanics reinforce the choice. A Byrd rule point of order is not debatable, the Congressional Research Service noted, so a motion to waive has to be made immediately, while the point of order is pending and before the presiding officer rules. A waiver may also be offered preemptively, and section 313(e) permits a single motion to waive to cover one, some, or all of the provisions named in a point of order raised against several provisions at once. One waiver motion could also reach the Byrd rule together with other provisions of the Congressional Budget Act and could extend, for specified language, through consideration of the conference report.

What a senator wins by defeating a waiver is correspondingly precise. When a point of order is sustained against a provision in a reconciliation bill as reported, the provision, and only the provision, is deemed stricken from the bill under section 313(a), the stricken matter may not be offered again as an amendment from the floor, and the measure proceeds with the remainder of its text intact. When the point of order is raised against an amendment, the amendment falls and the underlying bill proceeds without it. The Congressional Research Service described the application as surgical for exactly this reason: the point of order is raised against specific text, the presiding officer may sustain it as to all, some, or none of the provisions named, and the rest of the measure remains before the Senate. Faced with Title II language, the Congressional Research Service explained, a senator could raise the point of order under the Byrd rule and strike only the offending matter, or raise the point of order under section 310(g), which if sustained would be fatal to the overall measure and would remove the bill from the floor immediately.

Conference reports receive a dedicated procedure in section 313(d), because a conference report reaches the Senate as an unamendable package and a simple strike would leave the two houses without a text on which to agree. Section 313(d) applies the Byrd rule during consideration of a conference report and during the exchange of amendments between the houses, and the Congressional Research Service noted that the conference report provision reaches five of the six definitions of extraneous matter, omitting the definition based on committee jurisdiction in section 313(b)(1)(C). If a point of order against a conference report is made and sustained, the offending matter is deemed stricken, and the Senate proceeds without intervening action or motion to consider whether to recede from its amendment and concur with a further amendment, or to concur in the House amendment with a further amendment, where the further amendment consists only of the portion that was not stricken. The Congressional Research Service described the sequence in plainer terms: the matter is removed, the Senate proceeds at once to consider an amendment in the nature of a substitute containing the text of the conference agreement minus the stricken provisions, and if the Senate agrees to that amendment, the measure is sent back to the House for further consideration. A sustained point of order at the conference stage therefore reopens a bargain both houses had closed, and the House must act again on a text missing language its conferees had accepted. The conference report on the Tax Cuts and Jobs Act of 2017 demonstrated the sequence in December 2017, when, as the Congressional Research Service recorded, three violations of the Byrd rule were sustained, the Senate agreed to concur with an amendment in the nature of a substitute, and the House later adopted that amended text before enactment.

Waivers of the rule have been rare in practice, and the tally from the rule’s first decades shows how rarely the escape valve opened. The Congressional Research Service, in the tally published in its 2010 edition of The Budget Reconciliation Process: The Senate’s “Byrd Rule,” counted fifty-two motions to waive the Byrd rule that had been offered and disposed of, of which nine were approved and forty-three were rejected. Eight of the nine successful motions protected committee reported language in a bill or language in a conference report, and only one successful motion protected a floor amendment. The pattern fits the structure of the threshold. Reconciliation is valuable to a majority precisely because the majority lacks sixty reliable votes, since a coalition holding sixty votes could invoke cloture on an ordinary bill and legislate under the regular rules without accepting the Byrd rule’s content restrictions at all. A waiver motion asks that same short-handed majority to assemble, for one provision, the very supermajority whose absence justified the expedited vehicle, usually by persuading minority party senators to vote to suspend a rule written to restrain majorities. The mathematics occasionally worked for language that committees had reported or conferees had accepted, which is where the nine successes clustered, and failed for the floor amendments that reconciliation majorities most often wanted to protect. The sixty-vote escape valve therefore exists in the statute as a genuine possibility and functions in practice as a restatement of the rule itself: matter that can command sixty votes does not need the Byrd rule’s permission, and matter that cannot command sixty votes rarely receives it.

Why Tax Cuts Expire: The Sunset Mechanism at Work

Of the six tests that define extraneous matter under section 313 of the Congressional Budget Act, the fifth test operates differently from the others because its consequences remain visible in the statute book long after the floor fight ends. The first four tests and the sixth test strike provisions during consideration, and a provision struck under them simply disappears from the bill. The fifth test, codified at section 313(b)(1)(E), rarely strikes anything at all, because drafters absorb its demand in advance and write the demand into the text. A provision falls within the fifth test when it increases net outlays or decreases revenues in a fiscal year beyond the years covered by the reconciliation measure, and that increase or decrease is not offset by other provisions in the same title in that year. The result is a distinctive feature of modern tax legislation: the expiration date, or sunset, under which a rate reduction, a credit, or a deduction operates for a fixed span of years written into the statute and then ceases, with prior law reviving in its place, unless a later Congress enacts an extension.

The Congressional Research Service, in its accounts of the Byrd rule, describes the fifth test as a judgment rendered on a title as a whole rather than on a provision standing alone. The budgetary effects of the title reported by an instructed committee are measured together, year by year, into the years beyond the budget resolution window. If the combined effect of the title increases the deficit in any single fiscal year beyond the window, a single year being enough, then each provision in that title that increases outlays or reduces revenues in an outyear is exposed to a point of order as extraneous matter. Offsets inside the same title answer the test, because revenue raised or outlays saved by one provision in an outyear counts against revenue lost by another provision in that same outyear. The estimates that govern the measurement come from the Congressional Budget Office and, for revenue provisions, the Joint Committee on Taxation. Drafters therefore compose a reconciliation title the way an accountant closes a ledger, arranging gains and losses so that every year beyond the window balances, and the sunset is the most flexible entry in that ledger, since a provision scheduled to end contributes nothing at all to the years after its termination.

Why do tax cuts passed through reconciliation expire?

Tax cuts enacted through reconciliation expire because the fifth Byrd test treats a provision as extraneous when the title containing it increases the deficit in any year beyond the budget window, so drafters end revenue reductions inside the window and let the statute restore prior law afterward.

A drafter confronting the fifth test holds three options, and the choice among them explains most of the shape of modern tax statutes. The first option is permanent offset: a permanent tax reduction paired with permanent revenue increases or permanent spending reductions large enough to cover the loss in every outyear. The second option is a waiver, which requires sixty votes in the Senate and therefore surrenders the central advantage of reconciliation, passage by a simple majority. The third option is the sunset, under which the reduction operates inside the window and terminates before the outyears begin. The sunset converts an open ended revenue loss into a bounded one with a measurable cost, and it requires no offsetting sacrifice elsewhere in the title. For a majority that commands fifty-one votes rather than sixty, the arithmetic is decisive. A sunset also operates as a forcing event rather than an ending: a tax reduction scheduled to expire creates a future deadline at which a later Congress faces pressure to extend it, on the wager that a provision popular enough to enact once will prove difficult to allow to lapse.

The template for that wager was set in 2001. The Economic Growth and Tax Relief Reconciliation Act of 2001, Public Law 107-16, signed on June 7, 2001, moved through reconciliation under an instruction in the budget resolution that permitted a fixed total of tax reduction across the years the resolution covered. The act reduced marginal individual income tax rates in stages, created a new lowest bracket, widened the bracket structure for married couples filing jointly to relieve the marriage penalty, enlarged the child tax credit, expanded education and retirement saving incentives, and phased down the estate tax toward a single year of complete repeal. Each element carried a revenue cost that grew as the phase ins completed, so the statute as its proponents preferred it, a permanent restructuring of individual taxation, could not fit inside the instruction total or survive the fifth test in permanent form. The reconciliation instruction fixed how much revenue the measure was permitted to lose inside the window, and the fifth test forbade the measure from losing revenue beyond the window without offsets the majority had not assembled. Permanence failed on both counts at once.

The solution adopted in 2001 was a general sunset of extraordinary breadth. Rather than assigning separate termination dates title by title, the act provided that substantially its whole body of amendments would expire at the end of 2010, for tax years beginning after December 31, 2010, with the law reverting to the provisions in effect before enactment. The rate schedules carried the termination date. The estate tax provisions carried the same date, so the repeal scheduled for 2010 alone stood surrounded on both sides by years in which the tax operated, a structure produced entirely by the interaction of the phase down schedule with the general expiration. The enlarged child tax credit, the marriage penalty relief, the education incentives, and the retirement provisions all shared the single deadline. The Congressional Research Service recorded the reason directly in its description of the act: Congress included the sunset to avoid a Byrd rule violation in the Senate, because the rule treats as extraneous any language that increases the deficit in a fiscal year beyond those covered by the reconciliation legislation. A statute conceived as a permanent change in the tax code was thus enacted as a statute with a built in final year, and the final year was selected by the length of the budget window rather than by any judgment that the policy ought to end in 2010.

The design had consequences that unfolded across the following decade. Because the whole act shared one termination date, the expiration was not a series of small scheduled reviews but a single event at which the individual income tax would revert across nearly every dimension at once, rates, brackets, credits, and estate provisions together. Every intervening debate about extension took place against a deadline that Congress had set for procedural reasons in 2001. The act also established the pattern that later reconciliation tax bills followed: the sunset date became the most consequential single line in the statute for long range purposes, more consequential for the law as actually experienced than many of the rate schedules themselves, because the schedules only ever operated inside the span the sunset allowed. The 2001 act demonstrated that the fifth test does not merely constrain reconciliation bills at the moment of passage. It reaches forward and schedules the next decade of tax politics, fixing the year in which the argument must be held again.

The Jobs and Growth Tax Relief Reconciliation Act of 2003, Public Law 108-27, signed on May 28, 2003, answered the same constraint with a more differentiated architecture. The act accelerated rate reductions that the 2001 act had scheduled to phase in during later years, enlarged bonus depreciation and small business expensing for bounded periods, and reduced the tax rates on dividends and capital gains, taxing qualified dividends at the rates applicable to capital gains. Its drafters did not copy the single general sunset of 2001. The title containing the dividend and capital gains rate reductions carried its own termination date of December 31, 2008, earlier than the 2010 expiration that governed the accelerated individual rate provisions, which the act expressly tied to the sunset provisions of the 2001 act. Expensing and depreciation provisions ran on shorter schedules of their own. The result was a staggered set of expirations, in which different components of the combined 2001 and 2003 rate structure faced different deadlines, and later extension decisions arrived in pieces rather than in a single negotiation over the whole structure at once.

The 2003 amendment episode recounted in the preceding section, in which an amendment to end the offsets in a year beyond the window fell on a point of order, fixes the reciprocal logic of the mechanism, because the permanent revenue increases that a title uses to offset temporary reductions acquire a protected status under the title as a whole measurement, and the title survives the outyears only as long as the offsets remain in force for exactly as long as the losses they balance.

The Tax Cuts and Jobs Act of 2017, Public Law 115-97, enacted in December 2017, carried the mechanism to its most developed form. The act moved through reconciliation under a budget resolution for fiscal year 2018 whose instruction permitted a substantial increase in deficits inside the window while the fifth test continued to bar an increase beyond the window. The measure combined a comprehensive restructuring of individual taxation, a reduction and restructuring of corporate taxation, and a set of international provisions, all reported within a tax title whose combined effects the Joint Committee on Taxation and the Congressional Budget Office estimated year by year. Within that single title, the drafters separated the provisions into two classes with different lifespans, and the dividing line between the classes was drawn by the outyear arithmetic of the fifth test rather than by the economic character of the provisions themselves. Some provisions were written as permanent law. Others, of equal or greater importance to most taxpayers, were written to terminate on a date certain. The asymmetry between the two classes is the clearest demonstration in the statute book of how the fifth test shapes the lifespan of tax legislation.

On the expiring side of the 2017 act stood the core of its individual income tax provisions. The reductions in individual income tax rates were written to expire after December 31, 2025. The enlarged standard deduction was written to expire on the same schedule, as was the enlarged child tax credit. The deduction under section 199A for qualified business income from pass through businesses was written to terminate after December 31, 2025, as the text of Public Law 115-97 itself provides. The enlarged estate and gift tax exemption was likewise scheduled to expire after that date, returning the exemption toward its prior level as adjusted for inflation. The pattern is catalogued in full in the series guide to the Tax Cuts and Jobs Act of 2017, which sets out the act provision by provision. The shared termination date meant that the individual side of the 2017 act reproduced the structure of 2001: a broad body of law governing the taxation of households, enacted with a final year already printed in the text.

On the permanent side of the act stood its corporate centerpiece. The reduction of the corporate income tax rate to 21 percent, replacing a graduated rate structure with a single flat rate, was written without an expiration date. The asymmetry was possible because of the way the fifth test is applied. The test examines the reported title as a whole in each year beyond the window, and the tax title combined the permanent corporate reduction with permanent provisions that raised revenue, so that the title as measured did not increase the deficit beyond the window. The standard example of those permanent revenue raisers is the change in the method of adjusting the tax code for inflation: the act permanently replaced the traditional consumer price index with the chained consumer price index for indexing the individual rate brackets and other indexed amounts. Because the chained index rises more slowly than the index it replaced, bracket thresholds and indexed amounts grow more slowly year after year, producing a permanent and gradually increasing stream of revenue that operates in every outyear. That permanent gain, together with other permanent base broadening changes in the title, offset the permanent loss from the corporate rate reduction in the years beyond the window. The expiring individual provisions contributed nothing to those years, since their termination returned the law to its prior state before the window closed.

The counterfactual exposes the mechanism. Making the individual provisions permanent as well, rates, standard deduction, child tax credit, section 199A deduction, and estate exemption together, would have added a large permanent revenue loss in every year beyond the window, and the permanent revenue raisers in the title, the chained index change chief among them, were not large enough to cover both the corporate reduction and the individual reductions at once. The title would then have increased the deficit in fiscal years beyond the window, and its revenue reducing provisions would have been exposed to a point of order under section 313(b)(1)(E) that the majority could sustain only by waiver with sixty votes. The drafters therefore purchased permanence where the offsets could pay for it and accepted expiration where they could not. The division records no finding that individual rate relief deserved a shorter life than corporate rate relief. It records a measurement: the point at which the outyear ledger of the Finance Committee title ceased to balance. The 2025 date itself held no policy significance as a judgment about the proper duration of the provisions. It marked the outer edge of the budget window within which the instruction had authorized the deficit increase, translated into a termination date for the provisions the offsets could not carry.

The consequences of the earlier sunsets played out as dated history well before the 2017 act was drafted. The end of 2010 deadline in the 2001 act forced a negotiation in the final weeks of that year, because the alternative to agreement was a reversion of rates and credits across the individual tax system on January 1, 2011. The bargain that resulted was enacted as the Tax Relief, Unemployment Insurance Reauthorization, and Job Creation Act of 2010, Public Law 111-312, signed on December 17, 2010. The 2010 act extended the 2001 and 2003 rate structures for two additional years, through the end of 2012, coupled the extension with an extension of unemployment insurance benefits and a temporary reduction in the employee share of the payroll tax, and set estate tax terms for the extension period. The coupling visible in the title of the 2010 act was itself a product of the sunset: a deadline set in 2001 for procedural reasons assembled, nine years later, a coalition and a package that joined tax rates to unemployment benefits because both had to move before the same date or not at all.

The two year extension of 2010 merely relocated the cliff. As the end of 2012 approached, the extended 2001 and 2003 provisions stood scheduled to expire together on December 31, 2012, alongside other fiscal deadlines that combined to produce the negotiation known as the fiscal cliff. That negotiation was resolved, after the expiration date had passed, by the American Taxpayer Relief Act of 2012, Public Law 112-240, enacted on January 2, 2013. The Congressional Research Service overview of the act describes its central tax action as the permanent extension and modification of the 2001 and 2003 tax cuts, and that description captures the structural point. Most of the rate structure first enacted in 2001, and accelerated in 2003, was made permanent in 2013, with the rates applicable at the highest income levels allowed to return toward their earlier levels, and the permanence was achieved by regular legislation enacted after a lapse of two days, not by reconciliation. The provisions had spent more than a decade as temporary law, not because any Congress had judged a decade to be their proper span, but because the fifth test had required the 2001 drafters to print an expiration date, and each successive Congress had extended the date rather than settle the underlying question under regular order.

The lesson generalizes across the three statutes. Expiration dates in modern tax statutes are frequently procedural artifacts of the fifth Byrd test, chosen by the length of a budget window rather than by any judgment about when the policy ought to end. The 2010 date in the 2001 act, the 2008 date in the dividend and capital gains title of the 2003 act, and the date of December 31, 2025 written into the individual provisions of the 2017 act all encode the same information: the span of years a budget resolution covered, the size of the deficit increase an instruction authorized inside that span, and the quantity of permanent offset the drafters could assemble. None of the three dates encodes a conclusion that the underlying provision exhausted its usefulness in that year. A reader who treats a sunset as a policy judgment misreads the statute, and a reader who asks which test produced the date, what offsets stood in the same title, and how large the authorized window was reads the statute as its drafters were required to write it. The fifth test left the deepest marks on the law itself precisely because it so rarely struck provisions during debate: it instructed drafters, years in advance, to build the ending into the beginning.

The Adviser in the Room: Parliamentarian, Byrd Bath and the Overrule Question

The Byrd rule is a text, section 313 of the Congressional Budget and Impoundment Control Act, but the rule as senators experience it is a set of people in a sequence of rooms. A provision rarely dies in the manner the statute describes, struck on the floor after a point of order, a ruling from the chair and a failed motion to waive. Far more often a provision dies earlier and more quietly, in conference with an appointed staff officer whose advice carries no formal force, after arguments delivered by committee staff who hold no vote at all. Formally, the presiding officer rules, senators raise points of order, and sixty votes waive. In practice, the decisive examination happens before floor consideration begins, the examination is conducted by the Senate parliamentarian, and the judgment that emerges shapes the bill that reaches the floor so thoroughly that the formal stages often ratify a result settled in advance. A reader who understands only the six tests understands what the rule prohibits; a reader who understands the Byrd bath, the office and the one modern replacement understands who decides, and why that decision holds.

The pre-floor examination acquired a colloquial name in Senate staff usage, the Byrd bath, a term the Congressional Research Service records in its report on the Byrd rule as the label for the preemptive scrutiny of reconciliation legislation before floor consideration. No named coiner or dated first use anchors the term in the sources consulted for the rule, and the honest description is the modest one: staff who live through the process use the phrase, and its origin rests in oral staff culture rather than in any published coinage. A reconciliation bill enters the process carrying provisions accumulated through committee drafting, leadership negotiation and constituency pressure, and the bill emerges with some of that matter washed away, removed not by vote or by amendment but by the managers of the bill themselves after an advisory judgment that the matter would not survive.

The process begins while the bill remains in draft or shortly after a committee reports it, where the text is complete enough to be judged provision by provision but open enough to be changed without a floor fight. The participants assemble around the parliamentarian and the senior staff of that office. The majority side sends staff from the committees that received reconciliation instructions and from the Senate Budget Committee, whose chairman bears responsibility under the Budget Act for determinations about budgetary effects. The minority side sends parallel staff from the same committees, because the scrutiny is adversarial by design and minority staff hold the strongest incentive to identify extraneous matter that a point of order could later strike. Senators themselves rarely conduct the argument. The advocacy belongs to staff counsel and budget analysts who drafted the language under examination and who know, line by line, what each provision was written to do.

Argument in the bath proceeds provision by provision, and each provision draws the same structured contest. Minority staff, or majority staff anticipating the attack, identify the test the provision allegedly fails, most frequently the test that asks whether the budgetary effects are merely incidental to the nonbudgetary components. Majority staff answer with the provision’s fiscal mechanics, walking through the outlay and revenue effects as the Congressional Budget Office and, for revenue provisions, the Joint Committee on Taxation have estimated them. The estimates matter as evidence rather than as background. A provision with a substantial scored effect presents a different case from a provision whose scored effect is trivial beside its regulatory reach, and the size, direction and timing of the estimate feed into the incidental analysis, the instruction compliance analysis and the analysis of deficit effects beyond the window covered by the budget resolution. Precedent enters alongside the numbers. Staff cite earlier rulings and advisory judgments on comparable language, and the parliamentarian’s office, which compiles the precedents of the Senate in Riddick’s Senate Procedure, tests each citation for closeness of fit.

The parliamentarian signals rather than decides. In accounts given by participants in the 2017 health care cycle, the two sides presented arguments separately and together, the parliamentarian heard the assembled staff, and the judgment came back after deliberation rather than from the chair in the room. A signal could be firm, identifying a provision as a violation that a point of order would very likely strike. A signal could be provisional, identifying a provision as exposed, dependent on a close question of precedent or on whether a revised estimate changed the balance between budgetary and nonbudgetary effects. A signal could also clear a provision, and clearance carried its own consequence, because language that survived the bath entered floor consideration with an advisory judgment behind it that any senator raising a point of order would have to overcome in the ruling of the chair.

A provision dies at that stage through the decision of its own managers. Faced with a firm adverse signal, the majority staff who carried the provision weigh three options, and each explains why the quiet death is the common outcome. Striking the provision in the bath costs nothing on the floor, consumes none of the limited debate time reconciliation allows, and avoids a recorded vote in which the majority either loses the provision publicly or spends sixty vote waiver capital the majority does not possess. Redrafting offers a second path, narrowing regulatory reach or enlarging genuine budgetary effect until the advisory judgment changes. Carrying the provision to the floor despite the signal remains available, but the carrier accepts that a point of order will likely strike the language and that the ensuing waiver vote will advertise the shortfall. Because the first two options dominate, the Congressional Research Service observes that when senators respond to guidance by removing or redrafting questionable matter before floor consideration, no formal precedent is created and the reasoning never enters a published record. The bath thus produces outcomes without opinions, and interpretive practice grows in the files and memory of the parliamentarian’s office rather than in any reporter a reader can consult.

The Congressional Research Service, in its report on the office of the parliamentarian in the House and Senate, describes the Senate parliamentarian and the assistants in that office as charged with providing information and assistance to all senators and their staffs on a strictly nonpartisan and confidential basis, with the office situated within the office of the Secretary of the Senate. A companion Congressional Research Service report on offices and officials in the Senate states the appointment mechanism directly: the Senate parliamentarian is appointed by the Secretary of the Senate with majority leader approval. The Secretary, in turn, is elected by the Senate at the recommendation of the majority leader. No fixed term protects the holder, and the history of the office demonstrates that a parliamentarian serves only so long as the leadership of the chamber permits the service to continue.

Within that design the division of labor is exact. The parliamentarian advises and the presiding officer rules. The Congressional Research Service states the point without qualification in its parliamentarian report: as a staff official, the parliamentarian is empowered to make no decision binding on the Senate, and the advice offered to the presiding senator may be accepted or rejected, with rulings subject to appeal by individual senators. Off the floor the office confers with senators and staff in advance of the questions that floor sessions present. On Byrd rule questions the advisory character of the office coexists with a consistent practice: the presiding officer, advised by the parliamentarian, determines each alleged violation, and through the rule’s history to the vantage of March 2020 the chair’s determinations on the rule followed the advice the office gave. The advice was formally a recommendation. Functionally, over successive cycles under majorities of both parties, the recommendation settled the question.

Robert Dove tested the limits of that arrangement from inside the office, twice. Dove joined the parliamentarian’s office in 1966 and became parliamentarian in 1981, serving from 1981 to 1987. His first service ended in dismissal, when the Democratic Party obtained a Senate majority in the elections of 1986 and Democratic Majority Leader Robert Byrd dismissed him in 1987, replacing him with Alan Frumin, the senior assistant in the office. Dove returned as parliamentarian in 1995, when a Republican majority restored him and Frumin returned to the position of top assistant. The second service ran from 1995 to 2001, and its ending supplied the episode through which every later discussion of the replacement power proceeds. During Senate work on the budget and tax reconciliation measures of 2001, in a chamber divided evenly between the parties, Dove delivered advice that frustrated the Republican majority leadership. Contemporaneous reporting, summarized in later accounts drawing on The Washington Post, described rulings that made advancing President George W. Bush’s budget and tax agenda more difficult, among them a determination that Senate rules allowed only one revenue reconciliation bill in a year to receive protection against unlimited debate and advice against a Republican provision allocating funds in the 2002 budget for natural disaster assistance. In May 2001 Majority Leader Trent Lott dismissed Dove, and Frumin again succeeded him, becoming the first person to become parliamentarian under majorities of both parties. Senator Russell Feingold objected on the floor in May 2001, as the Congressional Record of that month preserves, protesting the summary firing of the parliamentarian because of advice on budget related issues. The objection changed nothing. Frumin took the advisory seat, and the tax reconciliation legislation of 2001 proceeded, its exposure to deficit increases beyond the budget window answered by sunset provisions rather than by any confrontation over the advice.

The Dove episode establishes two propositions at once. Replacement is real. A majority leader willing to absorb the institutional criticism that followed the dismissals of 1987 and 2001 could remove the adviser whose advice obstructed the majority’s program, and the removal required no vote, no hearing and no cause stated in any public instrument. Replacement is also bounded. The dismissal of May 2001 changed the person delivering the advice and changed nothing else. Section 313 stood unamended, the precedents compiled under it stood as compiled, and the successor inherited the same statutory text, the same body of prior rulings and the same obligation, embedded in the office’s nonpartisan charge, to read both as the record supported. Frumin’s subsequent retention through successive changes of party control, in June 2001, in January 2003 and in January 2007, underscores the point: the officer installed by a frustrated Republican majority in 2001 served afterward under a Democratic majority without removal. A parliamentarian can be replaced. An office, with its precedents and its memory of every prior bath, replaces far less easily.

Elizabeth MacDonough’s tenure carried the office across another change of party control and through the reconciliation cycles that dominate the pre 2020 record. MacDonough joined the parliamentarian’s office in May 1999 as an assistant parliamentarian and rose to senior assistant parliamentarian in 2002. She assumed the office of parliamentarian on February 2, 2012, succeeding Frumin, promoted under Democratic Majority Leader Harry Reid, and she was the first woman to hold the position, a fact recorded in the Politico account of her appointment in February 2012 and repeated in summaries of the office thereafter. Appointment under a Democratic majority settled nothing about her subsequent service. When Republicans took the Senate majority in January 2015, Majority Leader Mitch McConnell retained her, and the reconciliation cycles of 2015 and 2017 placed her advice at the center of a Republican program she had no part in writing.

The substance of those cycles belongs to the earlier sections of the article, and the institutional point belongs here: her advice cut in both directions within the same bills, and the majority absorbed the cuts because the same advice delivered the permissions the majority depended on. In the health care cycle of 2017, guidance issued on July 21, 2017 advised that central provisions of the Better Care Reconciliation Act would require sixty votes, among them the provision defunding Planned Parenthood for a period, the restrictions on tax credits for plans covering abortion, and the six month waiting period after a lapse in continuous coverage, with further guidance later in July 2017 advising against the state waiver provision as well. In the same cycle the office advised that the House passed American Health Care Act of 2017 complied with the reconciliation process, the determination that allowed the repeal effort to proceed under privilege at all, and advised that states could impose work requirements for Medicaid recipients within the process. In December 2017, during consideration of the conference report to the Tax Cuts and Jobs Act, three Byrd rule violations were sustained on December 19, 2017, including the provision allowing education savings accounts to pay homeschooling expenses and the exemption from the endowment tax for colleges with fewer than five hundred tuition paying students, and the bill returned to the House minus the stricken matter. A parliamentarian appointed by one party’s leader had, within a single year, preserved the procedural vehicle of the other party’s program, sustained provisions that party valued and removed provisions that party had campaigned on. Retention across the 2015 change of control and acceptance across the 2017 cycles form the strongest evidence the pre 2020 record offers that the office operated as an office, applying precedent to whichever majority stood before it.

Can the presiding officer ignore the parliamentarian?

Yes, in formal terms. The parliamentarian advises, the presiding officer rules, and no statute binds the chair to accept the advice given. A chair ruling against that advice on a Byrd rule point of order would stand unless sixty senators voted to sustain an appeal against the ruling of the chair.

The formal answer opens the genuine dispute, because the Senate that wrote the sixty vote appeal threshold into section 313 also built, around the advisory office, a practice in which the formal option went unexercised on the rule through March 15, 2020. The case for treating the option as a legitimate instrument of majority authority begins with the constitutional allocation of procedural power. The presiding officer is an officer of an elected chamber, in many reconciliation debates the Vice President or a senator answerable to colleagues and voters, while the parliamentarian is appointed staff, selected through the Secretary of the Senate with majority leader approval and accountable to no electorate. On that allocation, advice that settles the content of legislation inverts the intended order of authority, since a delegation of interpretive labor to an expert office cannot silently become a transfer of the decision. The text of section 313 supports the reading at every relevant point. The statute directs the presiding officer to rule, makes the ruling appealable, and sets the waiver and appeal thresholds at three fifths of senators duly chosen and sworn. A majority that cannot assemble sixty votes to waive a point of order has therefore not exhausted the statute’s own mechanisms when the chair has simply repeated advice the majority believes misreads the statute. The appeal threshold answers the objection that a contrary ruling would be lawless: the ruling would be tested immediately, in the open, by the most demanding vote the Budget Act imposes, and the series guide to the filibuster and cloture sets out the supermajority logic that the sixty vote figure carries across Senate procedure. Declining to use the option, on that reading, is a choice majorities make for reasons of prudence or institutional taste, and a choice cannot harden into a legal disability merely through repetition. Senate procedure treats rulings of the chair, sustained on appeal, as the mechanism by which the chamber itself declares what its rules mean, and a chamber that never rules against advice has allowed an appointive office to accumulate an interpretive monopoly the statute never granted.

The case against begins where the first case ends, with the observation that a precedent created by overruling differs in kind from a precedent created by waiver. The Congressional Research Service draws exactly that distinction in its report on the Byrd rule: a successful motion to waive recognizes that the matter violates the rule and allows it anyway, while a successful appeal establishes new precedent and changes the interpretation of what the Senate considers extraneous. A waiver is an exception, purchased at the price of sixty votes, that leaves the rule intact for the next bill and the next majority. A chair ruling against advice, sustained because the minority cannot reach sixty votes to reverse it, rewrites the meaning of merely incidental or of budgetary effect itself, and the rewritten meaning remains available to every future majority, including majorities of the opposite party pursuing programs the first majority would oppose. The drafting discipline the rule exerts depends on that distinction. The Byrd bath disciplines legislation precisely because drafters believe the advisory judgment forecasts a floor ruling with near certainty, and staff trim, redesign and abandon provisions months before any vote in reliance on that belief. One demonstrated overruling converts the forecast into a guess and the rule from a constraint majorities obey into a courtesy majorities extend to themselves when obedience costs little. On that reading the sixty vote waiver is the designed democratic escape, the route the statute provides for a majority convinced that a particular provision matters more than the discipline, and the price is set high deliberately so that exceptions remain exceptional. The historical record through the vantage of March 2020 gives the second reading its factual spine. Majorities of both parties absorbed adverse advice across the rule’s history, amending bills in 2017 within days of guidance that removed prized provisions, accepting the December 19, 2017 striking of tax provisions rather than contesting the chair, and the one remedy the institution used against an adviser was replacement of the adviser, in 1987 and again in May 2001, a remedy that left rule and precedent untouched. Pressure existed, in the published arguments of 2017 urging the chair to disregard advice on health care and in the frustration that preceded the 2001 dismissal, and the refusal recurred with the pressure. The two readings stand unreconciled because each rests on a true premise: the authority to rule belongs to the chair, and the rule survives only because the chair has declined, through March 2020, to rule against the adviser the chamber appointed to say what the rule requires.

Reading a Reconciliation Bill Through the Byrd Rule

A reconciliation bill yields its procedural secrets to a reader who interrogates the text in statutory order, because each test in section 313 corresponds to a question a drafter asks of any provision before the provision earns a place in the vehicle. The first question concerns money in the most basic sense: does the provision move money at all. A drafter asks whether the provision changes outlays or revenues, directly or through the terms and conditions under which outlays are made or revenues are collected, and whether an estimate from the Congressional Budget Office or the Joint Committee on Taxation records any change at all. A provision whose removal leaves an estimate untouched, however significant the statement the provision makes, carries no budgetary claim to expedited consideration, and the December 2017 striking of a short title, the name of an enacted tax act, demonstrated that the question admits no exception for importance.

The second and third questions concern the reporting committee, because reconciliation authority arrives addressed to named committees and reaches no further. A drafter asks whether the reporting committee sits inside the instruction the budget resolution delivered, measured title by title and year by year against the figures the resolution printed, and whether the provision lies inside the jurisdiction the standing rules assign to that committee. A title that misses a single named year exposes every expensive provision inside the title, while savings provisions survive untouched, and a provision belonging to a committee that received no instruction has no route into the bill at all, whether the provision arrives in a reported text or in a floor amendment. Compliance and jurisdiction are documentary questions, and a drafter who answers both from the resolution and from Rule XXV before drafting begins removes the two defects cheapest to cure and hardest to excuse.

The fourth question is the judgment question, and a drafter asks the question in the form the accumulated advice gives the inquiry: would the money matter if the policy vanished. Equivalently, a drafter asks whether the provision operates through its budgetary effect or merely casts a fiscal shadow from a decision taken on regulatory or social grounds. A provision that sets a rate, adjusts a penalty amount, defines eligibility for a payment, or sells access to a federal asset works through the budget, and the fiscal effect is the mechanism rather than an accompaniment. A provision that announces a rule of conduct and relies on downstream consequences for a score invites the comparison the fourth test orders, and agreement among participants that a provision matters greatly becomes evidence against the provision when the agreed importance is nonbudgetary. The fifth question concerns time: does the title balance in every year past the window. A drafter reads the title as a whole into the years beyond the final year the budget resolution covers, confirms that no single outyear shows a deficit increase, and answers any imbalance with an offset inside the same title and the same year or with an expiration date inside the window. The sixth question is categorical: does the provision touch Title II of the Social Security Act. A recommendation with respect to the old age, survivors, and disability insurance program fails without regard to size or merit, while Medicare, Medicaid, and the other titles of the same statute remain open to reconciliation drafting as a matter of routine.

The diagnostic explains the distribution of outcomes across the rule’s history, because the questions differ in the kind of answer each admits. Documentary questions settle in committee rooms, arithmetic questions settle in drafting through sunsets and offsets, and the categorical question settles before drafting begins, through exclusion of a subject from the vehicle altogether. The judgment question settles last and settles hardest, in argument before the parliamentarian during the Byrd bath, where defenders describe a budget measure with policy consequences and challengers describe a policy measure with budget consequences and the statute supplies no exchange rate between the descriptions. One undefined phrase governs American lawmaking, and the phrase governs precisely because every other question in the diagnostic can be answered by computation, leaving the character of a provision as the ground on which the practical boundary of majority legislative power gets drawn, cycle after cycle, by an appointed adviser applying precedent to text.

The six-test table collects each test, the matter the test excludes, a documented example from the record assembled in the preceding sections, and the pattern of invocation the record shows.

Test	What it excludes	Documented example	How often it is invoked
Test one, no change in outlays or revenues	Provisions that produce no change in outlays or revenues	The short title provision of the 2017 tax act, struck on December 19, 2017 after a waiver motion failed	Invoked often and the most frequently recorded ground in points of order
Test two, committee not in compliance with its instruction	Provisions increasing outlays or decreasing revenues inside a title whose committee failed its instruction	Medicaid spending struck from the 1996 reconciliation bill on July 18, 1996 after the Finance Committee submission missed its fiscal 2002 savings figure	Rare, with a single prominent strike recorded across the rule’s first three decades in Congressional Research Service accounts
Test three, outside the reporting committee’s jurisdiction	Provisions outside the jurisdiction of the committee that submitted the title or provision	The Manchin amendment of December 3, 2015, a Judiciary Committee subject offered to a bill instructed only to the Finance and the Health, Education, Labor, and Pensions Committees, with a waiver motion that failed 48 to 50	Invoked regularly, most often against floor amendments
Test four, merely incidental	Provisions whose changes in outlays or revenues are merely incidental to the nonbudgetary components	The one year bar on federal payments to Planned Parenthood in the 2017 health care reconciliation bill, advised extraneous in July 2017	The ground of nearly every genuinely contested dispute
Test five, deficit increase beyond the window	Provisions increasing the deficit in a fiscal year beyond the years covered by the reconciliation measure without an offset in the same title and year	The Sessions amendment of 2003, whose offsets were set to sunset in fiscal year 2015, beyond the fiscal 2003 to 2013 window, and which fell on a point of order	Answered in drafting through sunsets and offsets more often than contested on the floor
Test six, Social Security	Provisions recommending changes in the old age, survivors, and disability insurance program under Title II of the Social Security Act	Language amending Title II provisions on the use and collection of Social Security numbers and eliminating certain Title II reports, struck by point of order in 2015 before the language could be carried	Almost never invoked, because drafting stage deterrence operates before any bill is reported

Read against the table, the episodes of the preceding sections sort into a pattern a reader can carry to any later reconciliation text. Mechanical grounds dispose of names, findings, and decorative statements without argument about merit. Instruction and jurisdiction grounds dispose of provisions whose defect lies in the relationship between a committee and a resolution rather than in the provision alone. The window ground appears in the statute book as printed expiration dates rather than in floor records, and the Social Security ground appears as absence, the restructuring that no reconciliation bill between the adoption of the rule and the end of 2019 attempted to carry. The incidental ground alone generates a literature of argument, because the ground requires a characterization of purpose that estimates cannot supply and that precedent supplies only by analogy from one provision to the next.

A reader tracing the rule into later reconciliation drafting can turn to the series’ later reconciliation case study, framed as later series coverage of provisions shaped by this rule, and can test each provision in that coverage against the diagnostic in the order the statute presents the questions. The exercise rewards a particular discipline: begin with the estimate, proceed to the instruction and the jurisdictional assignment, reserve the characterization question for the provision that survives each documentary test, and treat every expiration date in the text as evidence about the window and the offsets rather than as a judgment about the provision. A reader who works in that order reads a reconciliation bill the way the Senate reads the bill, through a definition that decides what a majority may carry before the majority votes on whether carrying the provision is wise.

The six tests reward organized study, because each test pairs a statutory sentence with a body of example that no summary captures completely. A student preparing for coursework or an examination can keep your statute notes, citations, and case chronologies together free on VaultBook while tracing a provision from instruction to estimate to ruling, and can practice and revise US government and civics material on ReportMedic to fix the distinctions that decide outcomes, the distinction between a waiver and an appeal, between advice and a ruling, and between a provision struck from a bill and a bill removed from the floor. Mastery of the rule consists of those distinctions held together, applied in order, to text that changes with every budget cycle while the definition that judges the text remains the six sentences Congress enacted and left, in the decisive instance, undefined.

Frequently Asked Questions

Q: What does the Byrd rule prohibit?

The Byrd rule prohibits nothing in the manner of a criminal ban, and the precision matters for understanding enforcement. Section 313 of the Congressional Budget Act defines six categories of extraneous matter in a reconciliation bill, a reconciliation resolution, an amendment, or a conference report, and makes matter in those categories subject to a point of order in the Senate. Extraneous matter may be drafted, reported by a committee, and assembled into an omnibus bill, and removal occurs only when a senator raises a point of order and the presiding officer sustains the objection. A sustained objection strikes the offending provision or defeats the amendment carrying the provision, while the remainder of the measure continues in consideration. The Senate may also waive the point of order by a three fifths vote, so the rule operates as a procedural vulnerability attached to defined categories of matter rather than as an absolute legal bar on content.

Q: What are the six tests in the Byrd rule?

Section 313(b)(1) defines extraneous matter through six tests, and a provision failing any one test fails altogether. The first test reaches a provision producing no change in outlays or revenues. The second reaches a provision increasing outlays or decreasing revenues inside a title whose reporting committee failed to achieve an instruction. The third reaches a provision outside the jurisdiction of the committee that submitted the provision. The fourth reaches a provision whose changes in outlays or revenues are merely incidental to the nonbudgetary components of the provision. The fifth reaches a provision increasing the deficit in a fiscal year beyond the years covered by the reconciliation measure without an offset in the same title and year. The sixth reaches a provision recommending changes in the old age, survivors, and disability insurance program established under Title II of the Social Security Act.

Q: What does merely incidental mean in the Byrd rule?

The statute supplies no definition of merely incidental, no ratio between budgetary and nonbudgetary effects, and no dollar threshold, so the meaning comes from accumulated advice applying the phrase provision by provision. In practice, the test asks whether the money is the point of the provision or only a side effect of the point. A provision passes when changing spending or revenue is the mechanism that does the work, as with a rate adjustment or a leasing program that sells access to a federal asset. A provision fails when the fiscal change only accompanies a decision made for regulatory or social reasons, even where the scored effect is large. Advice in 2015 treated repeal of a nationwide coverage requirement as extraneous despite a saving estimated near 147 billion dollars over ten years, because the requirement reached approximately 270 million people as a rule of conduct.

Q: Why do tax cuts expire because of the Byrd rule?

Tax cuts enacted through reconciliation expire because the fifth test in section 313(b)(1)(E) treats a provision as extraneous when the title containing the provision increases the deficit in any fiscal year beyond the years covered by the budget resolution, unless another provision in the same title offsets the increase in the same year. A permanent revenue reduction loses money in every outyear, so drafters hold three options: assemble permanent offsets, obtain sixty votes for a waiver, or end the reduction inside the window and let prior law revive afterward. Majorities commanding fifty-one votes rather than sixty ordinarily choose the sunset. The act of 2001 expired at the end of 2010, the dividend and capital gains title of the act of 2003 expired at the end of 2008, and the individual provisions of the act of 2017 were written to terminate after December 31, 2025.

Q: Are retirement benefits excluded by the Byrd rule?

The exclusion is narrower than a general bar on retirement benefits, because the sixth test reaches one program defined by one title of one statute. Section 313(b)(1)(F) makes a provision extraneous when the provision violates section 310(g), which bars reconciliation recommendations with respect to the old age, survivors, and disability insurance program established under Title II of the Social Security Act of 1935. Programs under other titles of the same act fall outside the bar, and Congressional Research Service accounts name Medicare under Title XVIII, Medicaid under Title XIX, unemployment insurance under Title III, and the State Children’s Health Insurance Program under Title XXI as programs amended through reconciliation without violating the definition. The bar also reached administrative language in 2015, when language on the use and collection of Social Security numbers and on Title II reports was struck, so breadth inside Title II coexists with a boundary at the edge of Title II.

Q: How many votes waive the Byrd rule?

A waiver requires the affirmative vote of three fifths of the senators duly chosen and sworn, which is sixty votes in a Senate with no more than one vacancy. Section 904 of the Congressional Budget Act supplies the threshold for a motion to waive a Byrd rule point of order, and the same three fifths requirement governs an appeal of the ruling of the presiding officer on the point of order. The denominator is the full membership rather than the senators present and voting, so absences raise the effective burden on senators in the chamber. A point of order under the rule is not debatable, so a waiver motion must be made while the point of order is pending and before the presiding officer rules, although a waiver may also be offered preemptively and may cover some or all of the provisions named in a single point of order.

Q: Who advises on the Byrd rule?

The Senate parliamentarian advises on the Byrd rule, and the presiding officer rules. The parliamentarian is a staff official appointed by the Secretary of the Senate with majority leader approval, charged with providing information and assistance to all senators and staff on a strictly nonpartisan and confidential basis, and empowered to make no decision binding on the Senate. Advice operates at two stages. During the pre-floor review called the Byrd bath, majority and minority committee staff argue provision by provision before the parliamentarian, who signals whether a provision appears extraneous under one of the six tests. On the floor, the presiding officer determines each alleged violation with the advice of the parliamentarian, and through the history of the rule to March 2020 the determinations of the chair on the rule followed the advice the office gave, under majorities of both parties and across changes in party control.

Q: Can the parliamentarian be overruled on the Byrd rule?

Formally, the advice of the parliamentarian binds no one, because the parliamentarian advises and the presiding officer rules, and a ruling of the chair stands unless an appeal succeeds. An appeal under the Byrd rule requires three fifths of the senators duly chosen and sworn, so a ruling against the advice would stand unless sixty senators voted to sustain an appeal against the ruling. Through March 15, 2020, no presiding officer overruled the advice on the rule, and majorities of both parties absorbed adverse advice by amending or abandoning provisions instead. The one remedy the institution used against an adviser was replacement, in 1987 and again in May 2001, when Majority Leader Trent Lott dismissed Robert Dove after advice that frustrated the majority. Replacement changed the person delivering the advice and left the statute, the precedents, and the successor’s obligations intact.

Q: Does the Byrd rule apply in the House of Representatives?

The Byrd rule applies only in the Senate, and no equivalent provision governs House consideration of reconciliation bills. Section 313 sits inside the Congressional Budget Act as a definition of extraneous matter joined to a Senate point of order, a Senate waiver threshold of three fifths of the senators duly chosen and sworn, and a Senate conference procedure under section 313(d). House consideration operates under different institutional conditions, where the rules of the House on germaneness and special orders from the Rules Committee perform different work. The Senate only reach of the rule produces a practical consequence at the conference stage, because matter accepted by House conferees may face a point of order when the conference report reaches the Senate, and a sustained objection sends an amended text back to the House for further action, as the December 2017 tax conference demonstrated.

Q: What is a Byrd bath?

A Byrd bath is the pre-floor review in which a reconciliation bill receives provision by provision scrutiny before Senate consideration begins, a label from Senate staff usage that the Congressional Research Service records in accounts of the rule. Majority and minority staff from the instructed committees and the Budget Committee assemble around the parliamentarian and the senior staff of that office, and argument proceeds provision by provision. Challengers identify the test a provision allegedly fails, most frequently the merely incidental test, and defenders answer with fiscal mechanics drawn from estimates by the Congressional Budget Office and the Joint Committee on Taxation, together with precedent from earlier rulings. The parliamentarian signals rather than decides, and managers respond to a firm adverse signal by removing or redrafting the provision, so most matter that fails the rule dies in that room without a vote, a ruling, or a published record of the reasoning.

Q: What happens to a provision that is struck under the Byrd rule?

A sustained point of order removes the provision and leaves the remainder of the measure in consideration, a surgical effect that distinguishes the rule from a vote against a bill. Under section 313(a), the part of the title or provision containing the extraneous material is deemed struck from the bill, and stricken matter may not be offered again as an amendment from the floor. Where the point of order targets an amendment, the amendment falls and the underlying bill proceeds without the amendment. The presiding officer may sustain a point of order as to some provisions named in the objection and not others, because section 313(e) permits a single point of order to reach several provisions with a separate ruling on each. At the conference stage, a sustained objection strikes the offending matter and sends a substitute text back to the House for further consideration.

Q: Who was Robert Byrd, and why is the rule named for him?

Robert C. Byrd was a senator from West Virginia whose long career included a sustained defense of the institutional character of the Senate, including extended debate as part of the legislative test a controversial proposal had to pass. The rule carries the senator’s name because Byrd brought the objection to the floor on October 24, 1985, during consideration of the Consolidated Omnibus Budget Reconciliation Act of 1985, offering Amendment No. 878 to S. 1730 on behalf of himself and other senators. Byrd counted 122 items in the pending bill as extraneous and argued that such provisions provoked controversy without aiding deficit reduction and consumed the limited debate time reconciliation allotted. The Senate adopted the amendment by a vote of 96 to 0, and the proposal entered law in modified form as section 20001 of the 1985 act before permanence arrived in 1990.

Q: Does the Byrd rule apply to amendments as well as to bills?

The rule reaches amendments as well as reported bill text, and floor amendments form a large share of the documented applications. A point of order may target matter in a reconciliation bill as reported, an amendment offered during floor consideration, a conference report, or an amendment between the two chambers. The jurisdiction test operates with particular force against amendments, because an amendment reaching a subject assigned to a committee that received no instruction in that cycle has no route into the bill, and Senate consideration of the Health Care and Education Reconciliation Act of 2010 produced nine amendments that fell after points of order were sustained, all but one on jurisdictional grounds. The Manchin amendment of December 3, 2015 fell on the same ground after a waiver motion failed 48 to 50, and a sustained objection defeats the amendment while the underlying bill proceeds.

Q: What role do Congressional Budget Office estimates play in Byrd rule decisions?

Estimates supply the evidentiary foundation for most Byrd rule determinations, because four of the six tests turn on figures an estimate records. The Congressional Budget Office estimates outlay effects and the Joint Committee on Taxation estimates revenue effects, and determinations of budgetary levels for enforcing the rule are made by the Senate Budget Committee working from those estimates. The first test asks whether an estimate records any change in outlays or revenues at all. The second compares a title against an instruction using scored effects, and the fifth asks whether a title increases the deficit in a year beyond the window, a question the Congressional Budget Office may answer with a statement about outyear effects rather than point estimates past the window. Estimates also feed the merely incidental comparison, where the size, direction, and timing of a scored effect stand against the unquantified reach of a provision’s policy change.

Q: Why does Social Security have its own test in the Byrd rule?

Social Security holds a separate test because Congress placed the program outside the reconciliation vehicle through a freestanding bar that the sixth test complements. Section 310(g) of the Congressional Budget Act provides that consideration of a reconciliation bill, amendment, or conference report containing recommendations with respect to the old age, survivors, and disability insurance program under Title II of the Social Security Act is not in order in either house. Section 313(b)(1)(F) then defines a provision violating section 310(g) as extraneous matter, giving a senator two distinct points of order against Title II language. The pairing carries a practical distinction in consequence: a Byrd rule objection strikes only the offending matter and lets the measure proceed, while a sustained objection under section 310(g) is fatal to the overall measure and removes the bill from the floor immediately.

Q: What is the difference between waiving the Byrd rule and appealing a ruling under it?

A waiver and an appeal reach opposite conclusions about the provision under challenge, and the difference governs the future of the rule itself. A successful motion to waive recognizes that the matter violates the Byrd rule and records a decision by the Senate to allow the matter regardless, leaving the interpretation of extraneous matter intact for the next bill. A successful appeal asks the Senate to hold that the presiding officer decided the question incorrectly, and a sustained appeal establishes new precedent, changing the interpretation of what the Senate considers extraneous for every later measure. Both routes require the affirmative vote of three fifths of the senators duly chosen and sworn under section 904. A majority holding sixty votes that values a provision more than a precedent therefore waives, while a majority appealing and prevailing rewrites the boundary the next majority inherits.

Q: Did the Byrd rule change the text of the tax act before it passed?

The rule changed the text of the tax reconciliation measure of 2017 at the conference stage, after both houses had acted on a conference report. The conference report on H.R. 1 carried a provision stating the short title by which the act was to be cited, together with language allowing section 529 education savings accounts to fund homeschooling expenses and an exemption from the endowment tax for a narrow class of colleges. The parliamentarian advised that the provisions violated the rule, Senator Bernie Sanders raised points of order on December 19, 2017, a motion to waive failed 51 to 48, and the points of order were sustained. The Senate proceeded to a substitute text without the stricken matter, the House voted again on December 20, 2017 on the corrected text, and the enacted statute, Public Law 115-97, carries no short title at all as a result.

Q: Can a provision removed during a Byrd bath be restored later?

A provision removed during a Byrd bath disappears through a decision of the managers who carried the provision, rather than through a ruling, so no formal bar attaches to the text at that stage. Managers weigh removal against redrafting and against carrying the provision to the floor despite an adverse signal, and redrafting that narrows regulatory reach or enlarges genuine budgetary effect can change the advisory judgment on revised language. A provision struck on the floor after a sustained point of order stands in a different position, because section 313(a) provides that stricken matter may not be offered again as an amendment from the floor during consideration of that measure. Restoration in a later measure remains a separate question governed by the tests applied to the later text, and matter excluded from reconciliation retains every route available under regular Senate procedure, where extended debate and a cloture threshold apply.

Q: How is the Byrd rule different from the filibuster?

The filibuster and the Byrd rule govern different questions inside the same chamber. The filibuster is the practice of extended debate on ordinary Senate legislation, under which a final vote ordinarily requires cloture by three fifths of the senators duly chosen and sworn. Reconciliation legislation travels under a twenty hour limit on debate, so a reconciliation bill cannot be filibustered in the manner available against ordinary legislation and a simple majority can bring the bill to passage after the allotted debate expires. The Byrd rule polices the content that may claim that protection, by defining extraneous matter and exposing such matter to a point of order during Senate consideration. The two devices therefore interact as a boundary and a bypass: extended debate guards the ordinary road, reconciliation supplies an expedited road around the guard, and the Byrd rule decides which provisions qualify for the expedited road.

Q: What does the Byrd rule do to a conference report?

Section 313(d) applies the Byrd rule during consideration of a conference report and during the exchange of amendments between the houses, reaching five of the six definitions of extraneous matter and omitting the definition based on committee jurisdiction in section 313(b)(1)(C). A conference report reaches the Senate as an unamendable package, so the procedure differs from a simple strike. If a point of order against a conference report is made and sustained, the offending matter is deemed stricken, and the Senate proceeds without intervening action or motion to consider an amendment in the nature of a substitute consisting of the conference agreement minus the stricken provisions. Agreement to that substitute sends the measure back to the House for further consideration. The tax conference report of December 2017 followed that sequence after three violations were sustained, and the House adopted the amended text before enactment.