Every law that Congress enacts begins as two acts of approval, one in the House of Representatives and one in the Senate, and the Constitution counts the work as finished only when the two approvals attach to exactly the same words. A conference committee is the formal instrument Congress built for the frequent case in which the two chambers approve different words: a temporary joint committee, drawn from both chambers, that negotiates a single text and returns it to each floor for an unamendable vote. When the instrument functioned as designed, the final bargain of a major measure took written form as a conference report, signed by a majority of the managers of each House and explained in a joint statement that later readers could consult.

The authority under examination is procedural rather than statutory. Article I of the Constitution supplies the destination, the requirement that the House and the Senate approve the same bill in precisely the same form before presentment to the President. The machinery belongs to the chambers: House Rule XXII governs conferences and amendments between the houses in the House of Representatives, and Senate Rule XXVIII performs the same office in the Senate. The Congressional Research Service reports cited throughout the account, including Report RL34611 on resolving legislative differences and Report RS20454 on going to conference in the Senate, supply the procedural record on which the account rests.

The stakes of the mechanism are easiest to see in what each route leaves behind. Acceptance leaves a vote and nothing else. Amendment exchange leaves a trail of messages and countermessages scattered across two chamber records. A conference leaves a single negotiated document, a conference report, whose text and explanation can be cited by agencies, litigants, and courts. The choice among the routes therefore decides more than the wording of a statute. The choice decides who holds a seat when the final wording is settled, whose objections carry a procedural price, and what evidence of the bargain survives enactment.

The counts give the change its scale. Counts compiled from House and Senate records in an analysis by the Center for Congressional and Presidential Studies at American University recorded 38 formal conferences in the 106th Congress, covering 1999 to 2000, and 9 in the 110th Congress, covering 2007 to 2008. A fall of that size within a single decade, in an institution that had relied on conferences for its most complex measures across the 1970s and 1980s, calls for a mechanical explanation rather than a cultural one, and Senate procedure supplies it.

The account proceeds from the destination to the routes and from the routes to the price of the most formal among them. Three routes lead to identical text: a chamber may accept the text of the other chamber without change, the two chambers may exchange amendments until the texts converge, or the chambers may convene a conference. The conference route carried a defect at its entrance, located in the Senate. Before a single conferee could be named, Senate procedure demanded three separate motions, each open to debate and therefore to obstruction, and the cost of clearing that entrance explains why the most documented route to agreement became the least traveled. What followed from the substitution, in lost records and concentrated control, closes the account.

The Identical Text Requirement

Why must the House and the Senate approve the same text before presentment?

Article I requires passage by the House and the Senate before presentment to the President. Congressional Research Service Report RL34611 states that the two chambers must approve the same bill or joint resolution in precisely the same form, so disagreement must end in one identical text.

The starting point is the vesting language of Article I, section 1 of the Constitution. That provision places the legislative powers granted by the Constitution in a Congress consisting of a Senate and a House of Representatives. The grant runs to a single institution with two component chambers. Lawmaking power belongs to Congress as a constituted body, and Congress acts through the concurrence of its two chambers rather than through the separate action of either one. The constitutional design treats the two approvals as parts of one act of legislation, and a single act of legislation has to have a single object. That object is a text, and the requirement of identity attaches to that text once section 1 is read alongside the presentment language of section 7.

Article I, section 7 then supplies the sequence that gives the structure its practical force. The Constitution provides that every bill that has passed the House of Representatives and the Senate shall, before becoming law, be presented to the President. Passage by the House comes first in the naming, passage by the Senate comes alongside it, and presentment follows after passage is complete. Presentment is the delivery of a passed bill for presidential action, and the clause assumes that the thing delivered is the thing that passed. If the House passed one version and the Senate passed another, the condition stated in section 7 would describe no single measure. One version passed the House and failed to pass the Senate in that form, and the other version passed the Senate and failed to pass the House in that form.

The Congressional Research Service states the resulting rule in direct terms. In the summary of Report RL34611, Elizabeth Rybicki writes that the Constitution requires the House and the Senate to approve the same bill or joint resolution in precisely the same form before presentment to the President. The subject is the same bill or joint resolution, so the rule governs the measure as a whole and not a policy theme that two measures might share. The standard is precise sameness of form, which excludes similar measures, equivalent measures, and measures that differ only in details one chamber regards as minor. The timing is before presentment, which places the burden of agreement inside Congress and ahead of any presidential involvement. Agreement is a precondition that Congress satisfies internally, and presentment certifies to the President that the precondition has been met.

The identical text rule is an inference from the structure of Article I rather than a sentence that the Constitution spells out in those words. No clause of Article I uses the phrase identical text, and no clause instructs the chambers to reconcile divergent versions by any named method. Congressional Research Service Report RL34611 treats the requirement itself as constitutional while the means of satisfying the requirement remain open, and that distinction runs through the entire subject of conference committees. The destination is fixed by inference from sections 1 and 7 read together, and the road toward the destination is not fixed at all. The Constitution names no reconciliation device, prescribes no committee for resolving disagreement, and gives neither chamber authority to impose a version on the other.

Identity, under this requirement, attaches to the whole text that each chamber approves. Agreement on purpose does not satisfy the rule, because purpose can be shared across very different texts. Agreement on a framework does not satisfy the rule, because a framework leaves operative language open, and operative language is what a bill presents for enactment. When the House approves a bill, the approval extends to every section, every condition, and every qualification in the version before the House at the moment of passage. Senate approval extends in the same total way to the version before the Senate. The requirement is met only when those two total approvals land on one set of words arranged in one form.

Amendments complicate the picture without changing the test, because an amendment adopted by one chamber becomes part of the text that requires the agreement of the other chamber. A Senate amendment to a House bill is not an attachment standing outside the measure, and a House amendment to a Senate bill is not a suggestion traveling alongside a finished text. Once a chamber adopts an amendment, the amended version is the version that chamber has approved, and the other chamber faces a choice about that amended version as a whole. The other chamber may accept the amendment and complete agreement on the amended text, may reject the amendment and preserve disagreement, or may propose a further amendment and continue the convergence. Passage counts as agreement only when the full text in that form has drawn approval from the House and from the Senate alike.

The strictness of precise sameness deserves emphasis because looser standards are easy to imagine and impossible to administer. The formulation in Congressional Research Service Report RL34611 leaves no room for that description to do legal work. A condition present in one version and absent from the other is a difference in form. An exception written into one version and omitted from the other is a difference in form. A qualification that narrows a grant of authority in the Senate version and appears in no part of the House version is a difference in form. Any such difference means that one chamber approved words the other chamber never approved. Presentment of either version would then present language lacking the concurrence that Article I, section 7 makes the trigger for presentment.

Enrollment follows agreement, and the position of enrollment in the sequence reinforces the same point. Enrollment is the preparation of the final official text of a measure for presentment, and a final official text can be prepared only after agreement between the chambers has settled the content. Agreement comes first, because agreement decides what the text says. Enrollment comes second, because enrollment records what agreement decided. Presentment comes third, because presentment delivers the recorded text to the President, whose action under Article I, section 7 is action on that bill and on no other artifact. Enrollment has no power to cure a lack of agreement, because enrollment depends on completed agreement rather than substituting for agreement.

The Supreme Court decision that bears most closely on this structure is INS v. Chadha, 462 U.S. 919, decided on June 23, 1983. The holding, as the verified record for this article states the holding, was that the one House legislative veto was unconstitutional. A legislative veto device allowed action by a single House to carry legal effect outside the full legislative process described in Article I, and the Court struck that device down. The reasoning reached beyond the particular device. Action by Congress that alters legal rights and duties outside the legislative process is valid, under Chadha, only where the Constitution expressly authorizes such action. Legislation changing legal rights and duties has to travel the prescribed route of bicameral passage and presentment laid out in Article I.

Chadha reinforces the identical text requirement through that reasoning about sequence and exclusivity. If the Article I sequence is the constitutionally prescribed route for legislation, each element of the sequence has to be genuinely satisfied before the legal consequences of legislation can follow. Bicameral passage is one element of the sequence, and passage by two chambers of two different texts is not bicameral passage of a bill in any sense that section 7 can recognize. Two chambers producing divergent texts and asking for those texts to count as one passed bill would present a different defect, yet the structural answer has the same root. The sequence admits no substitutes for its steps. Concurrence on one text is what the passage step means when the legislature has two chambers.

The limits of Chadha are as important as the reinforcement, and this article states the decision at exactly the strength the record supports. Chadha did not itself involve two different bill texts. No House version and Senate version stood in disagreement before the Court, no conference committee figured in the case, and no question about enrollment or amendment exchange required an answer. The case concerned a one House legislative veto, and the judgment invalidated that device. The authority for identical text remains the structure of Article I, sections 1 and 7, as Congressional Research Service Report RL34611 formulates the inference. Chadha enters the account as confirmation that the sequence from which the inference is drawn is exclusive and mandatory, a route Congress must follow rather than a default Congress may set aside.

That allocation of authority has a practical consequence for every measure on which the chambers disagree. Disagreement between the House and the Senate is a normal stage of bicameral lawmaking, and nothing in Article I treats an initial difference of versions as a failure. The House may pass a bill in one form because a majority of the House prefers that form. The Senate may amend the bill or pass a companion measure in another form because a majority of the Senate prefers different language. Legitimate separate action is not yet legislation, however, and the space between separate action and legislation has to be closed by further agreement. Resolution has one definition only, convergence on a single text that each chamber approves in precisely the same form.

Convergence can be reached through several devices, and the variety of devices confirms that the Constitution regulates the destination rather than the journey. Congressional procedure offers acceptance, under which one chamber passes the text of the other chamber without change and completes agreement in a single step. Congressional procedure offers amendment exchange, under which the chambers send amendments back and forth until agreement on one text emerges. Congressional procedure offers the conference committee, a joint and temporary committee of conferees from the House and the Senate, appointed to resolve the matters in disagreement and to report a single text for approval. The identical text requirement is indifferent to those differences. Each device is judged by a single test at the end of the process, asking whether the House and the Senate have approved the same measure in precisely the same form.

Because the test is constant, the choice among devices is a procedural question with procedural stakes. Acceptance places the burden of agreement on the chamber asked to surrender a version entirely. Amendment exchange distributes the burden across a series of formal actions on the floor of each chamber, with each action visible as a vote on identified language. A conference committee concentrates the burden in negotiation among appointed conferees and returns the product of that negotiation to each chamber for approval as a conference report. Those differences matter for the politics of a measure, for the control that leaders and committees exercise over final language, and for the record the process leaves behind. None of those differences touches the constitutional standard waiting at the end of the process. A conference report that one chamber approves and the other chamber rejects has produced no agreement, however skilled the negotiation behind the report may have been.

Failure to converge carries a consequence that is simple to state and absolute in effect. A measure on which the chambers never agree in the same form never satisfies the condition that Article I, section 7 attaches to presentment, and a measure never presented cannot become law through the route that Chadha described as the constitutionally prescribed path. The process remains unfinished because bicameral support for related texts is a different thing from bicameral passage of one text. The distinction can look formalistic when the versions are close, and closeness is exactly the situation the precise sameness standard in Congressional Research Service Report RL34611 governs. Close is a description of distance between two texts. Identical is a statement that no distance remains.

The rule extends to joint resolutions on the same footing as bills, because the formulation in Congressional Research Service Report RL34611 names the same bill or joint resolution as the unit requiring approval in precisely the same form. That extension follows the logic of presentment rather than the label attached to a measure. A joint resolution that must be presented to the President before becoming law has to arrive at presentment as one agreed text, and related joint resolutions with matching purposes and divergent language cannot satisfy the requirement. Form controls over label throughout this subject.

The identity of this article follows from that allocation of authority, and the identity deserves a direct statement. The subject is a process and rules topic, not a single statute. No act of Congress creates the identical text requirement, and no statute supplies the conference committee with a constitutional warrant. The authority for the subject is the Article I requirement itself, joined with House Rule XXII in the rules of the House of Representatives and Senate Rule XXVIII in the rules of the Senate, the provisions under which each chamber organizes the handling of disagreement and conference within chamber proceedings. Questions about whether agreement exists are constitutional questions governed by the standard of precise sameness. Questions about how a chamber moves toward agreement, appoints conferees, receives a conference report, or disposes of amendments are questions governed by the rule of the chamber where the action occurs.

House Rule XXII and Senate Rule XXVIII matter to the identity of the subject for a further reason, namely that the two rules belong to separate procedural systems answering to a shared constitutional condition. The House writes rules for the House, and the Senate writes rules for the Senate, and neither chamber can amend the rule book of the other chamber. The constitutional requirement is therefore the one element of the subject binding the two chambers into a single inquiry. Chamber rules explain how each side arrives at agreement. Article I explains what agreement is, why nothing short of agreement suffices, and why agreement has to be agreement on words rather than agreement on intentions, themes, or expected effects.

A chamber rule, however detailed, cannot relax the constitutional standard, and that hierarchy disciplines the way procedural choices should be described. A rule may structure debate on a conference report, limit amendments at a particular stage, or channel a disagreement toward one device rather than another. A rule may not declare that two different texts count as one text for purposes of presentment, because the standard that presentment presupposes comes from Article I and not from the rule book of either chamber. Rules govern the internal proceedings of a chamber, and the identical text requirement governs the relationship between the proceedings of two chambers and the act of presentment following those proceedings. A shortcut changes the route, while a dispensation would change the destination, and no chamber rule has that power.

Seen from that fixed point, the identical text requirement performs a quiet but decisive function in bicameral lawmaking, converting two separate majorities into joint authors of a single law. A majority in the House and a majority in the Senate may want similar things and may say similar things in the versions each chamber passes, yet the legal order following presentment contains laws rather than similarities. The statute book holds texts that courts interpret, agencies administer, and citizens read for the obligations and rights the texts create, the connection that Chadha placed at the center of the analysis by tying legal effect to the prescribed legislative route. Authorship of such a text by a legislature with two chambers has to be joint authorship in the literal sense, exercised over the same words in the same order with the same conditions and the same exceptions. The requirement makes that literal joint authorship the price of enactment, collected at enrollment and at presentment before any word of the measure can claim the authority of law.

The Three Routes to One Text

How do the two chambers arrive at a single text?

Congress reaches one text through acceptance, amendment exchange, or conference. One chamber may pass the other chamber’s text unchanged, the two chambers may trade amendments until both concur, or appointed conferees may negotiate a single text and report it back for approval by each chamber without amendment.

The starting point for every route is a constitutional condition that admits no shortcut. Article I vests the legislative powers in a Congress consisting of a Senate and a House of Representatives, and Article I, Section 7 provides that every bill that has passed the House and the Senate shall, before it becomes law, be presented to the President. The Congressional Research Service, in Report RL34611 by Elizabeth Rybicki, summarized the consequence directly: the Constitution requires that the House and Senate approve the same bill or joint resolution in precisely the same form before presentment. Two similar texts are not enough. The requirement attaches to the words themselves, and a measure that clears one chamber in one form and the other chamber in another form has not completed passage at all. The identical text rule is an inference from the structure of Article I, and the Constitution names no device for satisfying it. It sets the condition and leaves Congress to build the machinery, and the machinery Congress built consists of the three routes described in this section. Each route is a different answer to the same problem, and the differences among them decide who controls the final words of a law, how long agreement takes, and what kind of written record the agreement leaves behind. That choice sits inside the wider sequence traced in how a bill becomes law in practice, where passage in each chamber forms only one part of the path that ends at presentment.

Acceptance Without Change

The simplest route asks the least of the machinery and the most of the accepting chamber. Under acceptance, one chamber takes up the text the other chamber has already passed and passes that text without change. The House may pass a Senate passed bill exactly as it arrived, or the Senate may do the same with a House passed bill, and at that moment the two chambers have approved precisely the same form. The condition of Article I is satisfied in a single step, with no negotiation, no drafting session, and no further exchange. The route demands nothing of procedure and everything of political will, because the accepting chamber surrenders every editorial instrument it ordinarily holds. It cannot strike a clause it dislikes. It cannot soften a provision, tighten a definition, or repair a drafting error that its own members have identified. The choice presented to the accepting chamber is binary in the strictest sense: take the text whole, exactly as the other chamber wrote and passed it, or refuse the text and lose the measure in its existing form. A single amendment, however small, destroys the route, because an amended text is a different text, and a different text must return to the other chamber for fresh agreement before presentment can occur.

That cost explains both the rarity and the strategic value of outright acceptance. A chamber accepts a text without change when the alternative paths look worse than the imperfections in front of it. Agreement through amendment exchange may fail. A conference may collapse and take the measure down with it. The calendar may leave no room for further rounds. In such calculations, the text in hand carries a decisive advantage over any text that might emerge from further bargaining: it exists, it has passed one chamber, and it can become the agreed text the moment the second chamber votes. The accepting chamber’s members may explain reservations in debate, and those explanations enter the record, but explanations do not alter the operative text. Acceptance is, in that sense, the route of maximum speed and minimum control, and a chamber chooses it with open eyes only when the measure’s survival outweighs its own hand in the drafting.

Amendment Exchange Between the Chambers

Amendment exchange resolves disagreement without surrender and without a conference, by sending the text itself back and forth between the two floors until one version commands the assent of both. The Congressional Research Service described the mechanics in Report 98-696, Resolving Legislative Differences in Congress, by Elizabeth Rybicki, and the mechanics repay close attention because they explain why the route behaves as it does. The exchange begins when one chamber passes a measure and the second chamber passes it with amendments and returns it. Formal messages between the chambers carry the text and its amendments from one floor to the other, and each message presents the receiving chamber with a defined set of choices. The receiving chamber may concur in the amendments, ending the exchange in agreement. It may amend the amendments and return them. It may disagree and insist on its own position.

The defining mechanical feature of the exchange is its built-in limit, and the limit comes from the law of amendment degrees. Under the practice described in Report 98-696, each chamber gets one opportunity to amend the amendments of the other. Parliamentary practice does not permit amendments beyond the third degree, so the exchange has a finite structure rather than an open-ended one. An amendment to a text sits in the first degree. An amendment to that amendment sits in the second degree. An amendment to the amendment to the amendment sits in the third degree, and there the ladder ends, because an amendment beyond the third degree is not in order. By capping the layers of amendment stacked on the underlying text, the degree rule caps the rounds of the exchange itself. As the degrees are consumed, the choices narrow toward their endgame forms: concur and settle the matter, recede from a position previously taken, or insist and place the burden of the next move, and possibly of the measure’s failure, on the other chamber. Where the gap between the chambers consists of identifiable provisions that can be traded in a small number of moves, the degree limit disciplines both sides toward settlement. Where the gap is wide and the contested provisions are many, the limit warns both sides early that the exchange may run out of room before it runs out of disagreement.

The two chambers do not meet the exchange on equal procedural footing, and Report 98-696 records an asymmetry that shapes how the bargaining unfolds. House amendments to Senate bills, and Senate amendments taken up in the Senate’s consideration of House measures, hold privileged standing in the Senate, which gives them a claim on the Senate’s floor business. Senate amendments to House bills hold no such privilege in the House. A Senate amendment arriving in the House does not carry its own right to prompt floor consideration, and the House commonly disposes of Senate amendments under a special rule reported by the Rules Committee. The special rule does procedural work that privilege does elsewhere. It brings the amendment to the floor on terms the House majority has chosen, it structures the debate and the form in which the amendment is considered, and it can close the process by limiting or foreclosing further amendment. The practical consequence is that the same amendment travels differently depending on its direction. Moving toward the Senate, it arrives with standing. Moving toward the House, it arrives as raw material for a rule, and the majority that controls the Rules Committee controls the conditions under which the Senate’s work is accepted, amended, or rejected.

The exchange ends in one of two ways, and the second way explains how measures die with both chambers professing to want agreement. Agreement comes through concurrence, when a chamber accepts the final amendment in the form in which it arrived and the two chambers stand on identical words. Failure comes through the vocabulary of insistence and recession. A chamber that insists adheres to its own amendment or its own disagreement and sends the matter back. A chamber that recedes withdraws from a position it previously took and accepts the other chamber’s terms. An exchange of insistence met by insistence consumes the remaining degrees without producing agreement, and when neither chamber recedes and no chamber concurs, the text never becomes identical and the measure ends unresolved. The endgame has a clarity that conference bargaining lacks. Each chamber’s position is stated in a motion, voted on the floor, and answered by the other chamber in kind, so responsibility for a failure attaches to identifiable votes rather than to a closed negotiation.

Amendment exchange keeps control of the text with chamber leadership and the floors themselves rather than with a committee, and the mechanics account for that distribution of power. No new body is created. The committee of jurisdiction does not reconvene to rewrite the measure. Leadership decides when a message is taken up, in the House under the terms of a special rule and in the Senate under the standing that privileged amendments carry. Leadership frames the amendment that a chamber offers in reply, because the reply must be drafted to survive the degree count and to present the other chamber with a choice it can accept. No agreement binds a chamber until its floor has voted it. The contrast with conference is structural. A conference delegates the drafting to a small appointed group and asks the floors to ratify the result. Amendment exchange delegates nothing, and the price of that refusal is paid in floor time and in the degree limit, while the benefit is that the final words of the measure are written through public motions by the chambers that must answer for them.

Conference in Overview

The third route delegates the disagreement itself. A conference committee is a joint, temporary committee of conferees, called managers, drawn from both chambers and appointed to resolve the matters in disagreement between the House and the Senate. The conference exists for one measure and one purpose. It is not a standing committee, it holds no continuing jurisdiction, and when its work on the measure ends, the conference ends with it. Its membership comes from the two chambers whose disagreement created it, and its charge is bounded by that disagreement: the matters the chambers have placed in dispute define what the conferees may settle.

The product of a conference is a single text. The conferees negotiate across the whole of the disagreement at once rather than trading amendments provision by provision through messages, and when a majority of the managers from each chamber reaches agreement, the conference reports its text back to the two chambers as a conference report. Each chamber then acts on that reported text as a whole. The bargain the conferees struck is presented for approval in the form in which it was struck, and the chambers’ choice resembles the choice under acceptance more than the choice under amendment exchange: the reported text is taken as a package, because reopening one provision would unravel the trade-offs that produced agreement on the rest. The appointment of conferees, the scope of their authority, and the consideration of conference reports belong to later sections. For the comparison among the three routes, the essential features are the delegation to a small temporary body, the negotiation of the entire disagreement in one setting, and the single take it or leave it text that comes back to the floors.

Set beside acceptance and amendment exchange, the conference occupies the opposite pole from acceptance on the question of authorship. Conference concentrates authorship in the conferees, the smallest group of the three arrangements, and gives the two chambers a ratifying vote on work done elsewhere. A chamber that sends a measure to conference trades direct control over the final words for the conferees’ capacity to settle many disputes in one negotiation, a capacity neither outright acceptance nor the degree limited exchange can match on a measure carrying a heavy load of disagreement.

Time, Control, and the Written Record

The three routes spend floor time in different currencies. Acceptance is the fastest route when it is available at all, because it requires exactly one further passage vote and no drafting, no messages, and no negotiation, yet its speed is conditional on a chamber’s willingness to swallow a text whole, and a chamber unwilling to pay that price cannot use the route at any speed. Amendment exchange spends time in rounds. Each message between the chambers requires floor action, so a disagreement settled in one exchange of messages is cheap and a disagreement that consumes the full degree ladder is expensive in the scarcest resource either chamber holds. Conference spends its time before the floors act rather than during floor action. The negotiation among conferees may run long, but the text that emerges asks each floor for a single decision rather than a sequence of them. A leadership choosing among the routes is therefore choosing where the time cost falls: in one decisive vote, in a series of floor engagements, or in a negotiation conducted away from the floor and ratified at the end.

Control follows the same pattern of trade-offs, and the routes can be read as three answers to the question of who writes the last version of a law. Under acceptance, the chamber that passed first writes it, entirely, and the second chamber’s control is the blunt control of refusal. Under amendment exchange, control stays distributed and sequential. Each chamber writes in turn, each chamber’s writing is exposed to the other’s reply, and leadership on both sides shapes the outcome through the management of floor action. Under conference, control is delegated and concentrated. The conferees write, the chambers ratify, and the members outside the conference room influence the result through the appointment of managers, through instructions, and through the final vote rather than through direct amendment of the text. Each route encodes a judgment, made measure by measure, about whether speed, floor control, or negotiating capacity matters most for the disagreement at hand.

The written record left behind differs as sharply as the control. Acceptance leaves the thinnest record of agreement between the chambers, because the accepting chamber contributes no text and its assent is expressed in a vote rather than in drafting. Amendment exchange leaves the most layered record. Its messages, amendments in successive degrees, motions to concur, to insist, and to recede, and the votes on each, build a step by step account of what each chamber demanded, conceded, and finally accepted. Conference leaves a record of a different kind, centered on the conference report itself: the single negotiated text, reported jointly by the managers of both chambers, speaks for the agreement in a way that the scattered record of an exchange does not.

The Health Care Sequence and the Price of Acceptance

The acceptance route usually operates on measures of modest consequence, where the price of swallowing a text whole is small. March 2010 showed the route operating at full scale on the largest measure of its Congress. The Senate had passed its health care text, and the House faced the choice the acceptance route imposes in its purest form. Amending the Senate text in any particular would have sent the measure back across the Capitol and reopened a Senate agreement that had been assembled with difficulty. The House instead accepted the Senate passed text and passed it without change, and no conference was held. The identical text requirement was satisfied by the House’s vote alone.

The sequel showed how a chamber can accept a text and legislate around it. The House’s acceptance did not mean that every member who voted for the Senate text regarded that text as final policy in every detail. Adjustments to the framework were carried in a separate reconciliation measure, a distinct vehicle with its own passage in both chambers, rather than in amendments to the accepted text itself. The distinction is the whole lesson of the episode. Amendments to the accepted text would have destroyed the agreement the acceptance created, because the chambers would again have held different words. The episode also displayed the political price of the route without disguise. House members who had reservations about particular Senate provisions voted for those provisions anyway, in the chamber’s own name, because the acceptance route offers no mechanism for voting for a text with exceptions. The reservation and the vote stand side by side in the record, and the text that was presented carried provisions its accepting chamber had publicly criticized.

Read against the other routes, the March 2010 sequence marks the boundary of what acceptance can do. Amendment exchange could not have carried that disagreement, because the contested matters were too many for a degree limited trade of messages, and the conference route was available in form while the House majority judged its risks greater than the price of outright acceptance. The measure became law on the strength of a single chamber’s decision that the other chamber’s finished text, taken whole, was worth more than any text the House might have written for itself.

Inside a Conference

How does a conference begin in the House?

A conference begins after one chamber disagrees to the other chamber’s amendments, or insists on its own, and the two chambers agree to meet. The House then requests a conference or accepts a Senate request, and the Speaker appoints the House managers who will negotiate.

The sequence is formal because a conference holds no authority until the disagreement between the chambers exists as a recognized parliamentary state. Two different texts passing two different chambers create a gap, and the gap by itself authorizes nothing. The account in CRS Report 98-381 describes the House as moving through three connected actions. The House disagrees to the Senate amendments, or insists on its own amendments where the House position is the one the Senate has altered. The House then requests a conference, or agrees to a conference that the Senate has requested. The Speaker then appoints the House conferees, called managers, and only at that point does a body exist that can negotiate in the House’s name. Each action depends on the one before it. Disagreement without a request leaves the measure suspended between the chambers, and a request without appointed managers leaves nobody empowered to sit across the table.

The motion to go to conference carries privilege in the House when it is offered by direction of the committee or committees of initial referral, and it is debated for one hour. Privilege gives the motion precedence over most other business, so the question of entering a conference does not wait in a queue behind unrelated measures. The one hour of debate places a fixed boundary around the argument over whether negotiation should begin at all. Because the motion comes by direction of the committees that first handled the measure, the proposal to go to conference arrives with the sponsorship of the members who know the text and the points of dispute. The House acts at each of its steps by majority vote, and a majority that wants a conference can therefore create one on a predictable timetable. Entry into a conference is, on the House side, an act of majority will executed through a short and structured procedure.

Appointment and the Shape of the Two Delegations

Formal appointment authority sits in different hands in the two chambers. The Speaker appoints the House managers. In the Senate, the presiding officer appoints the conferees. Practice channels both appointments through the committees that considered the measure, since appointment in each chamber is made in practice on the recommendation of those committees. The conferees came largely from the standing committees that had considered the measure, the same bodies whose work through the committee system and markups had shaped the text each chamber carried into the disagreement. The recommendation practice means that the formal appointing officers ratify a selection that the committees of jurisdiction have in substance already made, and the conference inherits the personnel of the earlier stages rather than drawing negotiators from the chamber at large.

Senior members predominate in that inheritance. The members recommended for a conference tend to be the senior figures of the committees involved, the chairs and ranking members and the members who managed the bill through committee and floor consideration. Seniority in this setting tracks knowledge. A senior conferee carries memory of why a provision took its form, which amendments the chamber accepted and which it refused, and where the floor majority showed its limits during debate. Negotiation over disputed text rewards exactly that memory, because every concession a conferee weighs has to be measured against what the appointing chamber will accept at the end. A delegation built from senior committee members arrives at the table with the chamber’s own negotiating history embodied in the people seated there.

Where a measure passed through more than one committee, the recommendation channel widens with it. The committees that considered the measure may be several, and a conference delegation assembled on their recommendations can carry members from each of them, so that every strand of the text in disagreement has a negotiator who worked on that strand at the committee stage. The delegation then mirrors the measure’s own path through the chamber. Provisions written by one committee are defended by that committee’s members, and provisions that crossed jurisdictions are negotiated by conferees who can speak for the committees that produced them. The appointment practice thus ties the conference back to the structure that built the bill, committee by committee, instead of treating the measure as the property of the chamber in the abstract.

No rule requires the two delegations to match in size. Unequal delegation sizes are permitted, and one chamber may send more conferees than the other without affecting the standing of either side. The reason lies in how a conference reaches its decisions, examined below: each delegation acts through a majority of its own members, so numerical weight across the table decides nothing. A large delegation gains no advantage from its size, because its internal majority produces one position, and a small delegation loses nothing, because its internal majority produces a position of equal procedural weight. Appointment is therefore free to serve representational purposes inside each chamber, spreading seats among the committees and factions with a stake in the measure, rather than serving any need for parity with the other side.

Instructions From the Two Floors

Before a conference meets, each chamber may tell its conferees what it wants, and neither chamber can make those directions bind. In the House, a motion to instruct conferees may be offered before the managers are appointed. The motion states the floor’s preference on the matters in disagreement, asking the conferees to hold a position, to accept a Senate provision, or to resist one. The instructions are never binding. No point of order lies against a conference report on the ground that the conferees departed from their instructions, so a report that defies an instruction in every particular reaches the floor with its procedural validity intact. The design is deliberate. A binding instruction would reduce the conference to a drafting errand and expose every genuine compromise to destruction on a point of order whenever the compromise crossed a line the floor had drawn in advance. Nonbinding instructions preserve the conferees’ freedom to bargain while giving the floor a recorded statement of its wishes, and a conferee who disregards a strongly supported instruction accepts a political risk rather than a procedural defect: the same floor that issued the instruction will later judge the report.

The House holds a second instruction device in reserve against delay. After a conference has been in session for 45 calendar days and 25 legislative days without reporting, a privileged motion to instruct the conferees becomes available. The two measures of time run on different clocks. Calendar days pass whether Congress sits or not, while legislative days accumulate only through the chamber’s own proceedings, so the paired thresholds wait for both a long passage of time and a substantial record of sitting days before the device unlocks. Privilege matters here as it does at the conference’s creation, because the motion can be called up ahead of competing business and forces the chamber to confront the stall in the open. The motion pressures rather than compels, since instructions remain nonbinding whenever they are offered, but a floor vote instructing conferees who have produced nothing for that span of time carries a message about patience that the conferees ignore at the cost of the report’s reception.

Senate practice gives instructions a wider and looser field. Motions to instruct Senate conferees may be offered after the motions that take the Senate to conference and before the conferees are appointed, and there is no limit on their number. Each such motion is debatable. Senators may therefore stack competing instructions on the same conference, one urging resistance on a provision and another urging acceptance, and each may be argued before the chamber votes. The absence of a numerical limit fits a chamber organized around extended debate and individual initiative, and it contrasts with the House’s placement of instructions at fixed points in a managed sequence. The legal character of the product is identical in both chambers, however numerous or hard fought the Senate motions become. Senate instructions are also nonbinding, advice backed by the weight of a floor majority and nothing more.

Open Meetings, Separate Votes, and Signatures

A conference conducts its formal business in the open unless the House votes to close the doors. House Rule XXII clause 6 makes conference meetings open by default, and closure requires the House itself, voting by roll call in open session, to order a meeting closed. The conferees cannot shut their own proceedings on their own authority; the decision to bargain in private belongs to the chamber that appointed them and must be taken on the record, member by member. The default reflects the position a conference occupies in the legislative process. The text a conference produces will return to the floors in a form that cannot be amended, so the open formal meeting is the one stage at which the bargaining over that final text takes place in view of the members who will later have to accept or reject the result whole.

The openness requirement carries a sanction with teeth. A failure to hold the required formal open meeting supports a point of order against the conference report, and the consequence of that failure is that the report is rejected and a new conference is automatically requested. The remedy does not patch the defective product or excuse the defect as harmless. The report dies, the negotiation restarts under a fresh conference, and the conferees must produce their agreement again in compliance with the rule. Senate Rule XXVIII also addresses open meetings, placing a parallel expectation in the Senate’s own conference rule, so the transparency obligation rests on the practice of both chambers rather than on a single chamber’s housekeeping.

Inside the room, the conference decides as two bodies, not as one merged committee. The two delegations deliberate and vote separately, and a majority of each delegation must agree before the conference can act. An agreement carried by a combined majority of all conferees seated together would be worthless if one delegation’s majority opposed it, because the conference exists to reconcile two chambers, and a product that one chamber’s own negotiators rejected by majority would misdescribe the state of agreement between the institutions. Separate voting keeps the bicameral structure intact inside the committee room. It also explains the tolerance for unequal delegation sizes, since each side’s internal majority is the only count that matters, and it defines the conference’s characteristic failure mode: agreement fails not when conferees in general balk, but when either delegation, voting alone, cannot assemble a majority for the terms on the table.

The signature rule certifies that the separate majorities existed. Under House Rule XXII clause 7(e), the conference report and the joint explanatory statement that accompanies it must be signed by a majority of the managers of each House. Signatures convert the internal votes into an attested fact that the floors can inspect. A report bearing the signatures of a majority of the House managers and a majority of the Senate managers proves, on its face, that each delegation approved the text through the channel the rules require. The joint explanatory statement travels with the signatures and sets out the agreement the managers reached, so the document the chambers receive is simultaneously the compromise text, the explanation of that text, and the certification that both delegations stood behind it by majority. A report short of majority signatures on either side fails to establish that a conference agreement occurred at all.

The signed product carries weight beyond the vote that disposes of it. The joint explanatory statement records the managers’ own account of the agreement they signed. The statement is not a floor speech or a committee aside; it is the explanation issued by the negotiators of both chambers, approved through the same majority signatures that validate the report itself, and presented to each chamber as part of the act of adoption. A text explained by the very body that wrote the final version, and ratified by both chambers in that explained form, enters the statute books with an unusually direct record of what its authors understood it to do.

The Report Returns to the Floors

The finished conference report returns to the two chambers with privilege in both, so it can be called up for consideration ahead of competing business in the House and in the Senate alike. It arrives on terms that define the end of the legislative process for the measure. The report is unamendable. Neither chamber may strike a provision, soften a figure, or attach a condition. Each chamber votes on approval or rejection of the report as a whole, accepting the entire compromise or refusing it entire. The rigidity is the price of the conference’s freedom to bargain. Any amendment adopted on either floor would alter the single text the conferees produced and recreate the disagreement the conference existed to end, sending the measure back into exchange between the chambers with the negotiated settlement unravelled. The take it or leave it structure forces each chamber to weigh the compromise as its authors framed it, as a package in which concessions on one side purchase gains on the other.

One escape remains, and it belongs only to the chamber that acts first. The first chamber to consider the report may recommit it, sending the report back to the conference for further work rather than approving or rejecting it outright. Recommittal presupposes that a conference exists to receive the report, and that condition expires the moment the first chamber adopts the report instead. Adoption dissolves the conference. The second chamber then faces the report without any recommittal option, because the body that would have reconsidered the text no longer exists, and its choice narrows to adoption or rejection. The order in which the chambers take up a report therefore carries real procedural consequence. The chamber moving first holds a middle course between acceptance and refusal that the chamber moving second cannot recover, and a chamber that expects difficulty with a report has reason to value acting while the conference survives to absorb a recommittal.

Senate consideration of a conference report begins under a protection that the report’s opponents cannot talk away. Under Senate Rule XXVIII paragraph 1, the motion to proceed to the consideration of a conference report is not debatable, so the threshold question of whether the Senate will take up the report is put to a vote without extended argument. The report itself, once before the Senate, is debatable, and senators may argue its merits at length in the ordinary way. The division between a nondebatable threshold and a debatable substance keeps the Senate’s decision from being consumed at the doorway while preserving full argument on the agreement itself. A statutory exception then caps debate for a defined class of measures. The Congressional Budget Act limits Senate debate on conference reports for budget resolutions and reconciliation bills to ten hours. Budget measures move on a statutory timetable, and the ten hour ceiling guarantees that the conference stage of that timetable ends in a vote, placing those reports on a faster track than the general run of conference business, where debate on the report proceeds without that statutory boundary.

The allocation of power across the whole sequence is consistent from appointment to final vote. Conferees receive wide freedom inside the room: instructions cannot bind them, their deliberations produce a text no floor may edit, and their agreement needs only the separate majorities and signatures of their own delegations. The floors receive a correspondingly blunt authority outside the room: an open meeting they can enforce by killing a report, instructions they can issue to mark their displeasure, and a final vote that accepts or destroys the compromise entire. Neither side of that division can do the other’s work, and the measure that emerges has been shaped by an institution built to hold the two kinds of power apart until the moment the chambers vote.

The Price of Admission in the Senate

Why could one objection turn a conference into three filibusters?

Each of the three motions required before the Senate could enter a conference was debatable under Senate procedure. When unanimous consent bundled them, adoption took moments. A single objection broke the bundle, and extended debate on each motion could then end only through a separate cloture vote.

The conference committee occupied a distinctive place among the routes by which the House and the Senate reached identical text. Acceptance asked one chamber to swallow the work of the other without change. Amendment exchange moved a text back and forth until one side concurred or receded. Conference asked something different. Conferees from each chamber sat together, bargained over the matters in disagreement, and reported a single text that each chamber then accepted or rejected as a whole. Before a single conferee could be named, the Senate had to clear a sequence of preliminary motions, and every motion in that sequence carried the potential for extended debate. The Congressional Research Service account of going to conference in the Senate, Report RS20454, described a threshold that looked clerical in ordinary times and formidable in contested ones, a threshold that could stop a conference before it started, not by a vote against the conference but by refusal to let the Senate reach a vote at all.

Senate procedure historically required three separate motions before a conference could begin, and Report RS20454 set them out as distinct acts of the chamber. By the first motion, the Senate disagreed to the amendments of the House or insisted on its own amendments, a step that placed the two chambers formally in disagreement and defined the ground the conference would occupy. By the second motion, the Senate requested a conference with the House or agreed to a conference that the House had requested, a step that converted disagreement into an invitation to negotiate. By the third motion, the Senate authorized the presiding officer to appoint Senate conferees, a step that supplied the named individuals who would carry the Senate position into the bargaining room. Each motion did real work. The sequence had a logic, moving from a declaration of difference through a decision to negotiate to the selection of negotiators. The difficulty lay elsewhere, in the procedural character that the three motions shared and in the way that character interacted with the Senate practice of transacting routine business by unanimous consent.

In practice, the three motions were usually combined and agreed to by unanimous consent. A senator acting for the majority would ask that the Senate disagree to the House amendments or insist on the Senate position, request or agree to a conference, and authorize the presiding officer to appoint conferees, all in a single request, and in the absence of objection the Senate adopted the whole package at once. The request passed in the rhythm of the many unanimous consent agreements by which the Senate arranged its daily work, and the conference began. That practice depended entirely on the acquiescence of the full chamber, because unanimous consent has a defining property that its name states plainly. Consent that must be unanimous can be destroyed by one voice. A single senator objecting outright ended the short route. The bundle came apart, and the Senate faced the three motions as separate questions, each to be moved, debated, and disposed of under the regular rules. What unanimous consent had compressed into a formality, objection expanded into a campaign.

The expansion mattered because each of the three motions was debatable. Senate debate, absent a unanimous consent agreement limiting time or a successful cloture motion, has no natural endpoint set by the presiding officer, and a debatable motion can therefore be talked against for as long as senators willing to hold the floor choose to talk. An objection to unanimous consent did not defeat the conference, and it did not need to command a single vote beyond the objector. The objection changed the procedural terrain. A majority that wanted a conference had to move the motion to disagree or insist and sustain debate on that motion until debate ended or cloture intervened, then move the motion to request or agree to a conference and repeat the process, then move the motion authorizing appointment of conferees and repeat the process a third time. Each stage presented a separate opportunity for extended debate, which meant that the threshold to conference contained the material for three distinct filibusters arranged in series. The majority could win every eventual vote by wide margins and lose days to the sequence anyway, because the contest at each stage concerned time as much as numbers.

The only regular instrument for closing debate against determined opposition was cloture, and the terms of that instrument set the arithmetic of the whole problem, a subject treated in the companion account of filibuster and cloture explained. Cloture required three fifths of the Senate, a supermajority threshold that a majority party short of that strength could not meet on party support alone and that even a larger majority had to assemble deliberately, senator by senator. Cloture also consumed floor time by its own operation. A cloture motion had to be filed, had to ripen under the rule, and had to be voted on in sequence, and the days occupied by that sequence were days the Senate could not easily give to other bills, nominations, or conferences waiting behind the one in dispute. Multiplied across the threshold motions, the arithmetic turned severe. One objection could oblige the majority to file cloture on the motion to disagree or insist, win three fifths of the Senate, and spend the attendant floor time, then file cloture on the motion to request or agree to a conference and spend comparable time again, then file cloture on the motion authorizing appointment of conferees and spend comparable time a third time. The majority paid in the scarcest currency the Senate possesses, scheduled floor time in a chamber where every measure competes for the same hours.

Floor time scarcity explains why the threat of the sequence often mattered more than the sequence itself. A majority leader arranging the Senate calendar weighed the conference for one bill against every other commitment pressing on the session, including appropriations measures with deadlines and nominations awaiting confirmation. A conference that required three cloture campaigns before the bargaining could even open presented a poor bargain, and leaders could rationally decline the bargain by turning to amendment exchange, by pressing the House to accept the Senate text, or by letting the bill wait. The minority understood that calculation and could price an objection accordingly. Delay at the threshold had value as leverage over the content of the eventual conference, over the identity of the conferees, over instructions the Senate might send with its negotiators, and over unrelated matters that a leader might trade to clear the path. The procedure translated intensity into power. A small number of senators who cared greatly about stopping or conditioning one conference could impose costs that a larger number of senators who cared moderately about holding that conference might refuse to bear.

The three threshold motions were not the only debatable instruments clustered at the entrance to a conference. Senators could also offer motions to instruct the conferees, directing the Senate negotiators, in the judgment of the Senate, toward particular positions on the matters in disagreement, and those motions to instruct were themselves debatable. No fixed limit capped the number of such motions that senators could put before the chamber in connection with a conference, so the device supplied an additional lever that operated alongside the three required steps rather than within them. A senator who had watched the majority grind through cloture on the threshold motions retained the ability to propose instruction after instruction, with debate attaching to each proposal and floor time draining away on questions that preceded any actual bargaining between the chambers. The threshold motions controlled whether and when a conference began, the instruction motions shaped the terms on which the Senate side entered the room, and both sets of motions drew on the same resource of debatable questions that only cloture or exhaustion could close.

The House of Representatives faced no equivalent problem at the corresponding steps, and the contrast locates the difficulty in Senate rules rather than in the nature of conferences themselves. In the House, the chamber disagreed to the Senate amendments or insisted on its own amendments, requested a conference or agreed to a conference requested by the Senate, and left appointment of House conferees to the Speaker. The motion to go to conference in the House was privileged when offered by direction of the committee or committees of initial referral, and the House debated that motion for one hour. Privilege secured the motion a place on the floor, the one hour limit bounded the argument, and the House majority could then act on its steps with majority votes. A determined House minority could speak against going to conference and could vote against the motion, but the minority held no instrument for converting each step into extended debate that only a supermajority could end. The House majority controlled its steps from start to finish and paid for a conference in votes rather than in days.

That veto, in its effects rather than in any partisan design, fell on whichever party held the Senate majority and needed a conference, and the history of the practice has a symmetry that the Congressional Research Service accounts preserve. Members of both parties used the objection when in the minority and deplored it when in the majority. A senator in the minority, confronting a bill shaped by the other party and a conference likely to ratify that shape, had every incentive to raise the price of admission, and senators of each party yielded to that incentive when the minority role was theirs. The same senators, rotated into the majority by an election and charged with moving a legislative program through a crowded calendar, discovered the virtues of dispatch and condemned the obstruction that an earlier Congress had practiced from the other side of the aisle. The incentive structure rewarded objection from the minority bench and punished delay from the majority bench, and officeholders responding to the structure sounded consistent within each role even as the roles reversed. The cost of the three motions therefore moved with majority status rather than with party identity. Any majority that needed conferences on contested measures faced the toll, and any minority that preferred different terms, a different vehicle, or simple delay could collect it.

The repair came in the 113th Congress, through Senate Resolution 16, adopted on January 24, 2013, and codified in Senate Rule XXVIII at paragraph 2, with the Congressional Research Service describing the change in Report R42996. The resolution attacked the multiplication problem at the threshold rather than the debatability of the threshold itself. Under the arrangement the resolution installed, the three steps that had once stood as separate motions were joined into a single compound motion, one question that carried the disagreement or insistence, the request for or agreement to a conference, and the authorization for the presiding officer to appoint conferees together. The compound motion was not divisible, which meant that the Senate could not split the package back into its component parts at the demand of a senator seeking separate debates and separate votes on each part. The rule closed that route. A senator opposed to the conference faced one motion, one debate, and one cloture question rather than a series of three, and the majority faced one campaign to open the conference door rather than three campaigns in succession. The architecture of delay lost its multiplier.

Those cloture terms rewrote the price schedule for the threshold. If cloture was filed on the compound motion, debate before the cloture vote was limited to no more than two hours, a cap that replaced the open ended ripening and argument that had made each earlier cloture effort so expensive in floor time. If cloture was invoked, meaning that three fifths of the Senate voted to close debate, the compound motion itself was then adopted by simple majority with no further debate. That second feature mattered as much as the first. Before the repair, success on cloture at one threshold motion purchased only the right to proceed to the next debatable motion, where the whole contest could begin again. Under Rule XXVIII paragraph 2 as the 2013 resolution framed it, success on cloture for the compound motion ended the threshold fight in a single act. The Senate that mustered three fifths to close debate adopted the motion to go to conference by simple majority, without another round of argument, and the presiding officer could appoint conferees. The supermajority requirement survived at the cloture stage, and the majority threshold governed the decision itself, an arrangement that kept the Senate protection for extended debate in a narrowed form while stripping out the serial repetition that had let one objection tax a conference three times over.

The limits of the repair deserve the same precision as the repair itself. Senate Resolution 16 did not make the compound motion nondebatable, and the resolution did not lower the cloture threshold for the motion. A senator could object to unanimous consent on the compound motion just as a senator could object to the bundled request under the older practice, and the objection retained real force, because the majority had to file cloture and had to find three fifths of the Senate to invoke it. What the objection could no longer do was multiply. One objection produced one debate under a two hour cap if cloture was filed, one cloture vote, and one disposition of the whole threshold question. The motions to instruct conferees, debatable and unlimited in number, remained part of the surrounding procedure, because the repair addressed the three threshold motions rather than every debatable question connected with a conference. The settlement that the 113th Congress struck can therefore be read as a judgment about proportion.

The test concerned the budget. In the spring of 2013, after the Senate had passed a budget resolution and the House had passed a budget resolution in differing form, Senator Reid repeatedly asked unanimous consent that the Senate go to conference with the House on the budget, and the requests met objection. Contemporaneous reporting by The Hill and Politico described the central objection, raised by Senator Cruz, as a demand that any agreement emerging from a budget conference rule out a tax increase and rule out a debt limit increase, conditions stated in advance of the conference rather than negotiated within it. The demand attached to the entrance of the conference sought to settle, before conferees met, two of the questions that a budget conference between a Senate majority of one party and a House majority of the other would exist to contest, and the objection that carried the demand drew its strength from the same consent practice that governed every conference threshold. The majority had options under the rule adopted in January of that year, including the compound motion and its shortened cloture path, and those options framed the standoff without ending it, because a cloture campaign on the budget conference, even a shortened one, required three fifths of the Senate and floor time that the majority had to choose to spend against the other demands of a session moving toward a confrontation over funding the government.

The conference was eventually convened in October 2013, after the government shutdown that began that month had rearranged the incentives of every participant in the budget dispute. Objection at the door of a conference rarely operated as a permanent bar, because the underlying disagreement between the chambers persisted and the need for a single text, or a single budget framework, pressed on both sides until some route to agreement became unavoidable. Objection operated instead as a tax on timing and terms, raising the cost of beginning negotiations until the majority judged the cost worth paying, extracting conditions as the price of passage, and shifting the bargaining that conferences were built to house into leadership negotiation over whether a conference would happen at all. The Cruz conditions reported by The Hill and Politico, that a conference agreement exclude a tax increase and exclude a debt limit increase, belonged to that shifted bargaining. The conditions were proposals about the outcome of a negotiation, presented as terms for permitting the negotiation to open, and the months between the first objections and the convening of the conference measured the time during which the two chambers conducted their budget argument at the threshold rather than at the conference table.

Weighed against the arrangement that Senate Resolution 16 displaced, the budget episode also shows what the repair accomplished and what the repair deliberately left untouched. A majority determined to force the budget conference question under Rule XXVIII paragraph 2 faced one compound motion, a cloture debate capped at two hours if cloture was filed, and adoption by simple majority after cloture, a path measurable in hours where the older threshold had threatened repeated cloture campaigns across three motions. The repair therefore changed the bargaining position of an objector without abolishing the position. Before January 2013, an objection could credibly threaten to consume the calendar, and the threat itself could kill a conference that a majority supported in principle but valued less than the rest of its program. After the repair, an objection could compel a bounded vote sequence and could deny a conference to a majority that lacked three fifths of the Senate for cloture, yet the objection could not credibly threaten the open ended multiplication that had made the threshold a graveyard for conferences whose supporters had counted the cost and turned away. The price of admission fell from three motions, each with a filibuster attached, to one motion with one filibuster attached and a clock running on the debate, and a price payable in supermajority votes and floor time is a real price even after the discount, which is why conferences after the 2013 repair remained decisions that majorities weighed rather than formalities that clerks recorded.

Scope, Airdrops, and the Honest Counter-Reading

What does scope limit conference committees to?

Scope limits conferees to the matters in disagreement between the two chambers. Conferees act as agents of their chambers for those matters only, and the rules of each chamber forbid inserting matter committed by neither House and striking matter that both Houses had already agreed upon in full.

The scope principle follows from the character of a conference itself. A conference committee was a joint, temporary committee of conferees, also called managers, appointed from both chambers to resolve the matters in disagreement and to report a single text. Appointment for a defined disagreement gave the conferees their authority, and the same appointment marked the boundary of that authority. Conferees were agents of their chambers for the matters in disagreement only. Agency language carried a practical meaning inside the legislative process. A chamber that had passed one provision and a chamber that had passed another had committed the difference between those two positions to conference. A chamber that had passed nothing on a subject had committed nothing on that subject, and a subject on which both chambers had already agreed presented no difference for anyone to resolve. Scope doctrine translated those simple observations into enforceable chamber rules, and each chamber wrote the doctrine into its own text in its own way.

Senate Rule XXVIII, paragraph 3(a), stated the Senate version by its terms. Under that paragraph, conferees shall not insert matter not committed to them by either House, nor strike matter agreed to by both Houses. The two prohibitions faced in opposite directions, and the difference between them mattered. The first prohibition guarded the entrance to the conference. Matter that neither chamber had passed could not enter a bill through the conference door, because no chamber had voted on it in the form in which the conference would present it and no disagreement over it existed for conferees to settle. The second prohibition guarded the exit from the earlier stages of the process. Matter that both Houses had agreed to had already survived the full sequence of passage in each chamber, and conferees held no commission to undo a settlement that the two chambers themselves had reached. Read together, the two clauses described a committee with a narrow warrant: work on the gap between the two texts, and leave everything outside that gap alone.

The House rule addressed the same problem through a closely related set of terms, as described in House Practice and in How Our Laws Are Made, the guide issued by the House Parliamentarian. In its terms, the House rule barred conferees from inserting matter not committed by either House, and it barred them from modifying a specific matter committed to conference beyond its scope as committed. Germane modifications were permitted. The House formulation added a refinement that the bare prohibition on insertion did not capture on its own. A specific matter could reach conference in a defined form, with a defined reach, and the commitment itself set the measure of what the conferees could do with it. A narrow provision committed to conference did not become, in the hands of the managers, a license to rewrite the surrounding subject at large. At the same time, the express permission for germane modifications recognized that resolving a disagreement required some room to adjust language, and that a conference forbidden to touch the wording of a committed matter at all could not perform the task for which the chambers had appointed it. The House rule therefore drew two lines rather than one: a line around subjects never committed, and a line around the scope of subjects that had been committed in a particular form.

A second latitude ran alongside those lines and applied where a chamber had committed an amendment in the nature of a substitute. The Congressional Research Service addressed that latitude in Report RS22733. Where the matter in disagreement took the form of a substitute, a complete alternative text offered in place of another complete text, the managers held wider room to work than a single narrow provision would have allowed. The governing test, as the report described it, asked whether a modification was sufficiently or reasonably related to the matter committed. The wording of the test did real work. A substitute committed an entire approach to a subject, and an entire approach could not be negotiated clause by clause under a standard that treated every sentence as a separate commitment with its own fixed perimeter. The sufficiently or reasonably related standard instead asked about relationship: whether the change belonged to the subject and the approach that a chamber had actually placed before the conference. Latitude under that test remained latitude within a commitment, rather than freedom from commitment altogether, because a modification unrelated to the substitute had no anchor in anything either chamber had passed.

The distance between the stated rules and the floor practice of the House formed part of the honest account of scope, and the Congressional Research Service described that distance in Report 98-696, Resolving Legislative Differences in Congress. In House practice, the Rules Committee commonly reported special rules waiving points of order against a conference report and against its consideration. A point of order was the enforcement device that gave a scope rule its effect on the floor: a member raised the objection, the presiding officer ruled, and a report that violated the rule faced the consequence the rule prescribed. A waiver removed that device for a particular report before debate on the report began. The rule on scope stayed written in the chamber rules, and the special rule for the individual measure set the objection aside for that measure. Both facts belonged in any accurate description. The House possessed a scope rule stated in restrictive terms, and the House also possessed a routine procedural instrument through which a majority could proceed with a conference report despite an objection that the rule would otherwise have sustained. Scope in the House therefore operated as a default that the majority could displace measure by measure, rather than as a barrier that no majority action short of changing the standing rules could move.

The Senate built a distinct enforcement regime around a distinct abuse, and Senate Rule XLIV, paragraph 8, supplied the relevant definition. That paragraph addressed new directed spending provisions, and the Congressional Research Service, in Report RS22733, recorded the colloquial label attached to such provisions: airdropped. The image inside the label described the procedural complaint with unusual precision. An airdropped provision arrived in a conference report without having passed either chamber beforehand. It had not been debated and approved as part of the House text, and it had not been debated and approved as part of the Senate text. It appeared at the stage of the process where the chambers faced the conference report as a finished package, a stage built for accepting or rejecting a settlement of disagreements rather than for examining new spending direction sentence by sentence. Directed spending sharpened the concern because the provision pointed resources toward a specified recipient or purpose, so the absence of any earlier chamber vote attached to a benefit that was concrete and identifiable rather than general and diffuse. The definition in Rule XLIV, paragraph 8, gave senators a textual category to invoke, and a textual category mattered because a point of order required a rule that the presiding officer could apply to identified language in a report.

The consequences that followed a successful Senate point of order depended on the kind of violation alleged, and the Senate rules treated the two classic violations differently. A point of order raised against matter that conferees had stricken, where that matter had been agreed to by both Houses, struck at the conference report as a whole: sustaining the objection rejected or recommitted the report. The logic of that consequence matched the character of the violation. Deleting agreed matter rewrote a decision that the two chambers had already made together, so the remedy sent the entire settlement back rather than attempting a surgical repair from the floor. A point of order raised against new matter under paragraph 5 operated with a finer instrument. Sustaining that objection struck only the offending matter, and the remainder of the report converted into an amendment that went back to the House. That second remedy preserved the parts of the settlement that stood within the conferees’ commission and removed the part that stood outside it, and the conversion into an amendment returned the surviving text to the ordinary channel through which the two chambers exchanged positions. Waiver of these protections required three fifths of the senators duly chosen and sworn. That threshold placed the scope protections beyond the reach of a simple majority acting alone on the question, and it meant that a Senate majority wishing to keep challenged matter in a report needed support drawn from beyond its own ranks, assuming attendance and party discipline held along expected lines.

The written Senate protections, described in that form, sounded settled, and the documented history of the Senate scope rule showed a period in which they were not settled at all. Between fall 1996 and 2000, rulings of the chair on scope points of order were overturned on appeal, and the accumulated effect of those reversals left the Senate without an effective scope limit. The mechanism of erosion deserved attention because no repeal produced it. The text of the rule survived the period; its enforcement did not. A point of order depended on a ruling, a ruling depended on the presiding officer’s reading of the rule and of precedent, and an appeal placed that reading before the Senate itself, where a majority could set the ruling aside. Each successful appeal weakened the precedential ground on which the next ruling would stand, until a senator contemplating a scope objection faced a rule whose words promised a protection that floor practice no longer delivered. Contemporaries saw the development clearly and named it in print. Bill Dauster addressed the concentration of effective power that unrestricted conference action produced in a 1996 essay in Roll Call titled “Five-Man Senate?”, a title that posed, as a question, the charge that a small group of conferees acting without a scope boundary could exercise a weight in lawmaking out of proportion to the body’s membership and to the votes cast on the floor.

Repair, when it came, arrived through legislation rather than through a simple internal reversal of mood. In December 2000, the Senate scope restriction was reinstated by section 903 of Public Law 106-554. The choice of vehicle carried its own significance for the account of how procedural protections endure. A rule that a majority could erode through appeals acquired a statutory restatement, and the restatement recorded, in a public law, the judgment that conferees required a stated limit after a period in which the limit had failed in practice. Later legislation returned to the subject and pressed further. The Honest Leadership and Open Government Act of 2007, Public Law 110-81, strengthened the rules governing conference conduct and the matter that conference reports could contain. The sequence of erosion and repair, read in order, resisted any simple claim that scope rules, once written, maintained themselves. The 1996 to 2000 period showed a rule hollowed out through the ordinary appellate machinery of the chamber that owned it. The December 2000 reinstatement and the 2007 strengthening showed the opposite movement, in which concern over conference practice produced first a restoration and then a tightening, each carried in enacted legislation with named statutory citations that later readers could trace.

Scope doctrine also reached beyond the chamber floors, because the product of a conference carried weight in later interpretation of the enacted law. A conference report arrived accompanied by a joint explanatory statement, and that statement entered the materials that courts and practitioners consulted as legislative history. Justice Rehnquist, dissenting in Simpson v. United States, 435 U.S. 6, at pages 17 to 18 (1978), treated a joint conference report as highly reliable legislative history, and guidance on finding legislative history sources reflects the place that conference materials held among the documents consulted after enactment. That interpretive weight raised the stakes of the scope question in a way that a purely internal account of chamber procedure could miss. Matter inserted beyond the scope of a commitment did not merely appear in a bill without an earlier chamber vote. The managers’ explanation of the settlement could also frame how the resulting statute was read afterward, so the small group that drafted the final text held a second kind of influence, exercised through the explanatory record that accompanied the text into law. That dual influence, over the text and over the explanation of the text, explains why a boundary drawn around the managers’ commission protected more than floor procedure alone.

Against the procedural account stood the counter-reading, and an honest treatment of conferences had to give that reading its full force rather than treating the scope rules as self-evident wisdom. The rules against out-of-scope matter answered a documented pattern of abuse, and the pattern had two faces. One face was insertion: provisions entering law through conference without passage in either chamber, the practice that produced the airdrop label and the Rule XLIV definition. The other face was exclusion. For most purposes, conferences met behind closed doors, despite open meeting rules that applied to them. Formal openness governed the meetings as a matter of rule, and the actual conduct of negotiation took place away from public view as a matter of practice. The gap between those two statements paralleled the gap in the House between a written scope rule and a routine waiver, and it paralleled the Senate gap between a written scope rule and rulings overturned on appeal. Across all three settings, the formal text described one institution and the working practice produced another, and readers weighing conferences had to hold both descriptions at once.

The conference on the Medicare legislation of 2003 gave the exclusion face of the pattern a documented and specific form. The political history of Medicare and prescription drug coverage, together with contemporary reporting in The New York Times by Robin Toner and Robert Pear, recorded the shape of the negotiations. Chairman Thomas allowed only Senators Baucus and Breaux, among Democrats, into the day to day discussions. Minority Leader Daschle was excluded. The three House Democratic conferees were reduced to signing or not signing the report. Each element of that account bore on a different part of the conference ideal. The admission of two named Democratic senators showed that participation across party lines existed in form. The exclusion of the minority leader showed that the form stopped where the minority’s own chosen leadership began. The position of the House Democratic conferees showed the endpoint of the process in its starkest version: conferees formally appointed to represent a chamber and a party in negotiation, presented with a finished product and a choice between signature and refusal, with signature itself carrying weight because a report needed the signatures of a majority of the conferees of each chamber to go forward in proper form. Negotiation, in that episode, described the activity of a narrower circle than the conference membership on paper.

The weighing that followed from the full record cut against nostalgia in both directions, and the evidence supported a judgment narrower than either celebration or condemnation of the institution. In favor of conferences stood features that no alternative route replicated in the same combination. A conference produced a public record: a report, an explanatory statement, named conferees whose signatures appeared on the result, and a text that each chamber voted on as the settlement of identified disagreements. Named accountability attached to that record. The senators and representatives who signed a report answered for its contents in a way that participants in informal, unrecorded negotiation between the chambers did not, and the scope rules, the airdrop definition, and the point of order machinery gave the parent chambers instruments for policing the agents they had appointed. Against those features stood the documented capacity of the same institution for exclusion and for inserted matter. Closed doors for most purposes, a minority leader shut out of the Medicare negotiations in 2003, appointed conferees confined to signing or refusing, agreed matter placed at risk before the repair legislation of 2000 and 2007, and new matter entering reports until definitions and enforcement provisions gave objectors a usable text: that record belonged to the same institution that produced the signed reports and the explanatory statements. The scope rules were the chambers’ answer to that record, written by institutions that had watched their own agents exceed a commission and had chosen, after a period in which the Senate limit failed outright, to state the commission’s boundary again and to arm it with consequences.

The Disappearance in Numbers and What Replaced the Conference

How far did formal conferences fall?

Formal conferences fell from 38 in the 106th Congress, covering 1999 to 2000, to 9 in the 110th Congress, covering 2007 to 2008, per counts compiled from House and Senate records in an analysis by the Center for Congressional and Presidential Studies at American University.

That pair of counts gives the decline a hard spine, and the rest of the evidence gathers around it in separate units that deserve separate handling. A conference report is a document. A formal conference is an event. A conference committee is a body that gets appointed and either reports or dissolves. The sources counted different things, and honest reading keeps those things apart rather than melting every figure into a single invented series. Read with that discipline, the record tells one consistent story. The conference went from a routine destination for hard legislation to an exceptional one, and then to a rarity. Each later measurement, taken with its own unit and its own source, sat at or below the measurement before it.

The starting point matters because the conference was never the route for most legislation, and exaggerating its former reach makes the decline impossible to measure honestly. Park, Smith and Vander Wielen, in The Changing Politics of Conciliation, treated conferences as the predominant route for complex legislation in the 1970s and 1980s, and that same study described the near evaporation of conference committees as the central change to explain. Predominant for complex legislation is a claim about weight rather than volume. The measures that reached a conference in that period were the measures with the deepest disagreements, the longest texts, and the most tangled trades between the chambers.

Set against that qualitative strength, the volume figure looks modest, and the modesty is part of the record. Donald Wolfensberger, in an analysis at the Center for Congressional and Presidential Studies, found that conferences were used for rarely more than about a quarter of public laws even in their strong period. About a quarter is a ceiling from a careful analyst, not a floor, and it disciplines every later comparison. Most public laws never needed a conference at all. The conference lived in the contested remainder, and within that remainder it held a commanding position during the 1970s and 1980s. It is the loss of the device that once processed the hardest disagreements, measured against a baseline where even success meant roughly one public law in four at the very most.

The documentary count from the middle of the 1990s shows the institution operating at a scale that later Congresses never approached. Congress.gov conference report records count 67 conference reports in the 104th Congress, which ran from 1995 to 1997. Conference reports are the output unit, distinct from conferences convened, because a conference that never agrees files nothing, and the count of reports therefore tracks completed reconciliations rather than attempted ones. It also describes a public archive. Each of those 67 reports entered the record as a document that members, journalists, agencies, and courts could read, cite, and argue about.

Between that Congress and the end of the following decade, the counts compiled from House and Senate records in the analysis by the Center for Congressional and Presidential Studies at American University registered the steepest part of the fall. Formal conferences numbered 38 in the 106th Congress, covering 1999 to 2000, and 9 in the 110th Congress, covering 2007 to 2008. The distance between 38 and 9 is the distance between a practice and an exception. At 9, a Congress could go months between conferences, and the appointment of conferees became an event worth remarking on rather than a step in a familiar sequence. The same counting approach, applied to House and Senate records at both ends of the span, found a practice reduced to less than a quarter of its former frequency within eight years.

The appropriations exception at the end of that span reveals how uneven the disappearance was across kinds of legislation. In the first sixteen months of the 111th Congress, ten conferences were held, and all ten were on appropriations measures, per the Wolfensberger analysis at the Center for Congressional and Presidential Studies. Ten conferences in sixteen months sounds, at a careless glance, like the beginning of a recovery, and the subject matter destroys that reading. Appropriations measures carry deadlines and consequences for inaction that most authorizing legislation does not, and the spending process retained a pull toward formal reconciliation that other processes had lost. Outside appropriations, the count for those sixteen months was zero. A device that survives only inside the one process where delay carries an automatic penalty has not survived as a general institution.

The endpoint of the measured arc belongs to the Congressional Research Service, and it must stay anchored to its edition rather than float as a loose claim about any later Congress. The 2018 edition of Report 98-696, Resolving Legislative Differences in Congress, recorded that the number of conference committees held in each Congress had decreased to fewer than five. Fewer than five is a different kind of statement from the precise counts earlier in the arc. Anchored to the 2018 edition, the figure closes the arc that began with 67 conference reports in the 104th Congress.

What replaced the conference in practice?

Leadership negotiation replaced the formal conference as the practical route to identical text. Leaders and staff assembled a finished bargain inside leadership offices, and the chambers then resolved that bargain through acceptance of one text without change or through a bounded exchange of amendments between the two chambers.

Substitution, rather than paralysis, is the fact to hold onto through every consequence that follows. Legislation did not stop reconciling its differences because the conference emptied out. The House and the Senate went on passing laws that required identical text, and the work of producing that identical text moved into rooms with fewer occupants, fewer procedural steps, and no obligation to file a joint account of what had been traded. Leadership negotiation names the destination of that moved work. The negotiation that a conference once conducted through appointed conferees, operating under instructions, scope limits, and reporting requirements, migrated upward into the offices of party leaders in both chambers, where the terms of agreement were settled among a small circle and then presented to each chamber as a finished product. What they lost was their place inside the negotiation itself, because by the time a leadership text reached either floor, the bargaining had ended and the remaining choice had narrowed to approval or rejection.

Acceptance is the purest form the substitution took, and the health care measure of 2010 supplied its clearest demonstration. In March 2010, the House accepted the Senate passed health care text, and a separate reconciliation measure adjusted the framework, with no conference held. Under acceptance, one chamber takes the other chamber text whole, swallowing provisions its own passed version had rejected or never considered, because opening the text to amendment would restart a process the leaders had decided to avoid. The separate reconciliation measure in that instance did the adjusting work that a conference report might otherwise have done, reshaping parts of the framework through a second vehicle while the main text moved untouched. It required only a calculation, made in both leadership offices at different moments in the same fight, that a formal conference exposed the measure to procedural dangers that acceptance avoided entirely. The chamber whose text gets accepted wins a victory so complete that its earlier compromises, struck for its own internal reasons, become the law without further negotiation.

Amendment exchange supplied the second vehicle, more flexible than acceptance and bounded by design. Under amendment exchange, the chambers send amendments back and forth until both agree on one text, and parliamentary practice limits the rounds, since each chamber gets one opportunity to amend the amendments of the other and amendments beyond the third degree are not permitted. Congressional Research Service Report RL34611 measured the scale of that route across a long span and found that amendment exchange resolved roughly 11 to 24 percent of the public laws enacted between 1992 and 2007. In some Congresses within that span the exchange carried about a ninth of enacted public laws, and in others it carried nearly a quarter, which means the substitute was never a marginal improvisation. It was a standing route with a substantial share of completed lawmaking throughout the very years when formal conferences were falling from 38 in one Congress to 9 in another.

The difference between a conference and an amendment exchange managed from leadership offices looks technical, and the technical difference conceals an institutional one. Conferees negotiate with authority to trade, within the scope of the differences committed to them, and their product returns as a conference report that both chambers must accept or reject whole. An amendment exchange, by contrast, moves a text between floors where leaders control recognition, scheduling, and the drafting of each successive amendment. The drafting pen stays in fewer hands for longer. A member who objects to one element of a leadership assembled package faces a structured choice among amendments that leadership has already screened, rather than a seat at a table where the package itself gets built.

The health care fight exposed the motive in words attributed at the time, and the attribution deserves exact handling. Government Executive reported in January 2010 that Democratic leadership aides attributed the decision to bypass a formal conference on the health care measure partly to Republican pledges to use the procedural hurdles a conference required. Aides speaking for the majority leadership named obstruction as the reason, and the minority pledges they pointed to were pledges about the conference process itself, about the motions and steps that going to conference placed in the path of a measure before any conferee sat down. A conference was not refused because its negotiating table lacked value. It was refused, on the majority account, because reaching the table required surviving a series of procedural exposures, and the minority had announced an intention to exploit those exposures. The bypass followed from the price of admission rather than from any judgment that bargains struck in leadership offices were better bargains.

The documentary consequence of the substitution arrives first because it shapes every other consequence. A completed conference produces a conference report and a joint explanatory statement, and together those documents give the final bargain a public form, a text with an accompanying explanation of what the conferees agreed and why the agreed language reads as it does. Leadership negotiation produces no equivalent pair. Where the final text emerges from leadership offices and gains approval through acceptance or amendment exchange, no conference report exists for the final text and no joint explanatory statement accompanies it, so no single public document records the negotiation. Fragments exist. Floor statements exist. Amendment texts and voting records exist. What does not exist is the one authoritative account, filed as part of the legislative process itself, that states the terms of the settlement in the negotiators own collective voice. The archive that results is thinner, and the thinness falls hardest on anyone who later needs to reconstruct why a provision reads one way rather than another.

Courts feel that thinness in a specific and historically grounded way, because the substitution steadily withdrew an interpretive resource that had carried unusual authority. A judge construing ambiguous statutory language searches for materials that speak for the enacting Congress with some authority, and a conference report with its joint explanatory statement spoke with unusual authority because it issued from the body that had settled the text and because both chambers acted on the measure in reliance on it. Floor debate under leadership management carries no comparable weight, since individual statements may conflict, may go unanswered, and may describe a bargain that the speaker had no part in making.

Members below the rank of leader lose something different, and more personal to the institution. Conferees held seats. Appointment as a conferee gave rank and file members, and members of the minority party in each chamber, a formal place inside the settlement of a measure, with access to the offers, the drafts, and the reasoning as the bargain moved. The minority presence at a conference was never equal to the majority presence, and conferee selection always favored the majority, yet the structure guaranteed the minority a defined role rather than a spectator position outside a closed door. It simply stopped convening the body in which the seats existed. A minority member confronting a leadership text under acceptance confronts a finished product with a vote at the end, and a majority member outside the leadership circle confronts the same product from a similar distance.

Accountability follows the bargaining, and the bargaining moved into leadership offices. When conferees settled a measure and filed a report, responsibility for the settlement had names, signatures, and a document attached to it, and dissatisfied members knew which colleagues had traded what. Under leadership negotiation, responsibility concentrates in leadership offices in both parties, where the circle of decision is small, the internal deliberations leave no report, and the connection between any single provision and any single decision maker can dissolve into the anonymity of a package. Each party used the blockage tools against conferences while in the minority, treating every procedural step toward a conference as an opportunity to delay, expose, or defeat the majority measure. Each party then regretted the same blockage while in the majority, discovering that the tools perfected against an opposing majority worked with equal efficiency against its own. It is an institutional equilibrium, held in place because the party out of power gains too much from the pricing of conferences to surrender the tools, and the party in power gains too much from leadership control of text to rebuild the older venue at the cost of its own convenience.

Three misunderstandings cling to the disappearance, and the record corrects each of them. The first holds that the conference vanished because members grew lazy, as if a demanding institution decayed through a failure of character or appetite for hard work. Leadership negotiation is laborious work, amendment exchanges require careful drafting and sequencing, and holding a majority together behind a take it or leave it text demands constant attention from leaders and staff. The conference disappeared from a calculation about procedural exposure and control, not from sloth, and any account built on declining diligence fails to explain why the same members who avoided conferences worked so intensely inside the substitute process that replaced them.

The second misunderstanding holds that conferences were abolished, struck from the rules by some reform that a later reform could reverse with equal simplicity. No rule abolished them. The conference remained available throughout the decline, authorized by the same practices that had sustained it during its strong period, and measures continued to reach conference whenever leaders judged the price acceptable, as the appropriations conferences of the 111th Congress demonstrated. Conferences were priced out at the door. The procedural steps required to convene one, each step an occasion for delay and obstruction in the Senate especially, raised the admission cost until rational leaders, acting on ordinary incentives, chose cheaper routes that delivered the same constitutional product.

The third misunderstanding concerns the repair attempted in 2013 and holds that the compound motion restored the conference by making the road to it cheaper. In 2013 the Senate adopted a compound motion approach that packaged the steps for going to conference, reducing the number of separate procedural exposures a measure had to survive before conferees could be appointed. The price fell. The price did not reach zero, because a compound motion remained a motion, debatable and blockable in ways that acceptance and amendment exchange were not, and the vote to proceed toward a conference remained a vote the minority could contest. At the same time, the substitutes kept their advantages. Leadership negotiation retained its control over drafting, its freedom from a joint explanatory statement that might constrain later argument, and its habit, entrenched by then across several Congresses, of treating leadership text as the normal starting point for final passage. A cheaper conference competing against entrenched substitutes with real advantages gains only the cases where its remaining price is worth paying, and the measurement that followed, fewer than five conference committees per Congress in the 2018 edition of Report 98-696, records that the restoration never arrived.

The article leaves a reader with a test rather than a lament, because lament cannot distinguish a genuine conference from a process wearing its name, and the distinction decides how any future measure should be read while it moves. The test counts the blockable steps between disagreement and negotiation, asks who controls the text at each step, and asks what document will record the bargain. A process with few blockable steps, where appointed conferees from both chambers control the drafting inside the scope of the differences and file a conference report with a joint explanatory statement at the end, is a conference in substance whatever difficulties attend it. A process where the steps multiply, where leadership offices hold the drafting pen from first offer to final text, and where no joint document records the settlement, is leadership negotiation in substance whatever title the participants give it. Where the blockable steps multiply and no document results, leadership negotiation has replaced the conference regardless of what the process is called.

What the Vanishing Conference Changed

The disappearance of the conference is a story about a doorway rather than about a room. Inside the room, the institution worked as its rules described: managers appointed from the committees of jurisdiction, negotiations bounded by the scope of the differences, a report signed by a majority of each delegation, an explanatory statement issued with the text, and an unamendable vote on the floors. The failure occurred at the entrance, where Senate procedure placed three debatable motions between disagreement and negotiation. Each motion handed any determined senator a separate occasion for delay, and the accumulated price of the occasions taught majority leaders of both parties to negotiate without ever asking the Senate for permission to negotiate.

The 2013 consolidation of the three motions into one compound motion, with a two hour limit on debate before a cloture vote, cut the entrance price without removing it. A majority that could command three fifths of the Senate could reach a conference in a single step; a majority that could not command three fifths faced the same closed door in a new frame. Leadership negotiation asked for no threshold at all. The text could be settled in the majority leader’s office, carried to the floors as an amendment exchange or as an acceptance, and sequenced around every other item of floor business. The end of session bargain, which the conference once housed in a named committee with named members, migrated into consolidated must pass vehicles, and the treatment of omnibus bills and riders describes the packages that absorbed the negotiations.

The honest ledger records losses on both sides of the change. The conference system in its strong decades produced airdropped provisions that neither chamber had passed, excluded minority conferees from the rooms where the text was written, and required two rounds of scope legislation, in 2000 and in 2007, to discipline its own managers. Those abuses were no invention of the system’s critics; the rules against them stand in Senate Rule XXVIII and Senate Rule XLIV precisely because the practices occurred. What vanished with the conference was not an innocent institution. What vanished was a documented one, in which the final bargain of a major measure existed as a public text with named authors and a signed explanation.

The substitution also redrew the map of influence inside each chamber. Conferees drawn from the standing committees once carried the committees’ work into the final settlement, and a seat at the conference table gave rank and file members and the minority party a formal place in the last stage of lawmaking. Leadership negotiation offers no seat, no signature, and no statement. Members confront a finished text and a choice between acceptance and blame for failure, a choice that recurs on measure after measure while the conference door stays shut. The three motions explain how the door closed; the practice that replaced the conference explains why no majority holding a working substitute has paid to reopen it.

The record loss compounds across Congresses. Each conference that never met is a joint explanatory statement that was never written, and the accumulation of unwritten statements thins the legislative history available for every statute enacted by substitution. Administrators construing an ambiguous provision, litigants arguing over its reach, and courts weighing the materials all work with floor statements and committee reports that describe earlier stages of a measure rather than its final bargain. The conference report was the one document that described the final bargain itself, in the words of the negotiators who struck it. Its absence is the quietest and most durable consequence of the price charged at the Senate door.

The procedural vocabulary of bicameral resolution rewards organized notes, since the same handful of motions, rules, and report forms recurs across every Congress: readers tracking the three motions against the two chamber rule books can keep your statute notes, citations, and case chronologies together free on VaultBook, and students preparing for examinations on the legislative process can practice and revise US government and civics material on ReportMedic.

The Three-Route Table

The three routes differ in who controls the text, in what can block them, and in what each one leaves behind.

Route	How identical text is reached	Procedural steps	Vulnerabilities	Documentary record
Accepting the other chamber’s text	One chamber passes the other chamber’s version without any change	Passage by the originating chamber, then passage of the same text by the second chamber, then enrollment	The accepting chamber surrenders all influence over wording and must take objectionable provisions whole or lose the measure	A vote in each chamber and no negotiated document
Exchanging amendments between the houses	The chambers amend each other’s amendments until both agree on one text	Messages between the chambers, amendments in successive degrees, motions to concur, insist, and recede	Parliamentary practice ends the exchange after a limited number of rounds, and Senate amendments reach the House floor only under managed procedures	Messages, amendments, and votes scattered across the records of two chambers
Convening a conference committee	Conferees negotiate the matters in disagreement and report a single text to both floors	Disagreement, request for or agreement to a conference, appointment of conferees, negotiation, signed report, unamendable floor votes	Three debatable Senate motions at the entrance, scope points of order against the report, and a debatable report in the Senate	A conference report and a joint explanatory statement signed by a majority of the managers of each House

Frequently Asked Questions

Q: Why has the conference committee nearly disappeared?

The formal route to conference in the Senate historically required three separate motions, to disagree or insist, to request or agree to a conference, and to authorize the appointment of conferees. Each motion was debatable, so a single senator could object to the unanimous consent that normally bundled them and force up to three filibusters, each curable only by cloture at three fifths of the Senate. Majority leaders of both parties concluded that the floor time and the sixty vote threshold cost more than the conference was worth, and leadership negotiation offered a substitute with no entrance fee. Counts compiled from House and Senate records in an analysis by the Center for Congressional and Presidential Studies at American University recorded the result: formal conferences fell from 38 in the 106th Congress to 9 in the 110th Congress.

Q: What are the three motions needed for a conference committee?

Senate procedure treated going to conference as three distinct actions, and for most of the conference era each required its own motion. The first motion put the Senate in disagreement with the House, by disagreeing to the House amendments or insisting on the Senate amendments. The second motion requested a conference with the House, or agreed to a conference the House had requested. The third motion authorized the presiding officer to appoint the Senate conferees. The Congressional Research Service described the three motions in Report RS20454, and noted that they were usually combined and adopted by unanimous consent. Senate Resolution 16 of the 113th Congress later combined the three actions into one compound motion under Senate Rule XXVIII, paragraph 2.

Q: What is ping pong instead of a conference committee?

Ping pong is the colloquial name for resolving bicameral differences by amendments between the houses instead of by conference. One chamber amends the other chamber’s text and sends it back, the second chamber amends the amendment and returns it, and the exchange continues until both chambers have agreed to a single text. Parliamentary practice bounds the game. Each chamber gets one opportunity to amend the amendments of the other, and amendments beyond the third degree are not in order, so the exchange has a limited number of rounds. The Congressional Research Service measured the route in Report RL34611 and found that amendment exchange resolved roughly 11 to 24 percent of the public laws enacted between 1992 and 2007.

Q: Can a conference committee add new material?

The rules forbid the practice in terms. Senate Rule XXVIII, paragraph 3, provides that conferees shall not insert in their report matter not committed to them by either House and shall not strike matter agreed to by both Houses, and House Practice states the same limits for House managers. Conferees hold latitude inside the disagreement itself: where a chamber committed an amendment in the nature of a substitute, the managers may report a germane modification of the subjects in disagreement. Matter that crosses the boundary exposes the report to a point of order, and waiver of the Senate points of order requires three fifths of the senators duly chosen and sworn.

Q: Is a conference committee report amendable?

A conference report cannot be amended in either chamber. The report returns to the House and the Senate as a single agreed text, accompanied by a joint explanatory statement, and each chamber votes the package up or down as a whole. The privilege of the report and the bar on amendments protect the bargain the conferees struck, since a floor amendment in one chamber would unravel the settlement and produce yet another disagreement. The choices open to a chamber are adoption, rejection, and, in the first chamber to act, recommittal to the conference. Points of order against out of scope matter operate under separate rules and do not convert the report into an amendable text.

Q: Why must both chambers pass identical text without a conference committee?

The requirement comes from the Constitution rather than from any conference rule. Article I vests the legislative power in a Congress of two chambers and provides that every bill that has passed the House of Representatives and the Senate shall be presented to the President before it becomes law, a sequence the Congressional Research Service reads to require approval of the same bill in precisely the same form. Two different texts cannot be presented, enrolled, or signed, because no single measure would have passed both chambers. A conference committee is only one device for reaching the required identity. Acceptance of the other chamber’s text and amendment exchange reach the same destination, and the Constitution is satisfied by whichever device produces one text approved by both.

Q: Who appoints members to a conference committee?

Appointment rests formally with the presiding officers. The Speaker of the House appoints the House managers, and in the Senate the presiding officer appoints conferees under the authority granted by the third of the Senate’s conference motions. Practice fills in what the rules leave open. Appointments follow the recommendations of the committees that considered the measure, and conferees are drawn largely from the membership of those committees, usually their senior members, since the conference defends and revises the text the committees produced. The two delegations need not be equal in number, and each delegation acts by its own majority when the conference votes and signs its report.

Q: What is an airdrop in a conference committee?

An airdrop is the colloquial name for a provision inserted into a conference report that appeared in neither chamber’s version of the measure. The Congressional Research Service used the term in Report RS22733 for matter the conferees added on their own authority rather than drawn from the disagreement committed to them. Senate rules answer the practice with points of order. Rule XLIV, paragraph 8, supports a point of order against a new directed spending provision in a conference report, and Rule XXVIII, paragraph 5, supplies the remedy for inserted matter generally: the offending provision is stricken and the remainder of the report is converted into an amendment returned to the House.

Q: What happens if one chamber rejects a conference report?

Rejection sends the measure back into disagreement rather than into law. If the first chamber to consider a conference report rejects it, the chambers return to the position they held before the conference reported, and the ordinary devices remain available: the chambers may request a further conference with a new set of managers, resume amendment exchange, or let the measure fail. Adoption by the first chamber changes the position, because the conference is then dissolved and the second chamber faces the report without any power to recommit it. A report rejected in the second chamber after adoption in the first ends the measure in the form the conference proposed, and any revival begins from the texts the chambers had passed.

Q: What is the joint explanatory statement that accompanies a conference report?

The joint explanatory statement is the document that accompanies a conference report and explains the agreement the managers reached. House Rule XXII, clause 7(e), requires the statement to be signed by a majority of the managers of each House, the same signature threshold that validates the report itself. The statement records how the conferees settled the matters in disagreement, which provisions of each chamber’s version survived, and what the settled language was understood to mean. Because the statement issued from the negotiators of both chambers and reached the floors with the text, courts and practitioners treated conference materials as legislative history of high standing for the enacted measure.

Q: Are conference committee meetings open to the public?

House rules have required open conference meetings since the reforms of the 1970s. House Rule XXII, clause 6, provides that conference meetings are open to the public unless the House votes by roll call in open session to close a particular meeting. House Practice attaches a sharp consequence to the requirement: a failure to hold the required formal open meeting supports a point of order against the conference report, under which the report is rejected and a request for a new conference is automatically made. Senate Rule XXVIII likewise addresses open meetings. In practice, much conference bargaining historically took place in informal sessions among subsets of the managers, a gap between the formal rule and the working habit that critics of the institution cited often.

Q: Can a conference report be filibustered in the Senate?

The report itself can be filibustered even though reaching it cannot. Senate Rule XXVIII, paragraph 1, makes the presentation of a conference report always in order and provides that the question of proceeding to its consideration, if raised, is decided without debate. Debate on the report is a different matter. Senators may debate a conference report at length, and a majority that lacks three fifths of the Senate for cloture can be kept from voting on adoption. One statutory exception stands out: the Congressional Budget Act limits Senate debate on conference reports for budget resolutions and reconciliation bills to ten hours, which places those reports beyond the reach of a filibuster.

Q: What is the scope of the differences that conferees may consider?

The scope of the differences is the territory committed to the conferees: the provisions on which the House and Senate versions disagree, together with the range between the two positions each chamber took. Conferees may settle any point inside that territory, split a numerical difference, blend two approaches to the same subject, or report a germane modification where a chamber committed an amendment in the nature of a substitute. The boundary excludes two things: matter that neither chamber committed, which the managers may not insert, and matter that both chambers agreed upon, which the managers may not strike. Senate Rule XXVIII, paragraph 3, states the boundary, and the Congressional Research Service described the working test as whether added matter is sufficiently or reasonably related to what was committed.

Q: Are instructions to conferees binding on a conference committee?

Instructions do not bind a conference committee in either chamber. The House may adopt a motion to instruct its managers after it decides to go to conference and before the Speaker appoints them, and the Senate may adopt motions to instruct its conferees after the conference motions and before appointment, but the instructions operate as advice from the floor rather than as orders. House Practice states the consequence plainly: no point of order lies against a conference report on the ground that the managers departed from their instructions. The floors retain other remedies, including rejection of the report and a privileged motion to instruct when a conference sits for 45 calendar days and 25 legislative days without reporting.

Q: What is the stage of disagreement between the chambers?

The stage of disagreement is the procedural position a measure reaches when one chamber has amended the other chamber’s text and the originating chamber has declined to concur, so that the two chambers stand committed to different versions of the same measure. The position matters because the formal devices of bicameral resolution open at that stage. Motions to go to conference are in order once the chambers are in disagreement, instructions to conferees belong to the interval between disagreement and appointment, and the scope of any conference is defined by the disagreement itself. Before disagreement, a chamber may only accept or amend; after agreement on a single text, no device is needed at all.

Q: Can a conference report be sent back to the conference committee?

A conference report can be recommitted, but only within a narrow window. The first chamber to take up the report may vote to recommit it to the conference committee, sending the managers back to revise their settlement. Once the first chamber adopts the report, the conference committee is dissolved, and the second chamber to act has lost the option: it must adopt or reject the text as reported. The asymmetry protects the bargain at the moment it begins to harden into law. A chamber that has already accepted the settlement cannot be made to renegotiate it by the later maneuvering of the other.

Q: Who has to sign a conference report before it can be filed?

A conference report carries the signatures of a majority of the managers of each House, and it cannot be filed without them. The requirement applies to the joint explanatory statement as well under House Rule XXII, clause 7(e). The managers of each chamber vote separately inside the conference, so the signature rule does double work: it certifies that a majority of the House delegation and a majority of the Senate delegation each accepted the settlement, and it prevents a report that speaks for only one chamber’s conferees from reaching either floor. A manager who refuses to sign does not block the report, provided the majority of the delegation signs.

Q: When did the use of conference committees begin to decline?

The counts show a strong institution through the middle of the 1990s and a steep fall across the following decade. Congress.gov conference report records count 67 conference reports in the 104th Congress, covering 1995 to 1997. Counts compiled from House and Senate records in an analysis by the Center for Congressional and Presidential Studies at American University then recorded 38 formal conferences in the 106th Congress and 9 in the 110th Congress, covering 2007 to 2008. The same analysis found only ten conferences in the first sixteen months of the 111th Congress, all on appropriations measures. The Congressional Research Service, in the 2018 edition of Report 98-696, recorded fewer than five conference committees per Congress.

Q: How long may a conference committee sit without reporting?

No general deadline forces a conference to report, and a conference that never agrees simply never files a report. The House built one pressure device into its rules. After a conference has been appointed for 45 calendar days and 25 legislative days without reporting, a privileged motion to instruct the House managers becomes in order, giving the floor a formal occasion to direct advice at its negotiators and to air the delay. Instructions remain nonbinding even then. The practical limits are political rather than parliamentary: the measure waits in disagreement, the Congress moves toward adjournment, and the chambers retain the option of abandoning the conference for amendment exchange or acceptance.

Q: What should a student remember about conference committees for an exam?

The essentials reduce to a destination, three routes, and a price. The destination is identical text: Article I requires both chambers to approve the same bill in the same form before presentment. The routes are acceptance of the other chamber’s text, amendment exchange, called ping pong, and the conference committee. The price explains the pattern of use: going to conference in the Senate required three debatable motions, one objection could multiply them into separate filibusters, and leaders substituted negotiation under their own control. A complete answer adds the report’s features: privileged, unamendable, signed by a majority of each delegation, accompanied by a joint explanatory statement, and bounded by the scope of the differences.