The Starting Position

On the morning of December 24, 2009, the Senate passed the largest health care bill in more than four decades, and the Democratic leadership that had steered it did not expect the text to become law in the form they were voting on. The plan was the ordinary one. The Senate would pass its version, the House would pass its version, and a conference committee would reconcile the two into a final statute. What made this bill different was that the ordinary plan never happened. The Senate’s draft became the final draft because the Senate’s arithmetic gave its leaders no second chance to touch it.

That is the problem this article names the unamendable-draft problem. The Patient Protection and Affordable Care Act bears the marks of a text written as an opening Senate offer rather than a finished product. Its structure reflects two committee bills that were merged in a leadership office, its vehicle was a six-page House bill about homebuyer tax credits for service members, and its final Senate vote was shaped to survive a filibuster rather than to invite amendment. None of this was accidental. Each choice responded to a constraint, and each constraint removed an option that the process would otherwise have kept open.

The stakes of the history were large enough to justify the contortions. The Senate Health, Education, Labor and Pensions Committee’s own press materials put the uninsured population at nearly 50 million Americans. News coverage of the committee’s July vote used a figure of 46 million. The bill was the most ambitious attempt to extend coverage since the creation of Medicare and Medicaid in 1965, and its coverage expansion would eventually carry a Congressional Budget Office score of roughly $848 billion over a decade. Everything about the process described here was downstream of that ambition colliding with the Senate’s rules.

This article promises a single argument, made in procedural order. First, the Senate did not write one bill. It wrote two, in two committees with different jurisdictions, different chairmen, and different political mandates. Second, Majority Leader Harry Reid fused those two bills into a leadership substitute that neither committee had voted on, because no single committee product could have assembled sixty votes. Third, the Senate draped that substitute over the hollowed-out shell of a House bill to satisfy the Constitution’s demand that revenue measures originate in the House. Each of these steps is usually told as background. Told in sequence, they are the explanation. The bill that became law reads the way it reads because the Senate could only get it to the floor by writing it twice, merging it once, and hiding it inside a bill about something else entirely.

The unamendable-draft problem also explains why the final statute contains provisions that read like committee compromises frozen mid-argument. A bill designed for conference carries placeholders, overlapping provisions, and tax schedules calibrated to satisfy one set of negotiators while leaving room for another. When conference disappeared as an option, those features hardened into law. The Senate text had been written with the expectation that House negotiators would rewrite it. They never got the chance.

The claim also carries a warning about how statutes are read. When a text written as a first Senate draft becomes final law without a conference rewrite, its ambiguities, overlaps, and placeholder compromises become the material that courts, agencies, and regulated parties must interpret. The history of the bill’s passage is therefore not decoration around the policy. It is part of the policy’s operating manual, because the reason a provision reads the way it does often lies in a procedural bargain rather than a substantive judgment.

One more preliminary matters. The Senate’s majority ran on arithmetic, and the arithmetic ran on a person. When Edward Kennedy died on August 25, 2009, the Democratic caucus lost the sixtieth vote it needed to break a filibuster on the health bill. Governor Deval Patrick’s appointment of Paul Kirk as interim senator on September 24 restored the number. The committee merger and the floor fight that followed happened inside that restored sixty, and every maneuver in the article’s span was an attempt to hold it. The number sixty is the silent character in every scene that follows.

Two Committee Bills, One Leadership Substitute

The Senate’s health care effort began not with a bill but with a jurisdictional split. The Health, Education, Labor and Pensions Committee, which Kennedy chaired until his illness removed him from daily management, owned the insurance regulation and labor provisions. The Finance Committee, under Chairman Max Baucus, owned taxes, Medicare, and Medicaid. Neither committee could produce the whole statute alone. The Senate would have to write the bill twice and then combine the two halves, a division of labor that made the merger step as consequential as the drafting.

The HELP Committee moved first. With Kennedy absent, Senator Christopher Dodd of Connecticut ran the committee’s work. Markup began in June 2009, and on July 15 the committee approved the Affordable Health Choices Act by a party-line vote of 13 to 10. The bill was the more expansive of the two Senate products. It required individuals to obtain health coverage, with a financial penalty of up to $750 a year for those who refused and could afford to comply. It imposed shared responsibility on employers, requiring businesses that did not provide coverage to contribute toward the cost of publicly sponsored coverage for their workers, with an exception for employers with 25 or fewer employees. And it created a government-run insurance plan, the Community Health Insurance Option, which would compete alongside private plans inside newly created state marketplaces the bill called American Health Benefit Gateways. Under the bill’s terms, the public plan’s provider payment rates would be negotiated by the Secretary of Health and Human Services and could not exceed the average of all gateway reimbursement rates.

The committee’s Democratic majority presented the bill as a fulfillment of Kennedy’s long career in health policy. Kennedy himself said the committee had done the hard work the American people had sent it to do, and that the search for unifying solutions had to continue as the bill moved to the floor. Dodd, who had managed the markup in Kennedy’s absence, argued that the bill was less expensive than early projections had suggested, pointing to a Congressional Budget Office estimate of less than $615 billion over ten years. The committee’s press materials emphasized that the public option responded to the wishes of the American people for an alternative to for-profit insurance companies, and that the subsidies for purchasing coverage would extend on a sliding scale to families at up to 400 percent of the poverty level, with additional credits for small businesses.

Republicans on the committee attacked the product as unworkable. Senator Judd Gregg of New Hampshire said the Democrats had struck out on health care, arguing that the bill left millions uncovered while burdening businesses in ways that could cost workers their jobs. The criticism foreshadowed the merger problem. The HELP bill was widely described in the press as the more liberal of the two Senate efforts, and Dodd declined to specify which of its provisions could survive combination with a Finance Committee product shaped by different political pressures. The bill also carried a structural limitation that mattered for the merger. The HELP Committee had no jurisdiction over taxes or Medicaid, so it could not pay for what it proposed and could not touch the program that would eventually cover a large share of the newly insured. The Congressional Budget Office’s accounting reflected the gap.

The markup itself illustrated the committee’s political position. Kennedy’s statement noted that the committee had considered hundreds of proposals, accepting principled compromises where it could and treating disagreements with patience where it could not. In practice, the committee’s Democrats moved the measure with almost no Republican support, and the 13 to 10 final vote reflected the partisan shape of the product. Dodd, managing the committee in Kennedy’s absence, framed the vote as historic while acknowledging that its toughest test still lay ahead, in combination with whatever the Finance Committee produced. The press coverage at the time treated the merger as the real legislative event and the HELP vote as a prelude, a reading the committee’s own leadership did little to dispute.

The timing of the July 15 vote carried significance beyond its content. The HELP Committee was the first of the five committees with jurisdiction over health care to approve legislation, which allowed its supporters to claim momentum and set a substantive marker that every subsequent draft had to reckon with. The public option, the employer fees, and the gateway structure all entered the legislative conversation through this bill, and although the Finance Committee rejected the first two, the merger could not simply pretend they had never been proposed. The HELP vote created the left boundary of the negotiation. The Finance vote would create the right boundary. Reid’s job was to find sixty votes between them.

The Finance Committee’s path was slower and more theatrical. Baucus spent the summer of 2009 pursuing a bipartisan compromise through extended negotiations with a group of Republican and Democratic members, and he made concessions designed to keep those talks alive. He jettisoned the public option, and he dropped the requirement that large companies offer insurance to their workers. In his opening statement at markup, Baucus preempted the criticism he knew was coming. He said the plan was no government takeover of health care, that it did not include a public option, that it did not include an employer mandate, and that it was paid for in full. The characterization was his own, and it defined the bill’s political identity.

The Finance Committee considered a Chairman’s Mark released in mid-September, and senators filed 564 amendments before markup concluded. On October 13, the committee approved America’s Healthy Future Act by a vote of 14 to 9, with Senator Olympia Snowe of Maine supplying the only Republican vote. The bill required all legal residents to obtain health insurance or face a financial penalty, established state-based health insurance exchanges where individuals and families between 100 and 400 percent of the federal poverty level could purchase coverage with government subsidies, and expanded Medicaid. It contained no employer mandate in the sense Baucus had disclaimed, though companies with more than 50 employees that did not offer coverage would face penalties when their full-time workers obtained subsidized exchange coverage. To pay for the coverage expansion, the bill relied on the Finance Committee’s jurisdiction over revenue. Its most discussed revenue provision was a 40 percent excise tax on high-cost insurance plans. The Congressional Budget Office put the bill’s ten-year price tag below $900 billion.

The Finance Committee’s deliberations were shaped by the long bipartisan negotiation that preceded them. Baucus had spent months in talks with Republican members, and the Chairman’s Mark he released in mid-September reflected the concessions those talks had extracted. He reduced the proposed penalties for defying the coverage requirement, eased the impact of the excise tax on high-cost plans, and, as his opening statement made explicit, left out both the public option and the employer mandate. The concessions did not buy bipartisan support. Grassley, the committee’s ranking Republican, said the White House and Democratic leaders had short-circuited the talks by imposing a mid-September deadline, calling the move utterly and completely appalling, and he criticized the coverage requirement, the taxes, and its provisions on abortion and immigration. When the committee voted on October 13, only Snowe crossed party lines. The 14 to 9 result meant the Finance product, like the HELP product, was a Democratic measure with a bipartisan veneer, and its path to sixty votes ran through the same leadership negotiation.

The two bills agreed on the core architecture of insurance reform. Both banned insurers from denying coverage because of pre-existing conditions, both created regulated marketplaces for purchasing coverage, and both used subsidies to make that coverage affordable. They differed on who bore responsibility and how the money moved. The HELP bill asked employers to share the cost directly and offered a government plan as a competitor. The Finance bill asked individuals to carry coverage, left employers largely outside the mandate framework, and paid for the subsidies through taxes the HELP Committee had no power to write. Neither bill, on its own, commanded sixty votes on the Senate floor. The HELP bill was too expensive and too regulatory for the chamber’s centrists. The Finance bill had shed the provisions that the chamber’s liberals considered essential. Merging them was not a matter of editorial preference. It was the only way to produce a bill that could survive cloture.

That merger happened in the majority leader’s office, not in a committee room. Reid took the two committee products and negotiated a leadership substitute that reconciled their differences, and he unveiled the result on November 18, 2009. The substitute, which took the title Patient Protection and Affordable Care Act, kept the Finance bill’s overall fiscal framework while adjusting its most politically sensitive revenue provision. The 40 percent excise tax on high-cost plans survived, but Reid raised the thresholds at which it applied, a change that reduced its revenue by roughly $50 billion relative to the Finance Committee version and was aimed at lawmakers and labor unions who had argued that the tax would punish workers who had traded salary increases for generous benefits. To replace the lost revenue, the substitute increased the Medicare payroll tax on high earners. The merged bill carried a coverage-expansion cost of about $848 billion over ten years, and the Congressional Budget Office estimated that it would reduce the federal deficit by $127 billion over the same period while extending coverage to 31 million people who lacked it.

The November 18 unveiling set the terms for everything that followed. The bill was no longer a committee product. It was a leadership product, assembled through private negotiation and presented to the full Senate as a substitute amendment. Senators who had spent months marking up committee bills were now asked to vote on a text they had not written, and the only procedural tool available for moving it was the one that required sixty votes at every step. The merger had solved the committee problem by creating a floor problem, and the floor problem would shape the vehicle, the timing, and ultimately the text of the statute. The unveiling was a beginning rather than an end. The leadership substitute still had to survive the Senate floor, and Reid chose to move it as an amendment to a House-passed vehicle, importing a second procedural history into the legislation’s story. The merger had concentrated the measure’s political risk in a single text controlled by the leadership. The next step would concentrate its parliamentary risk in a single number borrowed from the House.

The Shell Bill

The vehicle for the Senate’s health care bill began its life as an entirely different piece of legislation. H.R. 3590, as introduced, was the Service Members Home Ownership Tax Act of 2009, a six-page bill that modified the first-time homebuyer tax credit for members of the armed services. The House passed it on October 8, 2009, by a vote of 416 to 0 under suspension of the rules, the fast-track procedure reserved for uncontroversial measures. The Senate received it the same day.

Passage under suspension of the rules is worth pausing over, because it explains why the House’s action left no usable legislative history for the health care text. Suspension is the House’s fast track for noncontroversial measures: debate is limited, no amendments are permitted, and passage requires a two-thirds majority rather than a simple one. The 416 to 0 vote reflected genuine consensus on the underlying purpose, tax relief for service members buying homes, and it meant the House never debated, amended, or contemplated the health care provisions that would later travel under the same number. When the Senate later pointed to H.R. 3590’s House passage as the constitutional origin of its health care revenue provisions, the form was impeccable and the substance was empty. Nothing in its text had anything to do with health insurance, hospitals, or taxes on high-cost health plans. That was precisely the point.

The Senate needed a House bill because of the Constitution’s Origination Clause. Article I, Section 7 provides that all bills for raising revenue shall originate in the House of Representatives. The health care bill raised revenue. The excise tax on high-cost insurance plans, the increased Medicare payroll tax on high earners, and the various fees on insurers and other industry participants were revenue provisions, and a bill containing them had to begin its constitutional life in the House. The Senate could not simply introduce its leadership substitute as a new Senate bill without inviting an Origination Clause challenge. The solution was a parliamentary maneuver as old as the Senate itself. Take a House-passed bill, any House-passed bill, strike its entire text through an amendment in the nature of a substitute, and replace it with the Senate’s language. The bill number keeps its House origin. The text becomes whatever the Senate wants it to be.

In this case, the substitute amendment struck the six pages of the homebuyer tax credit bill and replaced them with the 2,074-page health care text. The maneuver is routine. The Senate uses shell bills regularly for measures that must originate in the House, and nothing about the technique violated the Senate’s rules. But its routine character did not make it politically invisible. The sheer disproportion between the original bill and its replacement, six pages of tax relief for service members becoming more than two thousand pages of health care law under the same bill number, gave critics a vivid emblem for their broader argument that the process had been engineered to evade normal legislative scrutiny. The shell became a constitutional talking point, cited by opponents who argued that the statute’s revenue provisions had not genuinely originated in the House in any meaningful sense.

The clause has a specific history. The framers gave the House the first move on revenue because the House, elected every two years, stood closest to the taxpayers who would pay. The Senate retained a broad power to amend House revenue bills, and the shell-bill maneuver lives in the space between those two provisions: the House originates, the Senate amends, and an amendment that replaces the entire text tests how much origination the form alone can carry. Defenders of the technique note that the House did pass H.R. 3590, that the Senate’s amendment power is plenary, and that the House would have further opportunities to act on the legislation before it became law. Critics answer that a six-page tax credit for service members is not a genuine origination of a two-thousand-page health care statute, and that the form has swallowed the substance the framers intended.

The technique matters for understanding the final text in a more practical way as well. A bill that arrives on the floor as a substitute amendment to a shell carries its committee history in a distorted form. The committee reports, the markup records, and the amendment files attached to H.R. 3590 belong to the homebuyer credit, not to the health care provisions that replaced it. The substantive committee work lived in the HELP and Finance Committee bills, which were never themselves brought to the floor. The Senate debated the most consequential domestic legislation in decades under a bill number whose official legislative history described tax relief for service members. For later readers of the statute, including courts and agencies charged with interpreting it, the paper trail of the bill’s construction runs through committee documents for bills that were never enacted, attached procedurally to a vehicle that was never about health care.

The substitution also scrambled the relationship between the measure’s cost estimates and its text. The Congressional Budget Office had scored the HELP and Finance Committee bills as committee products, with the assumptions and baselines appropriate to each. The leadership substitute was a new text, and it required new scoring, which arrived with the November unveiling and framed the public debate over the measure’s deficit effects. The $127 billion in ten-year deficit reduction that Reid cited at his press conference belonged to the substitute, not to either committee product. For a statute whose political survival depended on its deficit arithmetic, the fact that the decisive numbers attached to a leadership draft rather than to a committee product was itself a consequence of the merger.

The shell also shaped the timing of the floor fight. Because the substitute was offered as an amendment to H.R. 3590, the first question before the Senate was not the merits of the health bill but whether to proceed to the vehicle at all. On November 21, 2009, the Senate invoked cloture on the motion to proceed to H.R. 3590 by a vote of 60 to 39, recorded as Record Vote 353. The vote was the first demonstration that Reid’s sixty held, and it set the pattern for everything after it. Every subsequent step, from the consideration of the substitute amendment through final passage, would require the same sixty votes to overcome the same filibuster threat. The motion to proceed did not decide the bill’s content. It decided that the bill would be decided by attrition rather than by amendment, because no amendment could survive without the sixty votes that were already fully committed to the underlying text.

The 60 to 39 tally on the motion to proceed carried its own message. It showed that every Democrat and both independents who caucused with them would vote to advance the legislation, and that no Republican would join them. The precise sixty-vote margin, with no votes to spare, meant that any single defection could halt the measure at any subsequent cloture vote. The leadership would spend the weeks after November 21 negotiating not with the minority but with its own members, purchasing each vote with provisions tailored to individual senators. The motion to proceed had proven the sixty existed. Keeping it together would be the work of the floor fight.

The mechanics of the substitution deserve a closer look, because they explain how completely the Senate’s text displaced the House’s. An amendment in the nature of a substitute strikes everything after the enacting clause of the underlying measure and inserts new language in its place. Offered by the majority leader, such an amendment becomes the pending business of the Senate, and all further amendments are offered to it rather than to the original text. In the case of H.R. 3590, this meant that the six pages concerning the homebuyer credit vanished as a legislative matter the moment the substitute was agreed to, while the number, the House passage date, and the constitutional origin attached to that number survived intact. The Senate’s rules treat the number and the text as separable in exactly this way, which is why the maneuver, however jarring in its proportions, raised no point of order that the presiding officer sustained.

The displacement also clarifies a confusion that the measure’s opponents sometimes encouraged. The charge was occasionally framed as though the Senate had smuggled health care legislation past the House, as if the House never had a say. The parliamentary reality was less conspiratorial and more awkward. The House had passed H.R. 3590 unanimously for its original purpose, the Senate had replaced its contents through a public amendment offered on the floor, and the House would vote again on the resulting legislation before it became law. The objection was not that the House was bypassed. It was that the House’s origination of the revenue provisions was formal rather than substantive, a six-page tax credit standing in for two thousand pages of health care taxes and fees.

There is a final irony in the shell that the unamendable-draft problem makes visible. The Origination Clause exists to keep the power of the purse close to the chamber closest to the people. The shell-bill maneuver honors the clause in form while departing from its spirit, substituting a technical House origin for a substantive one. Whether that departure amounts to a constitutional violation became a question for later litigation and commentary. What is not in question is the practical effect. The Senate’s health care bill reached the floor wearing the number of a bill the House had passed unanimously for an unrelated purpose, and the senators who voted on the motion to proceed knew exactly what they were boarding. The vehicle was empty. The cargo was the entire American health care system.

Why did the Senate need a House bill as its vehicle?

The Constitution’s Origination Clause provides that bills raising revenue must originate in the House of Representatives. The health bill contained new taxes and fees, so it counted as a revenue measure. By amending a bill the House had already passed, the Senate preserved a House origin for text the House had never actually voted on.

What did the two Senate committee bills each contribute?

The HELP bill supplied the regulatory architecture: insurance market rules, employer responsibility provisions, and state purchasing gateways. The Finance bill supplied the fiscal architecture: the exchange subsidy design, revenue provisions, and Medicaid expansion. Reid’s substitute wove the two together around the Finance Committee’s tax-and-spending framework, since only that framework carried a CBO score that fit the budget.

The House Acts First

The House of Representatives moved first in the autumn of 2009. The 2008 elections had given Democrats 258 House seats and a sixty-seat Senate caucus, along with the presidency, and President Obama had made a sweeping health care bill the centerpiece of his domestic agenda. Through the spring and summer, three House committees, Ways and Means, Energy and Commerce, and Education and Labor, held hearings and marked up draft legislation, while an August recess filled with contentious town-hall meetings showed how combustible the issue had become. Critics filled those meetings, and former Alaska governor Sarah Palin charged that a provision on voluntary end-of-life counseling amounted to death panels, a characterization the bill’s supporters called a deliberate distortion; the provision was eventually dropped, but the episode showed the intensity of the opposition the leadership would have to overcome. On September 9, Obama addressed a joint session of Congress to press the case for action, promising that the legislation would give security to those who had insurance and coverage to those who did not, pledging not to sign a plan that added to the deficit, and endorsing a public insurance option while leaving the details to Congress. Seven weeks later, the Democratic leadership consolidated the committees’ work into a single vehicle and introduced it: H.R. 3962, the Affordable Health Care for America Act, on October 29, 2009. Speaker Nancy Pelosi presented the measure as the House answer to a system in which tens of millions of Americans lacked insurance and millions more feared losing the coverage they had.

The bill’s scope was vast. It barred insurers from denying coverage or charging higher premiums because of pre-existing conditions, ended lifetime caps on benefits, and created new insurance marketplaces, called exchanges, where individuals and small businesses could shop for coverage. The exchanges would operate under a new Health Choices Commissioner, a federal official charged with setting standards for the plans sold within them. Households of modest means would receive federal affordability credits, scaled to income and phased out as income rose, to help pay premiums. Medicaid, the joint federal-state program for the poor, would expand to cover more low-income adults. Most individuals would be required to carry insurance, and larger employers would face financial penalties if they failed to offer it. The measure also created a public option, a government-administered insurance plan that would compete with private carriers inside the exchanges, began closing the gap in Medicare prescription-drug coverage known as the donut hole, and established a voluntary long-term-care insurance program. The bill also required insurers to spend a minimum share of premium dollars on medical care and created a temporary high-risk pool for people locked out of the market by their medical histories. It was the most ambitious health care legislation the House had considered in a generation, and its fate depended on whether the Democratic majority could hold together long enough to pass it.

The bill that reached the floor had already survived one near-death experience. In the Energy and Commerce Committee, fiscally conservative Democrats known as Blue Dogs had stalled the markup in July, demanding changes to the bill’s cost and its treatment of rural providers, and Chairman Henry Waxman had negotiated a deal to win their votes. The episode previewed the autumn’s dynamic: the leadership’s majority was large, but its ideological spread meant every stage of the process required bargaining with its own members.

Paying for the expansion was its own fight. The bill combined projected savings in Medicare, including reductions in payments to private Medicare Advantage plans, with new levies on high-income households and fees on insurers, pharmaceutical companies, and medical-device makers. The Congressional Budget Office’s cost estimate gave the leadership the fiscal foundation it needed, and attention turned to the floor schedule. In the House, the path to the floor runs through the Rules Committee, often called the Speaker’s committee because its majority reliably reflects the leadership’s wishes, and the terms of debate would decide which fights the House actually had. The Rules Committee reported a special rule, H.Res. 903, and the House adopted it before general debate began. The rule waived points of order against the bill, set the terms of general debate, and closed the amendment process: rank-and-file members could not rise on the floor and offer changes of their own. Instead, the rule listed the only amendments the House would consider, a manager’s amendment carrying the leadership’s final technical and substantive adjustments and one additional amendment that would decide the bill’s fate. The Rules Committee had made in order an amendment on abortion offered by Representative Bart Stupak, Democrat of Michigan, and Representative Joe Pitts, Republican of Pennsylvania. That single decision set the terms for everything that followed, because a bloc of anti-abortion Democrats had made clear that the bill would not reach final passage without a vote on their language. The committee’s hearing drew testimony from the bill’s managers and its critics, and the panel’s Democratic majority reported the rule along the lines the leadership wanted. Allowing the Stupak vote was the price of that cooperation: without it, the rule itself might have failed on the floor, since the Stupak bloc had threatened to join Republicans in voting it down.

The Stupak bloc had been assembling for months. Stupak, a longtime opponent of abortion rights, argued that the health care bill as drafted would let federal money flow to insurance plans covering elective abortions. The Hyde amendment, a rider named for Representative Henry Hyde of Illinois and carried on annual appropriations bills since the 1970s, barred the use of federal funds for abortions except in cases of rape, incest, or danger to the life of the mother. Stupak and his allies contended that the affordability credits in H.R. 3962 created a new channel of federal spending outside the appropriations process, and that without explicit statutory language the Hyde restriction would not follow the money into the exchanges. The Energy and Commerce Committee had adopted a compromise associated with Representative Lois Capps of California, which permitted abortion coverage in subsidized plans so long as the funds were segregated, and the bill’s managers described that arrangement as consistent with existing law. Stupak called it a bookkeeping fiction and demanded a stricter standard. His group, several dozen strong, warned that it would vote against the rule governing debate, and if necessary against the bill itself, unless the House adopted his language. The threat was aimed at the rule vote deliberately: defeating the rule would have kept the bill off the floor altogether. Pelosi could afford few defections. With nearly all of the 177 Republicans expected to oppose the measure, she needed 218 votes from a caucus of 258, and the Stupak bloc was large enough to deny her that number.

Negotiations consumed the first week of November. The White House made clear that it opposed the Stupak language, and organizations supporting abortion rights lobbied the leadership to keep the amendment off the floor. But the arithmetic was the arithmetic, and on the eve of the floor debate the Rules Committee agreed to make the Stupak-Pitts amendment in order. The House took it up on the evening of November 7 as the central fight of the debate on H.R. 3962. Stupak and Pitts were joined as co-sponsors by Representatives Brad Ellsworth of Indiana, Dan Lipinski of Illinois, and Kathy Dahlkemper of Pennsylvania, Democrats who shared their opposition to abortion rights. The amendment’s text barred the public option from covering elective abortions and barred recipients of the bill’s affordability credits from using those credits to purchase any plan that covered elective abortions, while expressly allowing individuals to buy separate supplemental coverage for abortion services with their own private funds. Supporters described the provision as an extension of the Hyde principle to the new subsidies; opponents, including many Democrats who supported abortion rights, described it as a restriction that reached past federal dollars and into the private insurance market. The debate scrambled the usual alignments, with members of both parties crossing lines, and the roll call showed it: the amendment was adopted 240 to 194, with 64 Democrats voting in favor alongside nearly all Republicans. The sixty-four Democrats who joined the majority included many members who would go on to vote for final passage; the amendment gave them the political cover they said they needed. For the leadership, the vote was transactional: the amendment’s adoption purchased the bloc’s support for the bill itself.

With the amendment adopted, the Stupak bloc’s objections fell away, and the House moved toward final passage late that Saturday night. Republican speakers denounced the measure’s cost, its reductions in projected Medicare spending, and the scale of its new federal machinery, while Democratic speakers framed it as the fulfillment of a decades-long effort to guarantee coverage. The House also rejected a Republican substitute that would have replaced the bill with a narrower set of insurance reforms. The minority’s motion to recommit, the last procedural move available to the opposition, was defeated, and the House proceeded to the final roll call. It closed at 11:19 p.m. EST, and the tally stood at 220 to 215. Two hundred nineteen Democrats voted yes, joined by a single Republican, Representative Anh Cao of Louisiana, whose New Orleans district had voted heavily Democratic and who had announced his support that afternoon. Thirty-nine Democrats voted no, many of them fiscal conservatives who had criticized the bill’s cost throughout the autumn, joining all but one of their Republican colleagues in opposition. The margin was five votes, the narrowest the leadership could afford, and it illustrated the arithmetic that had governed the entire autumn: with the minority unified in opposition, every Democratic vote had to be earned individually. Had the Stupak amendment failed, a large share of its Democratic supporters had pledged to vote against the bill, and the 220-vote majority would have collapsed. The abortion vote had been the price of passage, and the leadership had paid it. For Pelosi, who had spent the autumn counting votes member by member, the five-vote margin was both a triumph and a warning: the coalition that had passed the bill was held together by bargains, like the Stupak amendment, that the Senate had not made.

The price, however, created a problem for the next stage. The Senate was writing its own measure, and its abortion language differed from what the House had just adopted. Where the House had accepted the Stupak-Pitts standard, the Senate bill carried language associated with Senator Ben Nelson of Nebraska, a less restrictive formulation that permitted subsidized plans to cover elective abortions under conditions the House language would not have allowed, including a mechanism for segregating federal and private funds. The two texts would have to be reconciled, and the plan at the time called for a conference committee to do the reconciling. That prospect gave the abortion question a second life. A conference committee draws its members from both chambers, named by the Speaker and the Senate leadership, and its report reconciling the House and Senate versions of a bill cannot be amended on either floor; each chamber must accept or reject the conferees’ product as a whole, and in the Senate a filibuster of the report would again require sixty votes to break. Pro-life House Democrats therefore had reason to insist that the conference preserve the Stupak language, and they said so publicly, warning that they would oppose any final agreement that weakened it. Stupak himself warned that he would urge his bloc to oppose the conference report if it dropped his language, a threat that gave the House negotiators a clear instruction and the Senate negotiators a clear warning. Members who favored the Senate’s looser standard had the opposite incentive, and the White House, which had opposed the Stupak amendment on the House floor, preferred maximum flexibility for the negotiators. The administration signaled that it wanted the conference to have room to maneuver, setting up a three-way negotiation among the House, the Senate, and the White House. Abortion was not the only gap awaiting the conference: the House bill’s public option had no counterpart in the Senate measure, and the two chambers differed on the structure of subsidies and the scope of the Medicaid expansion. But the abortion gap was the most politically charged of the differences, and both sides understood that whichever standard the conferees chose would effectively decide the abortion policy of the final law. The House had acted first, and it had acted by accepting a bargain its own negotiators would now have to defend.

Christmas Eve in the Senate

While the House had settled its abortion question by adopting the Stupak language, the Senate had settled its own version differently, and the difference would shadow the upper chamber’s work from the start. The Senate bill carried the abortion provisions associated with Senator Ben Nelson of Nebraska, a less restrictive formulation than the Stupak-Pitts amendment, and everyone involved understood that a conference committee would eventually have to choose between them or blend them. But before any conference could convene, the Senate had to pass something, and in the Senate of 2009 passing anything of consequence required sixty votes. The Senate’s bill had taken shape under different pressures than the House’s: a Finance Committee chairman determined to win Republican votes, a HELP Committee inclined toward a more expansive bill, and a majority leader who needed every member of a diverse caucus. The sixty-seat caucus was itself a recent achievement, secured only when Al Franken’s seating in July gave the Democrats their sixtieth vote. The Senate bill also omitted the public option the House had passed, leaving that gap for the conference as well.

Why did abortion language decide the House vote on November 7?

Anti-abortion Democrats, several dozen strong, conditioned their support on barring federal subsidies from plans covering elective abortions, and Speaker Pelosi needed 218 votes from 258 Democrats. The Stupak-Pitts amendment, adopted 240 to 194 on November 7, purchased the bloc’s votes and supplied the margin that carried the bill 220 to 215 that night.

The Senate’s answer to the same question took a different form, and the difference would become one of the conference’s central disputes.

The Senate’s path to sixty began in committee and ran through a summer of negotiation. The Finance Committee, under Chairman Max Baucus of Montana, spent months in talks, including a much-publicized effort by a bipartisan group of six senators, three from each party, to find common ground. The talks collapsed in September, and the committee marked up Baucus’s bill, which contained no public option, approving it in the autumn with a single Republican vote, Senator Olympia Snowe of Maine’s. The Health, Education, Labor, and Pensions Committee had approved its own, more expansive version in July under Senator Christopher Dodd of Connecticut, who had taken the gavel after Senator Edward Kennedy’s death in August. The HELP bill included a public option and passed on a party-line vote. The two committee bills reflected the Senate’s own divisions: the HELP bill’s public option and more generous subsidies against the Finance bill’s leaner structure and its reliance on nonprofit insurance cooperatives. Majority Leader Harry Reid then merged the two committee products into a single substitute, borrowing the Finance Committee’s structure while adding elements from the HELP bill. Reid built his floor vehicle on the Senate committees’ work rather than taking up the House-passed bill directly, and after the Congressional Budget Office scored the merged bill, he moved to bring it to the floor. On November 21, 2009, the Senate took up the motion to proceed, the parliamentary step that opens floor debate. The Senate convened on a Saturday, a rarity that underscored the leadership’s determination to keep the process moving. The motion was itself subject to a filibuster, so Reid filed for cloture, and the Senate voted 60 to 39 to invoke it, with every Democrat and both independents, Joseph Lieberman of Connecticut and Bernard Sanders of Vermont, voting yes. The sixty votes included two independents Reid had cultivated for months: Sanders, who wanted the bill to go further, and Lieberman, who wanted it to go less far. Holding both, along with fifty-eight Democrats spanning the party’s ideological range, would be the work of the next five weeks. Minority Leader Mitch McConnell argued the bill was being rushed to the floor before the country understood it, but the parliamentary question was only whether debate could begin. The vote decided that it could, not how it would end. Senators Blanche Lincoln of Arkansas and Mary Landrieu of Louisiana voted to proceed while publicly reserving judgment on the final product, a signal that their support had been rented for the procedural step rather than committed to the substance. Landrieu had secured additional Medicaid funds for Louisiana as the price of her procedural support, the provision critics would later call the Louisiana Purchase. Weeks of floor debate followed, with the Senate working through amendments in the slow, sequential fashion its rules impose.

Why did the Senate need exactly sixty votes?

Senate rules allow extended debate, and a minority can block final action unless three fifths of the chamber, sixty senators, vote to invoke cloture and close debate. With all forty Republicans united in opposition to the bill, Majority Leader Reid needed all sixty members of his Democratic caucus, including both independents, to reach the threshold at each procedural step.

The threshold applied at every stage: the motion to proceed, any amendment that drew a filibuster, the cloture vote on the bill itself, and eventually any conference report. The chamber’s filibuster and cloture rules explain why that number governed every procedural step.

The arithmetic gave every member of the Democratic caucus an effective veto, and the weeks of December became an exercise in holding sixty senators together through weekend sessions and late-night votes. Reid negotiated with individual members, adjusting the bill to secure each vote, and the concessions he made became some of the most discussed provisions of the Senate measure. Senator Joseph Lieberman’s refusal to support any bill containing a government-run insurance option forced the leadership to drop the public option from the Senate text, even though the House had passed one. Lieberman, who caucused with the Democrats but had won re-election as an independent in 2006, said in mid-December that he would support a filibuster of any bill containing the provision, and a late proposal to let adults aged fifty-five to sixty-four buy into Medicare met the same fate after he announced he would join a filibuster against it. A managers’ amendment consolidating the negotiated changes was assembled as the vehicle for the final text; released on December 19, it ran to hundreds of pages and carried the Nebraska and Louisiana Medicaid provisions, the Nelson abortion compromise, and dozens of smaller provisions sought by individual senators, the customary price of assembling sixty votes. The minority used the Senate’s procedural tools to slow the pace, forcing roll-call votes and consuming debate time, and Republicans offered amendments aimed at the individual mandate, the bill’s tax provisions, and its Medicare savings, forcing Democrats to cast difficult votes even as the outcomes were never in doubt.

The Senate bill shared the House measure’s architecture, exchanges, subsidies, an individual mandate, insurance market reforms, and a Medicaid expansion, but it differed in the details: no public option, a different subsidy formula, an excise tax on high-cost insurance plans, and the Nelson abortion language. To pay for the coverage expansion, the Senate bill relied on a mix of Medicare savings, fees on insurers and medical-device makers, and the excise tax on high-cost employer plans, which labor unions opposed and the House had not included. The bill required most Americans to carry insurance or pay a penalty, a provision Obama had opposed during the presidential campaign but embraced as necessary to keep premiums stable once insurers could no longer turn away the sick.

Nelson negotiated directly with Reid and the White House through the second week of December, and his price became clear in stages: first the Medicaid provision for Nebraska, then the abortion language. To win the support of Senator Ben Nelson, whose vote was indispensable, the bill included a provision under which the federal government would permanently cover the full cost of Nebraska’s Medicaid expansion. Nelson had made his position clear early: without the Medicaid provision and acceptable abortion language, he would not supply the sixtieth vote. Under the bill’s general formula, the federal government would pay the full cost of the expansion in its early years, with the federal share stepping down to ninety percent thereafter; the Nebraska provision froze the federal share at one hundred percent permanently, exempting the state from the step-down every other state would face. Republican critics dubbed the arrangement the Cornhusker Kickback, a characterization its defenders rejected as a routine accommodation of the kind the Senate regularly makes to assemble majorities. The provision became a symbol for opponents of the bill, who cited it as evidence that the sixty votes had been purchased rather than persuaded. A separate provision directed additional federal Medicaid funds to Louisiana, addressing the state’s hurricane-recovery costs after post-Katrina income data had distorted its federal matching rate; critics labeled it the Louisiana Purchase, drawing the obvious historical parallel. The Nebraska provision was later repealed in the reconciliation bill, while the Louisiana disaster-recovery adjustment survived in modified form, but in December both served their purpose: they kept wavering senators inside the sixty.

Nelson’s abortion language was the other price of his vote. The Nebraska senator, an opponent of abortion rights, would not support a bill whose subsidies could flow to plans covering elective abortions, but he also declined to embrace the full Stupak-Pitts standard the House had adopted. On December 8, the Senate had tabled a stricter abortion amendment Nelson himself had offered, by a vote of 54 to 45, clearing the way for the compromise. The compromise associated with his name required that federal subsidy dollars be segregated from any funds used for elective abortion coverage, with subscribers paying for such coverage through separate payments drawn from private funds, and it permitted states to bar plans covering elective abortions from their insurance exchanges altogether. In practice, the language meant an enrollee seeking abortion coverage would pay for it separately from the plan itself, with the insurer keeping the abortion premium dollars and the federal subsidy dollars in separate accounts. The language was less restrictive than the House version: where Stupak-Pitts had barred subsidy recipients from buying plans that covered elective abortions at all, the Nelson formulation allowed the purchase so long as the abortion coverage was paid for separately. The gap between the two chambers’ texts thus survived the Senate’s deliberations intact, and it joined the list of differences awaiting the expected conference.

The endgame arrived in the week before Christmas. Reid filed cloture on the bill itself, and in the early hours of December 23 the Senate voted 60 to 39 to invoke it, Record Vote 395, with the Democratic caucus again holding together and every Republican present voting no. By then the outcome had been certain for days; the remaining drama was procedural, measured in hours of required debate rather than in persuadable senators. Republicans, resigned to the outcome, used their remaining debate time to denounce the process that had produced it. Thirty hours of post-cloture debate followed, the maximum the rules allow, and then, at about 7:16 a.m. on the morning of December 24, the Senate passed the bill 60 to 39, Record Vote 396. Every Democrat and both independents voted yes; every Republican present voted no; Senator Jim Bunning of Kentucky, a Republican, was absent and did not vote, the only senator not to cast a ballot. Vice President Joe Biden presided over the chamber for the passage vote, a ceremonial acknowledgment of the moment’s weight. Reid himself briefly voted no by mistake during the roll call before correcting his vote to yes. The moment drew a ripple of confusion in the chamber; under Senate practice, a senator may change a vote before the result is announced. The chamber was nearly empty when the clerk called the roll; most senators voted from their desks in a matter of minutes, and the result, long expected, was greeted without celebration. No Republican voted for the bill at any stage. The scene was subdued for legislation of such scale: a morning vote on Christmas Eve, thin galleries, and a result that had been certain since the cloture tally the day before.

At that moment, the expectation was that the two bills would go to conference. House and Senate leaders spoke of appointing conferees in the new year, and the ordinary machinery of bicameral legislating was set to engage: the conferees would reconcile the House and Senate texts, produce a conference report, and send it back to each chamber for an up-or-down vote without amendment. The plan envisioned the House naming its conferees and the Senate naming its own in January, with the two delegations bargaining over the abortion language, the public option, the subsidy formulas, and the tax provisions. The abortion language, the Medicaid provisions, the scope of the subsidies, the structure of the exchanges, and the fate of the public option would all be settled in that negotiation. The vote positioned the Senate bill as the vehicle for the final negotiations. The Senate had done its part on Christmas Eve morning, and senators scattered for the holiday with the understanding that the hardest bargaining still lay ahead: reconciling two bills that differed on abortion, the public option, taxes, and subsidies. The final shape of the law now depended on what the conference would make of the differences between the two chambers’ work.

The Sixtieth Vote Vanishes

The late Edward Kennedy had held his Massachusetts Senate seat for decades, and when he died in 2009 the vacancy fell to a special election scheduled for January 19, 2010. Democrats entered that contest expecting a routine result. Massachusetts had not elected a Republican to the Senate in a generation, and the national party treated the race as a formality on the way to replacing Kennedy with a Democrat who would keep the caucus at sixty votes. The Republican nominee, state senator Scott Brown, campaigned as a break from the pattern, and the Democratic nominee, state attorney general Martha Coakley, ran as the heir to the seat. On election night the assumed outcome collapsed. Brown defeated Coakley by a margin of about 52 to 47 percent, a result that stunned the Democratic leadership in Washington and immediately redrew the legislative arithmetic of the health care fight.

Brown was sworn in on February 4, 2010, becoming the forty-first Republican in the Senate. The number mattered more than the name. With his arrival the Democratic caucus fell to fifty-nine, counting the two independents who aligned with it, and the sixty-vote supermajority that had let Democrats break a filibuster on their own was gone. In the Senate, sixty is the number that ends debate. A motion to invoke cloture requires sixty votes, and without it a determined minority can keep a measure from ever reaching a final vote. The loss of the sixtieth vote therefore did not merely narrow the majority. It removed the mechanism by which the majority had planned to force the health bill through the upper chamber against unified Republican opposition.

To grasp what vanished, it helps to recall the plan as it stood before January 19. The Senate had passed its version of the health legislation in December, and the House had passed its own version the previous November. The two chambers had written substantially different bills. The ordinary way to resolve such differences is a conference committee, where appointed members of both houses negotiate a single compromise text, the conference report, which each chamber then votes on without amendment. The Senate would have needed sixty votes to invoke cloture on that report and bring it to a final vote. With sixty Democrats and aligned independents in the caucus, the leadership believed it could supply those votes. A conference report, once adopted by both chambers, would have gone to the President for signature. That was the path, and it depended entirely on the sixtieth vote.

Brown’s victory destroyed it at a stroke. A conference report is debatable in the Senate like any other major measure, and with only fifty-nine votes available to the Democratic side, opponents could filibuster the report and defeat cloture. No amount of negotiation inside a conference committee could fix a shortage that existed on the Senate floor. The report would emerge, stall, and die. Congressional leaders understood this immediately, and the planned conference was abandoned before it ever convened. The House would never send its members to a negotiation whose product could not survive the Senate.

What followed was a period of open uncertainty that stretched through late January and February. The bill’s supporters in Congress split over what to do next. Some argued that the House should simply pass the Senate bill as written, accepting provisions that House Democrats disliked in order to get a law enacted. Many House members resisted that idea, saying the Senate text differed too sharply from what they had voted for and from what they had promised their constituents. Others proposed moving the entire package through the budget reconciliation process, which requires only a simple majority in the Senate, but that route raised its own problems, since reconciliation carries strict limits on what provisions it can carry and House leaders had no appetite for rebuilding the whole measure under those constraints. Still others, watching the political cost of the fight, urged the party to set the bill aside and turn to other business. None of these options commanded a majority, and for weeks the legislation appeared to be drifting toward failure.

The arithmetic, however, pointed toward a narrow opening. The Senate had already passed its bill with sixty votes in December, and that vote could not be undone. If the House agreed to the Senate text exactly as the Senate had passed it, without changing a single word, then the Senate would not need to act again. A bill that both chambers have passed in identical form goes directly to the President. The sixty-vote threshold would never come into play, because there would be nothing left for the Senate to vote on. The obstacle was political rather than procedural. House Democrats would have to swallow the Senate bill whole, including the provisions they had spent months criticizing, and trust that a second measure would fix what they could not change in the first.

That second measure was the reconciliation bill, and the pairing of the two became the strategy that carried the legislation to enactment. The Congressional Research Service later described the approach as a sidecar arrangement, with the reconciliation measure riding alongside the base bill to make the adjustments both chambers had agreed upon. The design solved the central problem. The base bill needed no further Senate action, so the missing sixtieth vote was irrelevant to it. The reconciliation bill needed only a simple majority in the Senate, so the missing sixtieth vote was irrelevant to it as well. What had looked like a fatal shortage became, through this two-step construction, a problem the leadership could walk around rather than through.

The weeks of drift ended when the House leadership committed to the maneuver and began counting votes for the pieces in sequence. The decision required House Democrats to accept the Senate text first and trust the reconciliation fix to follow, a sequence that demanded unusual discipline from a caucus that had just watched its Senate margin collapse. By mid-March the leadership believed the votes were there, and the House moved.

The trust problem was the heart of the matter. Members of the House who disliked the Senate bill had to vote for it before the fixes existed in law. The reconciliation bill would make the agreed changes, but it would be passed after the base bill, and nothing in the rules could guarantee its fate in the Senate beyond the simple-majority threshold. If the reconciliation measure somehow failed, the Senate bill would stand alone as the law of the land, with all of its provisions that the House had opposed. Asking a majority of the House to take that risk, weeks after an election that many read as a rebuke, required a level of coordination that the chamber had rarely attempted.

Some House Democrats tried to reverse the sequence. They asked the Senate to pass the reconciliation fixes first, so the House would never have to vote on the unamended Senate bill alone. The Senate leadership rejected the idea as unworkable under the chamber’s rules. A reconciliation bill amends existing law or pending law through a defined budget process, and the leadership concluded it could not sensibly precede the base bill it was meant to repair. The House would have to go first. That answer deepened the trust problem, because it confirmed that the House vote on the Senate bill would be a leap taken on the strength of promises from across the Capitol rather than on the security of enacted text.

The Rules Committee briefly considered another device that might have eased the politics of the first vote. Under a self-executing rule, sometimes called deem and pass, the House could have been deemed to concur in the Senate amendments at the moment it adopted the rule governing debate, sparing members a separate recorded vote on the Senate bill itself. The idea drew immediate criticism from Republican members, who characterized it as an attempt to pass the bill without voting on it, and it made many Democrats uneasy as well. The leadership dropped the device before March 21 and put the House on record with a direct vote. The decision traded political cover for procedural clarity, and it meant the concurrence vote would stand as a clean, countable act.

Leadership addressed the risk by binding the two bills together politically rather than procedurally. The Speaker and the Majority Leader presented the votes as a single package in everything but name, and the President engaged directly with wavering members to hold the coalition. The message to the caucus was that the Senate bill was a vehicle and the reconciliation bill was the destination, and that neither made sense without the other. Members who objected to particular Senate provisions were told those provisions would be altered by the reconciliation measure, and members who doubted the reconciliation measure would survive were told the Senate leadership had committed to passing it. The strategy rested on commitments between the chambers that the rules could not enforce but that the party’s leaders treated as binding.

Meanwhile the calendar pressed on the process. The longer the House waited, the more the political environment deteriorated, and opponents of the bill used the interval to argue that the two-bill maneuver was an abuse of procedure. Republican members, led by Senate Minority Leader Mitch McConnell, characterized the plan as an effort to ignore the message of the Massachusetts election and force through a bill the public had rejected. House Speaker Nancy Pelosi answered that the Senate bill had earned its majority through the regular process in December and that every step of the two-bill sequence used established parliamentary tools. The dispute over legitimacy ran alongside the dispute over policy, and both shaped the final weeks of the fight.

The mechanics of the Senate’s rules gave the dispute its edge. Even motions to take up a conference report are debatable, and after cloture is invoked the rules still allow up to thirty hours of post-cloture consideration. With fifty-nine votes, the Democratic caucus could not even reach cloture, which meant opponents could extend debate indefinitely at multiple stages of a conference report’s journey. The leadership’s count was not a matter of persuasion. It was a matter of arithmetic, and the arithmetic said that no version of the conference path could survive the floor.

The sixtieth vote that the December tally had relied upon belonged, in a literal sense, to a placeholder. After Kennedy’s death, Massachusetts law provided for an interim appointee to hold the seat until the special election, and Governor Deval Patrick named Paul Kirk, a former Kennedy aide, to fill it. Kirk was in place when the upper chamber passed its health legislation in December, and his vote was one of the sixty that carried it. The Democratic supermajority had therefore always been a temporary construction, assembled around an appointed senator whose tenure was scheduled to end the moment the voters chose a successor. Brown’s victory did not take the sixtieth vote by surprise so much as it replaced the placeholder with an opponent, converting a borrowed margin into a permanent deficit.

The symbolism compounded the arithmetic. The seat was known nationally as Kennedy’s seat, and Kennedy had made health care reform the cause of his final years in public life. Losing that particular seat to a Republican who had campaigned explicitly against the health bill allowed opponents to frame the result as a verdict on the legislation itself. Republican members argued that the voters of the most Democratic state in the country had rejected the bill. Democratic leaders argued that the race had turned on the flaws of the Coakley campaign and on anger about the economy, not on the merits of health reform. Both readings were political characterizations of the same 52 to 47 result, and both shaped the weeks that followed, because the meaning assigned to the election determined how much risk the majority was willing to take to finish the job.

By the third week of March the uncertainty had narrowed to a single question, which was whether the House could assemble majorities for each piece in the required order. The leadership had solved the procedural puzzle. What remained was the human puzzle of persuading more than two hundred members to walk through it together. The answer came on March 21, when the House took up the bills in sequence and the maneuver moved from theory to law.

The Two-Bill Maneuver

The maneuver that carried the Affordable Care Act to enactment had two moving parts, and each part solved a different half of the problem that Scott Brown’s election had created. The first part was the House’s adoption of the Senate bill, H.R. 3590, exactly as the Senate had passed it in December. The second part was a separate reconciliation bill, H.R. 4872, that made the changes both chambers had negotiated. The order mattered as much as the content. The House would pass the Senate bill first, clearing it for the President without any further Senate action, and then both chambers would pass the reconciliation bill under rules that required only a simple majority in the Senate.

The logic of the first step rested on a basic constitutional rule. A bill becomes law when both chambers pass the identical text and the President signs it. The Senate had already passed H.R. 3590. If the House concurred in the Senate’s amendments without altering a word, the text in both chambers would be identical, and no further Senate vote would be necessary. The bill could go straight to the President’s desk. This was the feature that made the maneuver workable after the loss of the sixtieth vote. There was no conference report to filibuster, no amended bill to send back across the Capitol, and no second Senate vote on the base text at which the missing sixtieth vote could matter. The sixty-vote threshold simply never arose, because the Senate had nothing left to decide on that bill.

The logic of the second step rested on the special rules of budget reconciliation. Reconciliation is a parliamentary process created by the Congressional Budget Act that allows certain budget-related measures to move through the Senate under expedited procedures. Its defining feature is the vote threshold. A reconciliation bill needs only a simple majority in the Senate, and it cannot be filibustered, which means the sixty-vote cloture requirement does not apply. With fifty-nine votes in hand, the Democratic caucus could pass a reconciliation bill on its own. The second bill would carry the adjustments the House and Senate leadership had agreed upon, altering the provisions of the Senate bill that the House found objectionable and adding the policies that the two chambers had negotiated after December.

What parliamentary path survived the loss of the sixtieth vote?

Before the election, the plan was to merge the House and Senate bills in conference and pass the report with sixty Senate votes. After Brown became the forty first Republican senator, Democrats held only fifty nine votes, so any conference report could be filibustered. Leadership abandoned conference and designed a strategy that required no Senate vote on the base bill.

It is important to understand what this maneuver was not. It was not the passage of the entire health care act through reconciliation. The base bill, the measure that created the act’s coverage expansions, insurance market rules, and major programs, passed as ordinary legislation. The House concurred in the Senate amendments by a regular majority vote, and the President signed the resulting law through the normal presentment process. Reconciliation touched only the second bill, the package of amendments and fixes. Observers sometimes described the whole episode as a reconciliation end-run, but that description misstates the mechanics. The act itself traveled the standard legislative route. Only the sidecar traveled under reconciliation’s expedited rules.

The distinction mattered for the debate over legitimacy. Opponents argued that using reconciliation at all was an abuse of the process, and Republican members characterized the two-bill sequence as a way to dodge the consequences of the Massachusetts result. Supporters answered that the base bill had won its sixty votes fairly in December, when the Senate passed it before Brown took his seat, and that reconciliation was being used for the limited purpose the budget rules allow, which is adjusting budget-related provisions by majority vote. The Congressional Research Service, in its neutral institutional voice, described the arrangement as a sidecar strategy, a term that captured the design without endorsing either side’s political characterization.

The House executed the sequence on March 21, 2010, in a single day of extraordinary parliamentary activity, and the choice of day was itself a signal of urgency. March 21 was a Sunday, and the House rarely conducts major legislative business on a Sunday. The leadership had scheduled the session for the weekend to keep the momentum of the final vote count from dissipating and to complete the maneuver before opposition could regroup. Members debated through the afternoon and into the evening, with the minority using its time to restate the case against the legislation and the majority using its time to defend the two-step construction.

The day began with the rule. The House adopted H.Res. 1203, the special rule governing debate on the health measures, by a tally of 224 to 206. A special rule sets the terms under which the House considers a measure, including how long debate runs and what amendments are permitted, and its adoption is the procedural gateway to the substantive votes that follow. The rule structured the consideration of both the base legislation and the reconciliation package, and its passage confirmed that the leadership had the organization, if not yet the final tallies, to proceed. With the rule in place, the House moved to the central vote of the day. It concurred in the Senate amendments to H.R. 3590 by 219 to 212, recorded as Roll no. 165. That vote was the moment the base legislation cleared Congress. Because the House had accepted the Senate text unchanged, the upper chamber was not required to act again, and the measure was cleared for the President.

The discipline of the unchanged text deserves a final emphasis, because it was the load-bearing element of the entire design. Had the House altered so much as a comma in the Senate’s amendments, the altered text would have had to return to the Senate for another round of consideration, and that round would have required sixty votes to survive a filibuster. The leadership understood that any amendment, however minor or popular, would have reintroduced the very threshold the maneuver was built to avoid. The House therefore voted on the Senate’s words exactly as the Senate had written them, accepting provisions its members disliked and postponing every correction to the reconciliation vehicle. The purity of the concurrence was not an accident of drafting. It was the condition that made the rest possible.

Why could no conference report survive the Senate after January 19?

A conference committee could still meet and agree on a report, but the report would have to pass the Senate again. With only fifty-nine Democratic votes, opponents could block cloture on the report, killing it. Meeting in conference therefore offered no path to enactment, so leadership dropped the idea entirely.

The House then turned to the second bill. It passed H.R. 4872, the Health Care and Education Reconciliation Act, by 220 to 211. This was the sidecar, the package of changes that the leadership had negotiated to make the Senate bill acceptable to the House majority. Its passage in the House completed the chamber’s half of the maneuver. The reconciliation bill then went to the Senate, where it could pass by simple majority, protected from filibuster by the reconciliation rules. The two votes together, 219 to 212 on the base bill and 220 to 211 on the reconciliation measure, represented the majorities that the maneuver had been designed to assemble, each just above the threshold its parliamentary track required.

The Senate’s handling of the reconciliation bill illustrated why the process fit the second measure so neatly. Reconciliation begins with instructions in a budget resolution, and the resulting bill moves under expedited procedures that limit debate and bar the filibuster. Amendments are still permitted, but they face strict germaneness and budgetary tests, and the Byrd rule allows senators to strike provisions deemed extraneous to the budget. Those limits are precisely why the leadership never attempted to run the entire health care act through reconciliation. The act’s insurance market rules, coverage mandates, and delivery system reforms were regulatory provisions, not budget provisions, and much of that architecture would have been vulnerable to Byrd rule challenges. The base bill therefore traveled as ordinary legislation, where no such limits applied, and reconciliation carried only the budgetary adjustments that the rules permit.

Enrollment and presentment followed the ordinary course. Once the House concurred in the Senate amendments, the bill was enrolled, the formal process by which Congress certifies the identical text passed by both chambers, and it was presented to the President. The constitutional requirement was satisfied. Both houses had passed the same words, and the President’s signature would make them law. None of this required the Senate to lift a finger after December, which was the entire point of the construction.

Two days later the process reached its formal conclusion. On March 23, 2010, President Obama signed H.R. 3590 at a White House ceremony, enacting it as Public Law 111-148, the Patient Protection and Affordable Care Act. The signing closed the sequence that had begun with the collapse of the conference plan in January. The base law was in place. The reconciliation measure would follow through the Senate under its majority-vote rules, completing the adjustments the two chambers had agreed upon. The signing itself, on March 23, transformed the parliamentary sequence into a public fact. At the White House ceremony, the President presented the law as the completion of a legislative effort that had survived the loss of the Senate supermajority through the two-step construction the leadership had designed in February. The enrollment of the text, the presentment to the President, and the affixing of the signature followed the constitutional forms exactly, and the designation as Public Law 111-148 entered the measure into the statute books under the same procedures that govern any enacted legislation. The extraordinary part of the story was never the signing. It was everything the House had done on March 21 to make the signing possible without a sixtieth senator.

The two-bill construction had done what the conference could not. It had moved a major health care bill from a deadlocked Senate to the President’s desk without ever needing the sixtieth vote that Scott Brown’s election had taken away.

The narrowness of the House margins deserves emphasis, because it explains why the maneuver is remembered as much for parliamentary design as for vote counting. The concurrence tally stood at 219 to 212, a margin of seven. The reconciliation package passed 220 to 211, a margin of nine. In a chamber of 435 members, those are fragile majorities, and they held only because the leadership had sequenced the decisions to give each faction something it needed. Members who disliked the Senate text got the reconciliation fix. Members who doubted the fix got the signed base law first. The parliamentary structure and the political coalition were built together, each reinforcing the other.

The leadership’s whip operation in the final days illustrated how thin the margin was. Individual members were courted with policy concessions, district-level commitments, and direct appeals from the President, and the public positions of undecided members shifted by the hour in the week before March 21. The Congressional Research Service’s sidecar description captured the mechanics, but the mechanics alone do not explain how seven votes were held together under sustained pressure from an opposition that treated the Massachusetts result as a mandate to stop. The explanation lies in the pairing itself. No member was asked to support the Senate text on its own merits. Every member was asked to support a sequence whose second step repaired the first, and that framing converted opponents of the Senate language into supporters of the package.

The episode left a lasting imprint on how Congress handles closely divided chambers. It demonstrated that the loss of a supermajority does not always end a legislative effort, provided the majority can find a parliamentary path that fits the votes it still holds. It also demonstrated the cost of such paths. The House majority had to accept a Senate text it disliked, trust a second bill to repair the damage, and defend the whole arrangement against charges of procedural overreach. The maneuver succeeded on its own terms. Whether it was a model or a warning depended, as it always does, on which side of the aisle was asked.

Reconciliation and the Byrd Rule

Reconciliation is a legislative procedure created by the Congressional Budget Act of 1974. In ordinary Senate practice, most bills can be debated without limit and defeated by filibuster, which takes sixty votes to break. A reconciliation bill moves under different rules: debate in the Senate is capped at twenty hours, no filibuster is permitted, and passage requires only a simple majority. The procedure was designed for budget business, for deficit reduction and the revenue and spending measures that implement a budget resolution. Its expedited character makes it the only reliable path when a majority exists but a supermajority does not, and for that reason it has been used by both parties for major fiscal legislation since the early 1980s. The 1974 Act created reconciliation as part of a new congressional budget process, giving Congress a way to bring spending, revenue, and debt-limit measures into line with the budget resolution it had adopted. In the decades that followed, presidents and congressional majorities of both parties turned to it for tax cuts, deficit reduction packages, and welfare reform, precisely because it offered a route around the Senate’s supermajority requirement. By 2010 it was a familiar, if controversial, instrument, and its rules were well understood by everyone who would have to operate under them.

The health care fight arrived at exactly that situation in early 2010. The House had passed its own health bill, and the Senate had passed a different one on December 24, 2009. Normally the two chambers would reconcile their differences in a conference committee and vote again on the conference report. But the January 2010 special election in Massachusetts reduced the Democratic caucus to fifty-nine Senate seats, one short of the sixty needed to break a filibuster on a new conference report. The leadership’s answer was a two-step plan: the House would pass the Senate bill unchanged, and a second bill would carry the agreed corrections under reconciliation’s majority-vote rules. The Senate bill would become the law, and the reconciliation bill would amend it.

That second bill was H.R. 4872, the Health Care and Education Reconciliation Act, which the House had already passed as the reconciliation vehicle. What it could contain, however, was not a political question but a parliamentary one, governed by the Byrd rule. The Byrd rule takes its name from Senator Robert Byrd of West Virginia, who persuaded the Senate to adopt it in the mid-1980s and to make it permanent in 1990. It provides that any senator may raise a point of order against provisions in a reconciliation bill that are extraneous to the budget. A provision counts as extraneous when it does not produce a change in spending or revenue, when its budgetary effect is merely incidental to a non-budgetary policy change, when it falls outside the jurisdiction of the committee that reported it, or when it increases the deficit beyond the years covered by the budget resolution. Overcoming a Byrd rule point of order requires sixty votes to waive. Because the majority held only fifty-nine, waiver was not available to it, and the rule therefore functioned as a hard gate on the bill’s contents.

In practice, the test that does the most work is the “merely incidental” standard. A provision that changes spending or revenue only as a side effect of a larger policy change, where the budgetary consequence is secondary to the non-budgetary purpose, fails the test and can be struck. The parliamentarian applies this and the other tests against a body of precedent built up over decades of rulings, so senators drafting a reconciliation bill study past rulings the way lawyers study case law. Provisions that fall outside the jurisdiction of the committee that reported the bill are also vulnerable, as are provisions that increase the deficit in years beyond the budget window. The cumulative effect is a discipline that narrows reconciliation bills to their fiscal core, no matter how broad the majority’s ambitions.

The parliamentarian sits at the center of this process. The parliamentarian is the Senate’s nonpartisan procedural referee, and on Byrd rule questions the parliamentarian advises the presiding officer whether a challenged provision is extraneous. Points of order are raised on the floor by senators, but they are raised under the parliamentarian’s advice, and the presiding officer’s rulings follow that advice in the ordinary course. Senators therefore draft reconciliation bills in the shadow of the parliamentarian’s expected rulings, and provisions that cannot survive a point of order are either rewritten to carry genuine budgetary effect or left out.

What made the Byrd rule the reconciliation bill’s real drafter?

The reconciliation bill could carry only provisions with direct budgetary effects, because the Byrd rule allowed any senator to strike extraneous matter by point of order unless sixty senators waived it. Spending, revenue, and deficit changes qualified. Non-budgetary policy, no matter how desirable to the bill’s authors, could not survive the point of order and had to stay out.

That gate shaped the bill’s contents from the first draft. The student loan overhaul could ride along because it moved money: the measure ended new lending under the Federal Family Education Loan program, in which private lenders originated federally guaranteed student loans, and shifted federal student lending to direct lending by the government, with the resulting savings scored against the budget. Agreed policy changes with no budgetary effect, by contrast, could not be added no matter how widely both chambers supported them. And the most controversial of the Senate bill’s Medicaid bargains, the Nebraska provision, was repealed in the reconciliation measure rather than defended, while the Louisiana disaster-recovery adjustment survived in modified form. The reconciliation bill thus carried a mix of health care corrections, education provisions with fiscal consequences, and the removal of the provisions its authors could no longer sustain. The education title was not an accident of packaging. Because the shift to direct lending was scored as producing savings for the Treasury, those savings helped the overall bill meet its deficit targets, which gave the leadership a fiscal reason to attach the student loan overhaul to the health care corrections. Some agreed fixes were made, and others were not, because the Byrd rule drew the line between them. The result was a bill whose contents reflected not only what the leadership wanted but what the parliamentarian would permit.

The Senate took up H.R. 4872 on March 23, and consideration continued through March 24 and March 25 under reconciliation’s compressed timetable. The procedure’s amendment rules shaped the fight: after the twenty hours of debate expired, senators could offer amendment after amendment in the marathon session known as vote-a-rama, and each amendment had to survive the same Byrd rule scrutiny as the underlying text. Vote-a-rama is one of the Senate’s distinctive spectacles, a rapid succession of roll-call votes with little debate between them, and the minority treats it as an opportunity to force the majority to take uncomfortable positions on the record. The minority used the open amendment process to force difficult votes, offering a series of Republican amendments and motions to recommit, each of which was defeated. Nine Republican amendments fell to Byrd rule points of order: when the parliamentarian advised that provisions in the amendments were extraneous, the points of order were sustained, and the motions to waive the rule were rejected, since waiver required sixty votes the minority could not supply. On March 25 the Senate passed the bill by 56 to 43.

Near the end of that day, however, the Byrd rule struck the bill itself. Senator Judd Gregg of New Hampshire raised two points of order, offered on the floor under the parliamentarian’s advice, against two brief provisions in the education portion of the bill dealing with the Pell grant program. The provisions were ruled extraneous, and both were struck from the text. The strikes were minor in policy terms but decisive in procedure: the bill the Senate had passed was no longer the bill the House had passed, and the measure had to go back to the House for further action. A reconciliation bill that has been amended on the floor, even by subtraction, cannot become law until the other chamber agrees to the change.

What turned a Senate amendment back into a House vote?

The parliamentarian advised that two Pell grant provisions in the education title violated the Byrd rule, so Senator Gregg’s points of order struck them on March 25. The struck text changed the bill the Senate passed, and the House could only concur in the Senate amendment as adopted. Concurring required a new House vote, which it took the same day.

The House acted within hours. On March 25 it adopted a special rule, H.Res. 1225, governing consideration of the Senate amendment to H.R. 4872, and then concurred in that amendment by 220 to 211. A special rule is the House’s own procedural instrument: reported by the Rules Committee, it sets the terms of debate, determines which amendments may be offered, and structures the vote. Here it provided the vehicle for the House to accept the Senate’s amended text without further change. With that concurrence, the reconciliation process was complete, the Senate bill having already been signed into law. President Obama signed H.R. 4872 as Public Law 111-152 on March 30, 2010.

The episode is worth studying as procedure rather than as partisan combat. The Byrd rule acted as an editor with a red pen, deciding which provisions belonged in the expedited vehicle and which did not. It struck nine minority amendments, it struck two provisions of the majority’s own bill, and its operation forced an additional House vote on the same day the Senate finished. None of those outcomes required anyone to break a rule; each followed from rules that had governed the Senate for decades. The reconciliation bill that became Public Law 111-152 was, in a real sense, the text the Byrd rule allowed to exist, and the two Pell grant strikes are the clearest proof that the rule constrained the majority as well as the minority.

The Executive Order and the Final Votes

The last obstacle to the Senate bill’s House passage was abortion funding. The House had earlier adopted the Stupak-Pitts amendment, offered by Representative Bart Stupak, Democrat of Michigan, and Representative Joseph Pitts, Republican of Pennsylvania, which would have barred federal funds from paying for abortion or subsidizing health plans that covered it, with exceptions for rape, incest, and danger to the mother’s life. The Senate’s version did not include that language. The Stupak-Pitts amendment had been adopted by the House as part of its own health bill, but the Senate declined to include it in the version it passed in December 2009, leaving the two chambers with incompatible positions on abortion funding. Stupak said that he and several other anti-abortion Democrats in the House would not support the Senate bill unless its language restricting federal funding of abortion was strengthened. Without their votes, the House could not pass the Senate bill unchanged, and the two-step plan would collapse. The bloc numbered roughly a dozen members, enough to decide the outcome in a closely divided House, and Stupak made clear that the issue was not negotiable for them.

On March 21, 2010, the White House announced the agreement that resolved the standoff: President Obama would issue an executive order on abortion funding, and Stupak and the members aligned with him would vote for the Senate bill. The order, Executive Order 13535, bore the title “Ensuring Enforcement and Implementation of Abortion Restrictions in the Patient Protection and Affordable Care Act.” It was announced on March 21 and signed on March 24. Its stated purpose was to reinforce the commitment to the Hyde Amendment’s policy of restricting federal funds for abortion within the context of the new health care legislation. In substance, the order established that the act would be implemented consistent with that policy: it directed the Office of Management and Budget and the Department of Health and Human Services to develop guidance ensuring compliance, and it provided that tax credits and cost-sharing reduction payments in the health insurance exchanges could not be used to pay for abortion services except in cases of rape, incest, or danger to the life of the woman. The order also contemplated the segregation of funds within the exchanges, so that the federal subsidies flowing through the new marketplaces would be kept separate from any payments for abortion coverage. It was, by its terms, an enforcement document rather than a new restriction: it purported to ensure that the Hyde policy would be carried into the new law’s implementation, not to write new abortion law by presidential directive.

Stupak kept his side of the bargain. After the agreement was announced, he voted for the Senate bill, and other members of the anti-abortion bloc followed. Their votes supplied support the bill’s managers needed at the decisive moment, and the House passed the Senate bill that evening of March 21. The executive order had done its political work before the ink on it was dry: announced on the day of the vote, signed three days later, it converted opponents into supporters on the strength of a presidential commitment. For Stupak, the order represented a tangible extension of the Hyde principle into the new law’s machinery, and he defended it as consistent with the sanctity-of-life position he had held throughout the debate.

The choice of an executive order, rather than an amendment to the bill, was itself dictated by procedure. The entire two-step plan depended on the House passing the Senate bill verbatim: any House amendment to the Senate text would have sent the bill back to a Senate where the sixty-vote majority no longer existed. A presidential commitment was the only instrument available that could deliver strengthened abortion language without touching the text of the bill. The order therefore occupied a precise procedural niche. It could not change the statute, but it did not need to; it needed only to give a dozen House members a stated reason to vote yes, and it did.

The order satisfied almost no one outside the bloc it was designed to move. Major anti-abortion organizations, including the Susan B. Anthony List, the National Right to Life Committee, the United States Conference of Catholic Bishops, and the Family Research Council, condemned it as ineffective. They argued that executive orders can be rescinded at any time by a later administration, and the National Right to Life Committee said the order did not correct the provisions in the law that it had identified as objectionable. Legal analysts for the United States Conference of Catholic Bishops went further, arguing in a published analysis that where the order purported to fix shortcomings in the act it was likely to be legally invalid, and where it was likely to be legally valid it did nothing to fix those shortcomings. In their telling, the order could neither supply what the statute lacked nor add anything the statute did not already contain. Abortion-rights organizations opposed it from the other direction: the National Organization for Women, Planned Parenthood, NARAL Pro-Choice America, and EMILY’s List were among the groups that condemned the order, questioning the President’s commitment to the abortion-rights position. Planned Parenthood’s president, Cecile Richards, called the order a symbolic gesture. The White House, for its part, presented the order as a restatement of existing policy: press secretary Robert Gibbs said the President believed the legislation had not changed the status quo on abortion funding and that the order reiterated it. Stupak himself defended the instrument, noting that executive orders had long been used to implement policy and comparing the order to President George W. Bush’s 2007 order on stem cell funding, which he said had been welcomed by the same anti-abortion community that now dismissed Obama’s.

Taken on its own terms, the order changed no statutory text. It could not amend the Senate bill, and its legal force rested on the President’s direction to the agencies charged with implementing the law. Its significance was political rather than legislative: it gave a group of House Democrats a stated basis for voting for a bill whose abortion language they had opposed, and it gave the White House the votes to pass that bill unchanged. The order also illustrated the limits of presidential commitments as legislative instruments. A future administration could rescind it, a court could narrow it, and an agency could implement it loosely, which was exactly the anti-abortion groups’ objection. The criticisms from both sides recorded the same underlying fact in different registers. Anti-abortion groups said the order proved the statute needed fixing; abortion-rights groups said the order conceded ground the statute had already settled. Both treated the order as an instrument of vote-getting first and a policy document second, and on that much the historical record agrees with them.

The Verdict: Procedure Wrote the Text

The Affordable Care Act’s passage presents a problem that can be stated plainly, and it is the claim this article asks the reader to carry into the rest of the series. Call it the unamendable-draft problem. The Senate passed H.R. 3590 on December 24, 2009. The House then adopted that Senate text unchanged rather than sending the two chambers’ versions to a conference committee. No conference ever reconciled the House and Senate bills, no conferees scrubbed the language, and no conference report received a final vote in both chambers. The text that entered the United States Code was therefore a Senate draft that had never passed through the institution traditionally charged with finding and correcting drafting errors. Whatever imperfections, ambiguities, and infelicities the Senate text contained entered the Code uncorrected, because the procedure chosen for passage left no stage at which to correct them. A conference committee exists for exactly this purpose. In the regular order, each chamber appoints conferees, the conferees negotiate the differences between the two versions, and they produce a conference report accompanied by a joint explanatory statement describing what the final language does and why. The conference is the last point at which drafting errors can be caught, ambiguities clarified, and provisions harmonized before the text becomes law. By forgoing it, the House accepted the Senate’s draft as a finished product, and the Code received a text that had never undergone that final review.

A fair account must immediately complicate that claim, because the complication is where the strongest defense of the process lives. The principal statute was not itself a reconciliation product. The Senate considered H.R. 3590 under regular order in late 2009: it was debated, amended, and subjected to the filibuster, and it passed on December 24 with sixty votes, a cloture-proof margin that defeated extended debate by the rules everyone acknowledged. Only the amending act, H.R. 4872, traveled the reconciliation track with its twenty-hour debate cap and simple-majority threshold. The charge that the entire law was forced through on a party-line reconciliation vote misstates the record. The sixty-vote Senate passage of the underlying bill is the fact that the process’s defenders have always led with, and it is a genuine fact.

Each side’s strongest case deserves to be stated in its best form, with the characterizations attributed to those who made them. The supporters’ case is that every step followed standing parliamentary rules. The Budget Act of 1974 created reconciliation; the Byrd rule, adopted at Senator Byrd’s urging and made permanent in 1990, governed what reconciliation could carry; the parliamentarian advised and the presiding officer ruled; the special rule H.Res. 1225 structured the final House vote; and each vote was a recorded vote under rules the Senate and House had operated under for decades. Reconciliation itself was no improvisation: both parties had used it for major legislation, from deficit reduction packages to welfare reform to tax cuts, and its expedited procedures were a familiar feature of the institution. On this account, nothing was improvised and nothing was broken. The majority used the tools the institution provided, the minority used its own tools in opposition, and the parliamentarian’s rulings cut against both sides, striking nine Republican amendments and two of the majority’s own Pell grant provisions. The minority’s amendments were defeated by votes and points of order, not suppressed by fiat, and the majority’s bill was itself amended on the floor when the parliamentarian found it wanting. Legitimacy, in this telling, means compliance with the rules as written, and the record shows compliance.

The critics’ strongest case is not that rules were broken but that the rules used were the wrong ones for the work at hand. On this account, the two-step design circumvented the conference committee, the institution in which House and Senate conferees traditionally reconcile differences between the chambers’ bills, negotiate final language, and correct drafting errors before the text becomes law. Critics of the process argued that skipping conference deprived the bill of the deliberation that step was designed to supply, and that a measure of this scope deserved the scrutiny a conference provides. The House never had the opportunity to amend the Senate text at all: the choice presented to it was to accept the Senate bill verbatim or to let the entire effort fail, which is a thinner form of legislative deliberation than the Constitution’s bicameral design contemplates. The unamendable-draft problem is the critics’ exhibit: whatever one thinks of the statute’s policies, its text never received the final scrubbing that the regular order exists to provide, and the ambiguities that entered the Code are the traceable consequence of that omission. This is a case about deliberation, not legality, and it does not require showing that any rule was violated.

The verdict this article supports is that the procedural history explains the statute’s final shape better than the policy debate alone does. Procedure wrote the text. The Byrd rule acted as a co-drafter, determining which corrections the reconciliation bill could carry and which had to be left out: budgetary fixes survived, non-budgetary policy did not, and the parliamentarian’s rulings, including the two Pell grant strikes that forced a second House vote on March 25, set the boundaries of what was possible. The sixty-vote waiver threshold meant the majority could not simply overrule the parliamentarian, so the bill that became Public Law 111-152 was the text the rules permitted. At the same time, the decision to forgo conference meant the underlying Senate text entered the Code with its drafting imperfections intact. Both halves of the two-step design left fingerprints on the law: reconciliation shaped the amendments, and the absence of conference shaped the base text. Even the executive order belongs in this procedural accounting. It was not legislation, but it performed a legislative function, supplying the final votes that the parliamentary arithmetic required. The order, the Byrd rule rulings, the special rule, and the skipped conference are all entries in the same ledger: the law that emerged was the product of the rules under which it was forced to pass, quite as much as the policies its authors intended.

There is no contradiction in holding both of the contending judgments at once. The process was lawful under the standing rules, and it was thinner in deliberation than the regular order would have supplied. The supporters are right that every step followed the rules, and the critics are right that following those rules still left the statute’s text unscrubbed. A reader who keeps both propositions in view will understand why the law looks the way it does: a Senate draft enacted verbatim, amended by a reconciliation bill whose contents the Byrd rule dictated, with an executive order supplying the final votes. That is what the procedure produced, and the procedure is the explanation. To study the passage of the Affordable Care Act without studying the procedure is to mistake the outcome for the whole story.

Use this article as the frame for the series’ other entries. The articles on the Affordable Care Act’s key provisions describe what the law did, section by section. The vote analysis articles record who voted and how, in each chamber and on each motion. This article explains why the rules of passage left their fingerprints on the text itself: why some corrections were made and others were not, why the House voted twice in two days on related measures, and why the statute that emerged carried both its negotiated compromises and its uncorrected rough edges. Read together, the three kinds of articles give the full picture: what passed, who passed it, and how the procedure wrote what they passed.

The Two-Bill Path

Every recorded vote on the path from committee products to signed law, in the order it happened.

Date	Bill	Chamber	Tally	Function	Text enacted
2009-10-08	H.R. 3590	House	416-0	Passed the shell bill, the Service Members Home Ownership Tax Act, under suspension of the rules	Original six-page tax text; no health provisions
2009-11-07	H.R. 3962	House	240-194	Adopted the Stupak-Pitts amendment on abortion	Amendment text inside the House bill; never enacted
2009-11-07	H.R. 3962	House	220-215	Passed the House health bill	House text; never enacted; set aside after the Massachusetts election
2009-11-21	H.R. 3590	Senate	60-39	Cloture on the motion to proceed, Record Vote 353	None; opened floor debate
2009-12-23	H.R. 3590	Senate	60-39	Final cloture, Record Vote 395	None; closed debate
2009-12-24	H.R. 3590	Senate	60-39	Passage, Record Vote 396	Senate health text, the base of the final law
2010-03-21	H.Res. 1203	House	224-206	Special rule structuring the final votes	None; procedural
2010-03-21	H.R. 3590	House	219-212	Concurrence in the Senate amendments, Roll no. 165	Senate text unchanged; became Public Law 111-148 on March 23
2010-03-21	H.R. 4872	House	220-211	Passage of the reconciliation bill	Amendments package; not yet law
2010-03-25	H.R. 4872	Senate	56-43	Passage under reconciliation rules, as amended by Byrd rule strikes	Amended text returned to the House
2010-03-25	H.R. 4872	House	220-211	Concurrence in the Senate amendment	Reconciliation text; became Public Law 111-152 on March 30

Study This History

The two-bill maneuver rewards close study because its logic is entirely procedural. To trace each step yourself, open the legislation study notebook and map every vote in the table above to its parliamentary function: which votes required sixty, which required a simple majority, and which votes the strategy was designed to avoid entirely. Note how the Byrd rule narrowed the reconciliation bill, how the Origination Clause dictated the shell vehicle, and how the loss of a single Senate seat closed the conference path. The pattern repeats whenever a narrow majority must fit a large bill through rules written for a supermajority.

Frequently Asked Questions

Q: How did the Affordable Care Act pass Congress?

The Affordable Care Act traveled an unusual path through Congress because its final form depended on the rules of the Senate at each stage. The House first passed its own health bill, H.R. 3962, on November 7, 2009, by a vote of 220 to 215. The Senate meanwhile took up H.R. 3590, a House-passed shell bill, substituted its own health text, and passed it on December 24, 2009, with exactly 60 votes. After Senator Scott Brown of Massachusetts won the January 19, 2010 special election, Senate Democrats lost the 60-vote supermajority needed to pass any revised compromise. The House therefore concurred in the Senate amendments to H.R. 3590 unchanged on March 21, 2010, by a vote of 219 to 212. A separate reconciliation bill, H.R. 4872, then made targeted changes to the enacted law. President Barack Obama signed the principal statute on March 23, 2010, and the reconciliation measure on March 30, 2010.

Q: Did the Affordable Care Act pass through reconciliation?

No, not the principal statute. The Affordable Care Act itself, Public Law 111-148, passed under the Senate’s regular order, which required 60 votes to end debate. The Senate approved it on December 24, 2009, by a 60 to 39 vote, and the House concurred in the Senate amendments on March 21, 2010, by a vote of 219 to 212. Reconciliation entered the story only afterward. Because Senate Democrats had lost their 60-vote supermajority after the Massachusetts special election, they used the budget reconciliation process for H.R. 4872, the Health Care and Education Reconciliation Act, which amended the already-enacted law. Reconciliation allows certain budget-related measures to pass with a simple majority and limited debate. The Senate passed H.R. 4872 on March 25, 2010, by 56 to 43, and President Barack Obama signed it as Public Law 111-152 on March 30, 2010.

Q: How did Scott Brown’s election change the Affordable Care Act?

Senator Scott Brown’s victory in the Massachusetts special election on January 19, 2010, stripped Senate Democrats of their 60-vote supermajority when he was sworn in on February 4, 2010. That supermajority had been essential: the Senate health bill had passed on December 24, 2009, with exactly 60 votes, and any revised bill produced by a House-Senate conference would have needed 60 votes again to end debate. With only 59 Democrats and independents, no further compromise version could clear the Senate. The election therefore forced a new strategy. The House abandoned its own bill and its planned conference with the Senate, and instead concurred in the already-passed Senate bill unchanged. Democrats then used budget reconciliation, which required only a simple majority, to enact a separate package of changes in H.R. 4872. The bill that reached the President’s desk was thus the Senate’s December text plus the reconciliation amendments, not a negotiated compromise.

Q: Why did the House pass the Senate version of the Affordable Care Act unchanged?

The House passed the Senate bill unchanged because there was no other route to get a bill to the President. After the Massachusetts special election, Senate Democrats held only 59 seats once Senator Scott Brown was sworn in on February 4, 2010, so a conference report reconciling the House and Senate bills could not have mustered the 60 votes needed to end Senate debate. The Senate bill, H.R. 3590, had already passed that chamber on December 24, 2009, with 60 votes. By concurring in the Senate amendments without alteration, the House avoided a second Senate vote on the principal bill entirely; the measure went straight to the President’s desk. House leaders paired that vote with H.R. 4872, a reconciliation bill that carried changes the House wanted, since reconciliation needed only a simple majority in the Senate. The House concurred 219 to 212 on March 21, 2010.

Q: What bill number became the Affordable Care Act?

H.R. 3590. As introduced in the House, that number belonged to the Service Members Home Ownership Tax Act, a short tax measure that the House passed on October 8, 2009, by a vote of 416 to 0. The Senate then used the bill as a legislative vehicle: it stripped out the original text and substituted a complete health care bill, a step that let the revenue provisions of the health measure move through a House-originated bill as the Constitution’s Origination Clause requires. The Senate passed the substituted H.R. 3590 on December 24, 2009, and the House concurred in the Senate amendments on March 21, 2010. President Barack Obama signed it as Public Law 111-148 on March 23, 2010. The companion reconciliation measure carried a different number, H.R. 4872.

Q: What was the Cornhusker Kickback in the Affordable Care Act?

The Cornhusker Kickback was critics’ nickname for a provision in the Senate bill that gave Nebraska full federal funding for its share of the Medicaid expansion. Because Medicaid expansion costs were normally shared between the federal government and the states, permanent full federal funding for one state stood out, and opponents seized on it as a special deal. The provision was widely understood as connected to the support of Senator Ben Nelson of Nebraska, whose vote was the sixtieth Democrats needed to pass the Senate bill. Senator Nelson had also secured separate language on abortion coverage in the Senate bill. The Nebraska provision did not survive: it was removed in the reconciliation process, so the final enacted law contained no special Nebraska Medicaid funding arrangement.

Q: How many months did the Affordable Care Act take to pass?

There is no single canonical figure, so the answer depends on the endpoints chosen. Measured from the House’s floor vote on its own health bill, H.R. 3962, on November 7, 2009, to President Barack Obama’s signing of the reconciliation amendments on March 30, 2010, the span was just under five months. Measured from the Senate’s first cloture vote to begin debate, on November 21, 2009, to the signing of the principal statute on March 23, 2010, it was about four months. The core Senate floor fight itself, from the motion to proceed through final passage on December 24, 2009, took just over a month. If the count starts with the committee work of the preceding summer and fall, the full legislative effort stretched roughly nine to ten months. The commonly cited shorthand of about four months refers to the November-to-March floor and signing sequence.

Q: What did the Stupak amendment do to the Affordable Care Act?

The Stupak-Pitts amendment, offered by Representative Bart Stupak of Michigan, tightened abortion restrictions in the House health bill by barring the use of federal funds for abortion coverage beyond the limits of the Hyde Amendment. The House adopted the amendment on November 7, 2009, by a vote of 240 to 194, and then passed H.R. 3962 the same day. The Senate bill, however, carried different abortion language associated with Senator Ben Nelson, and after the Massachusetts special election the House had to accept the Senate text unchanged, without the Stupak-Pitts language. Abortion opponents in the House, led by Stupak, then threatened to withhold their votes. The impasse was resolved when President Barack Obama announced on March 21, 2010, that he would issue Executive Order 13535 enforcing Hyde Amendment restrictions, which secured the final votes for passage.

Q: What was H.R. 3590 before it became the health care bill?

Before it carried a single word of health policy, H.R. 3590 was the Service Members Home Ownership Tax Act, a six-page measure that extended a homebuyer tax credit for members of the armed services. The House passed it on October 8, 2009, by a unanimous vote of 416 to 0, and it went to the Senate. Senate leaders then used it as a shell: they stripped out its original text and inserted the full Senate health care bill, a maneuver that allowed the health measure, which contained revenue provisions, to travel through a bill that had originated in the House, as required by the Origination Clause of the Constitution. The number H.R. 3590 thus followed two entirely different bills through the legislative process, and it is the Senate’s substituted text that President Barack Obama signed as Public Law 111-148.

Q: What was the November 21, 2009 motion-to-proceed vote?

The motion-to-proceed vote was the procedural step that allowed the Senate to begin formal debate on the health care bill. Under Senate rules, moving to consideration of the measure required 60 votes to cut off debate on the motion itself. On November 21, 2009, the Senate voted 60 to 39 to invoke cloture on the motion to proceed to H.R. 3590, with every Democrat and independent voting in favor. The vote was closely watched because it tested whether the Democratic majority could hold together its 60-vote coalition, and it signaled that the bill would advance to weeks of floor debate and amendment. It did not decide the substance of the bill; it decided only that the Senate would take it up at all.

Q: Why did the Senate hold its final health care vote on Christmas Eve 2009?

The Christmas Eve timing was the product of the Senate’s procedural calendar. Majority Leader Reid had filed cloture on the bill, and after the Senate invoked cloture on December 23 by 60 to 39, the rules permitted up to thirty hours of post-cloture debate, which pushed the final roll call into the morning of December 24. Democratic leaders had scheduled the endgame to finish before the new year, keeping the sixty-vote coalition intact through the final vote rather than letting the fight drag into January. The Senate passed the bill that morning by 60 to 39, with Senator Jim Bunning absent. The timing meant the House would take up the Senate bill in the new year.

Q: Why were two Pell grant provisions removed from the reconciliation bill?

The two Pell grant provisions were removed because they violated the Byrd rule, the Senate’s restriction on what budget reconciliation bills may contain. The Byrd rule bars provisions that do not produce a change in federal spending or revenues, that increase the deficit beyond the budget window, or that fall outside the jurisdiction of the committee that reported them. Any senator may raise a point of order against such a provision, and it takes 60 votes to waive the rule. In the case of H.R. 4872, the Senate parliamentarian advised that the two Pell grant provisions ran afoul of the rule, and the provisions were struck from the bill. Because the Senate had altered the House-passed version, the House had to vote on the bill again, concurring 220 to 211 on March 25, 2010.

Q: What happened to the House-passed health care bill, H.R. 3962?

H.R. 3962, the Affordable Health Care for America Act, was the health bill the House passed on November 7, 2009, by a vote of 220 to 215, after adopting the Stupak-Pitts amendment on abortion by 240 to 194. The Senate never took up the House bill; instead it built its own version inside H.R. 3590 and passed that on December 24, 2009. The original plan was for a House-Senate conference committee to merge the two bills into a final compromise. That plan collapsed after the January 19, 2010 Massachusetts special election cost Democrats their 60th Senate vote, because a conference report could not have survived a Senate filibuster. The House therefore set H.R. 3962 aside and concurred in the Senate bill unchanged, so H.R. 3962 never became law.

Q: What was the Louisiana Purchase in the Senate health care bill?

The Louisiana Purchase was opponents’ nickname for a provision in the Senate health bill that gave Louisiana special Medicaid funding treatment. Like the better-known Cornhusker Kickback for Nebraska, it was attacked as a state-specific deal inserted to secure a senator’s vote during the drive to assemble 60 votes. The provision was associated with the support of Senator Mary Landrieu of Louisiana. Unlike the Nebraska provision, which the reconciliation act repealed outright, the Louisiana-related disaster-recovery Medicaid adjustment survived in modified form rather than as a Louisiana-specific earmark. The episode, along with the Cornhusker Kickback, became shorthand for the bargains struck to hold the Senate’s 60-vote coalition together.

Q: What was the Nelson amendment?

The Nelson amendment is shorthand for the abortion-coverage compromise language in the Senate health bill associated with Senator Ben Nelson of Nebraska. Nelson, the sixtieth Democratic vote, insisted on tighter restrictions on the use of federal funds for abortion coverage than the House had accepted, and the compromise language associated with his name imposed them in the Senate bill. Anti-abortion House members considered even the Nelson language too weak, and after the Massachusetts special election the House could no longer add the stronger Stupak-Pitts restrictions because it had to accept the Senate text unchanged. The Nelson language thus defined the abortion compromise that survived into law, paired with Executive Order 13535, in which President Barack Obama committed to enforcing Hyde Amendment restrictions. Nelson’s vote, and the Nebraska Medicaid provision linked to his support, made him the pivotal senator in the endgame.

Q: What did Executive Order 13535 require?

Executive Order 13535, which President Barack Obama announced on March 21, 2010, and signed on March 24, 2010, committed the executive branch to enforcing the abortion funding restrictions of the Hyde Amendment within the new health care law. The order directed federal agencies to ensure that no federal funds under the Affordable Care Act would be used for abortions beyond the Hyde Amendment’s exceptions, which cover cases of rape, incest, and danger to the life of the mother. The order was a political commitment, not a statute: it was issued to win the votes of anti-abortion House Democrats led by Representative Bart Stupak, who had threatened to oppose the Senate bill because it lacked the Stupak-Pitts restrictions. With that assurance in hand, Stupak and his allies voted yes, and the House concurred in the Senate bill 219 to 212.

Q: What was H.R. 4872?

H.R. 4872 was the Health Care and Education Reconciliation Act, the budget reconciliation bill that amended the Affordable Care Act after its principal statute was enacted. Because it moved under reconciliation rules, it needed only a simple majority in the Senate rather than the 60 votes required to end a filibuster, which mattered after Democrats lost their supermajority in the Massachusetts special election. The House passed it on March 21, 2010, by 220 to 211. The Senate passed it on March 25, 2010, by 56 to 43, after the parliamentarian’s Byrd rule rulings struck two Pell grant provisions, and the House concurred the same day by 220 to 211. President Barack Obama signed it as Public Law 111-152 on March 30, 2010. It repealed the Cornhusker Kickback, among other changes.

Q: What were the final vote margins for the Affordable Care Act?

The decisive votes were all close. In the Senate, the motion to proceed to the bill won cloture on November 21, 2009, by 60 to 39, and the Senate invoked final cloture on December 23 by the same 60 to 39 margin. The Senate passed the bill on December 24, 2009, by 60 to 39, with Senator Jim Bunning absent. In the House, the rule governing the final vote, H.Res. 1203, was adopted on March 21, 2010, by 224 to 206; the House then concurred in the Senate amendments to H.R. 3590 by 219 to 212 and passed the reconciliation bill, H.R. 4872, by 220 to 211. The Senate passed H.R. 4872 on March 25, 2010, by 56 to 43, and the House concurred that day by 220 to 211. Every margin showed how narrow the path was.

Q: On what dates was the Affordable Care Act signed into law?

The Affordable Care Act was signed on two dates because it reached the President in two pieces. President Barack Obama signed the principal statute, the Patient Protection and Affordable Care Act, on March 23, 2010, two days after the House concurred in the Senate amendments; it became Public Law 111-148. The companion reconciliation measure, the Health Care and Education Reconciliation Act, followed a separate path through the Senate under budget reconciliation rules and was signed on March 30, 2010, as Public Law 111-152. Together the two laws formed the complete Affordable Care Act: the first established the framework, and the second amended it, removing provisions such as the Cornhusker Kickback.

Q: What role did H.Res. 1203 play in the final House vote?

H.Res. 1203 was the House rule that set the terms for the climactic March 21, 2010 votes. Adopted by a vote of 224 to 206, it structured the floor debate and provided for the House to concur in the Senate amendments to H.R. 3590 without changing a word of the Senate text, a condition that the post-Massachusetts arithmetic made unavoidable. The rule also arranged consideration of H.R. 4872, the reconciliation bill carrying the changes House members wanted but could not insert into the Senate bill itself. Because the rule was adopted on a nearly party-line vote, it showed how tightly House leaders controlled the endgame. Its passage cleared the way for the 219 to 212 concurrence vote and the 220 to 211 vote on H.R. 4872 later that day.