The Wall Street Reform and Consumer Protection Act, known everywhere by the names of its two sponsors, became law on July 21, 2010, as Public Law 111-203. In the Statutes at Large it is cited as 124 Stat. 1376. The two hundred thirty-one days between introduction and signing hold one of the most instructive passage stories in modern congressional history, because the journey breaks into five distinct stages, each governed by a different procedural logic, and because the pivotal moments are unusually well documented. A reader who follows the whole path can explain how a measure that drew no votes from the minority party in one chamber nonetheless depended on minority votes to survive in the other, can name the amendments that were added and the more far-reaching ones that failed, can describe a televised conference committee that reopened after it had finished its work, and can account for a majority-party senator who voted against the finished product for being too weak.

The paradox at the center of the story is easy to state and harder to absorb. In the House of Representatives, the bill passed in December 2009 without a single Republican vote. In the Senate, where sixty votes were required to break a filibuster, the majority party did not command sixty reliable votes on its own at the decisive moments, so the final text was shaped by a handful of senators from the minority whose support had to be earned provision by provision. The House record supports the reading that one party wrote the law; the Senate record does not. Both records are true at once, and the tension between them is the point of this account.

What follows is a stage-by-stage reconstruction drawn from the congressional record, roll-call tallies, and contemporary reporting from 2010. It begins with the House vote of December 11, 2009, follows the Senate through three failed attempts to begin debate and a successful cloture vote on a substitute amendment, catalogs the floor amendments that were adopted and the restrictive alternatives that were defeated, reconstructs the public conference of June 2010 and its extraordinary reopening, and ends with final passage and the signature that made H.R. 4173 of the 111th Congress into Public Law 111-203. The through line is procedure determining text: at each stage, the rules of the chamber decided which provisions lived, which died, and which were softened enough to secure the next vote.

The article also answers the counter-reading directly. If the statute had been written entirely by one party, the conference committee would not have been reopened at the demand of a single minority-party senator, the swaps push-out provision would not have been softened to keep his vote, and the nineteen billion dollar pay-for would not have been replaced with a different funding mix. Those events happened because sixty votes were required and the majority did not have them alone. The record of which provisions each pivotal vote bought is unusually complete, and it is laid out here in full.

Legislative histories of major statutes often blur the sequence of events into a single narrative of partisan struggle, and the blur hides the mechanism. The mechanism here is unusually visible because each stage left a numbered record. Roll Call 968 in the House, Record Votes 124, 126, 127, 158, 160, and 162 in the Senate, the 20-11 and 7-5 conference tallies, Roll Call 413 on the conference report, and Record Votes 206 and 208 on the final Senate action form a chain in which every link can be inspected. The chain shows a majority party that could pass a bill on its own in one chamber and could not take a single dispositive step without minority help in the other. That asymmetry is the engine of the story, and it explains why the final text contains provisions the House majority never sought and lacks provisions the Senate majority could not hold.

The five stages also map onto five different procedural regimes, which is why the article treats them separately. The House stage was governed by majority control of the floor. The April Senate stage was governed by the cloture rule on the motion to proceed. The May Senate stage was governed by amendment procedure under the shadow of cloture on a substitute. The conference stage was governed by the rules of reconciliation between the chambers, conducted for once in public. The final-passage stage was governed by the up-or-down arithmetic of the conference report, which cannot be amended on the floor. At each stage, different actors held influence, and the text moved accordingly. Readers who keep that mapping in mind will find the details that follow easier to place.

The documentation is worth appreciating on its own terms. The Congressional Record preserves the floor statements, congress.gov preserves the action sequence and the roll-call numbers, the House Clerk preserves the individual votes on the conference report, and contemporary press accounts from 2010 preserve the contemporaneous understanding of what each tally meant. Every roll call, date, and margin in the account below has been checked against those primary sources. Nothing that follows rests on recollection or reconstruction; every number comes from the record, and where the record is silent, as on the question of a standalone floor tally for the swaps push-out, the article says so plainly.

The test this article sets for itself is the one stated at the outset: a reader who finishes it should be able to explain how a bill opposed by nearly the entire minority in one chamber nonetheless required minority votes to survive a filibuster in the other, name the amendments added and the far more radical ones that failed, describe the televised conference committee that reopened after it had finished, and understand why a senator from the majority party voted against the finished product for being too weak. Each element of that test corresponds to a section below. The minority-vote paradox is the Senate phase. The amendment ledger is the table and the two sections around it. The reopened conference is the June narrative. Feingold’s no vote closes the final-passage section. Hold the article to that test, and judge whether it passes.

The House Moves First

H.R. 4173 entered the House on December 2, 2009, introduced by Representative Barney Frank of Massachusetts, the chairman of the House Financial Services Committee. The bill text was timestamped at 2:43 that afternoon, and from that moment the House moved with the speed of a chamber whose rules give the majority firm control of the floor. Nine days later, on Friday, December 11, 2009, the House passed the measure 223-202 on Roll Call 968. The tally did not include a single Republican vote. Twenty-seven Democrats joined every Republican present in opposition, and contemporary press accounts were unanimous on the point: no House Republican voted for the bill.

How did the House manage to pass the bill without any minority votes?

The House’s rules let a majority move legislation with a simple majority of those voting, and the Democratic caucus in December 2009 held enough seats to lose twenty-seven of its own and still prevail. The motion to recommit and the Bachus substitute were defeated on party lines, and the 223-202 tally required nothing from across the aisle.

The day’s earlier votes sketched the shape of the opposition. A motion to recommit with instructions, offered by Representative Dent, failed 190-232 on Roll 967. A substitute amendment offered by Representative Bachus, the ranking Republican on the Financial Services Committee, failed 175-251 on Roll 966. Both votes confirmed that the minority had an alternative it preferred and that the majority had the numbers to reject it. The House Democratic leadership did not need a single Republican, did not seek one in any serious way, and did not get one. The bill that left the House on December 11 was, in partisan terms, a purely Democratic product, and opponents of the statute have cited that tally as evidence that the law was written by one party.

That reading is accurate as far as the House is concerned, and the article does not dispute it. The House majority used its procedural advantages, the minority offered its substitute, the substitute lost by a wide margin, and the bill passed on party lines with Democratic defections. Nothing in the later Senate history rewrites that chapter. But the House chapter is only the first of five stages, and the rules changed completely when the bill crossed the Capitol. The Senate received H.R. 4173 on January 20, 2010, and from that point forward the majority needed something it did not possess in full: sixty votes.

The contrast between the chambers would define everything that followed. In the House, the question was whether the majority could hold its own members, and twenty-seven defections showed the limits of that hold without endangering the outcome. In the Senate, the question was whether the majority could assemble sixty, and the answer at every decisive moment was that it could not do so alone. The senators who supplied the missing votes did not supply them for free. Each extracted changes, and the changes they extracted are visible in the enacted text, from the debit-interchange amendment that sailed through with minority support to the swaps provision that was softened in conference to the pay-for that was rewritten after a single senator objected. To understand how those changes got into the law, the story must slow down and follow the Senate vote by vote.

Frank’s nine-day sprint from introduction to passage reflected the advantages of a House majority operating under rules it writes for itself. The Rules Committee sets the terms of debate, the majority controls the amendment process through structured rules, and the minority’s options narrow to the motion to recommit and whatever substitutes the rule permits. The Dent motion to recommit and the Bachus substitute were the minority’s two permitted avenues on December 11, and both lost by margins that showed the majority’s control was comfortable but not total. The 190-232 defeat of the recommit motion and the 175-251 defeat of the substitute were wider than the 223-202 passage margin, which tells its own story: some Democrats who would not vote for the minority’s alternatives also would not vote for the bill.

The twenty-seven Democratic defections on final passage deserve attention because they complicate the partisan narrative from the other direction. The bill that left the House was a Democratic product, but it was not a unanimous Democratic product. Twenty-seven members of the majority party voted no, joining every Republican present. Their reasons varied, and the record does not supply a single explanation, but their existence meant the House leadership had to manage its own right and center as well as the opposition. The leadership succeeded; 223 votes were enough, and the bill moved on.

The zero in the December tally is doing a great deal of work in later debates about the statute’s legitimacy, so it is worth stating exactly what it proves. It proves that no House Republican was willing to vote for the House’s version of financial reform in December 2009. It does not prove that no Republican in Congress was willing to vote for financial reform in any form, because the Senate record shows otherwise within months. The distinction between a chamber-specific tally and a Congress-wide conclusion is the kind of distinction this article insists on throughout, because the passage story punishes anyone who generalizes from one chamber to the other.

The December margin also set the baseline against which the June conference-report tally can be read. On June 30, 2010, the House agreed to the conference report 237-192, fourteen votes better than the December result, with three Republicans voting yes. The improvement came from two sources: the conference’s concessions, which brought a few additional Democrats aboard, and the three minority votes, the first Republican support the bill had ever received in the House. Fourteen votes is not a transformation, and the bill remained a partisan measure in the lower chamber, but the direction of movement mattered. The Senate’s demands had made the bill marginally more passable in the House, which is the opposite of what one would expect if the Senate had simply rubber-stamped the House product.

But the very completeness of the House’s control is what made the Senate’s rewrite possible. Because the House had passed a bill, the Senate had a vehicle. Because the vehicle existed, the Senate could substitute its own text without starting the legislative process over. And because the Senate’s rules gave the minority leverage the House’s rules denied it, the substituted text was negotiated under constraints the House bill had never faced. The December bill was a statement of what the majority wanted. The May substitute was a statement of what the majority could get sixty votes for. The distance between those two statements is the subject of this article.

When the Senate received H.R. 4173 on January 20, 2010, the bill entered a chamber where the majority’s control of the floor is conditional rather than absolute. Any senator can threaten extended debate, and ending that debate requires sixty votes under Rule XXII. The House’s structured process, in which the majority decides what gets a vote, has no Senate equivalent; instead, the Senate proceeds by unanimous consent agreements that any single senator can block, or by cloture motions that require a supermajority. The financial reform bill would test both paths. The weeks between the bill’s arrival and the first cloture vote were consumed by negotiation over whether the Senate would take the bill up at all, and the April roll calls answered that question in the negative three times before the chamber found another way forward.

Three Failures Before a Word of Debate

The Senate’s consideration of the financial reform legislation began not with debate but with a procedural wall. The vehicle before the Senate was S. 3217, Senator Christopher Dodd’s bill, and the majority leader moved to proceed to it. Under Senate rules, ending debate on the motion to proceed required sixty votes, and the first cloture vote, held on April 26, 2010, failed 57-41 on Record Vote 124. The majority was three votes short. The next day, April 27, the Senate voted again on reconsideration, and cloture failed a second time by the identical margin of 57-41 on Record Vote 126. On April 28, a third attempt failed 56-42 on Record Vote 127, with the majority now four votes short of the threshold.

Three failures in three days might have suggested that the bill was stuck, and for a brief period the outcome was genuinely uncertain. But the Senate has more than one way to begin debate, and on the afternoon of April 28, after the third failed cloture vote, the chamber proceeded to the measure by unanimous consent. The motion to reconsider the failed vote was withdrawn, the third cloture motion was rendered moot, and the bill was laid before the Senate without a successful cloture vote on the motion to proceed ever having occurred. That detail matters because later summaries sometimes imply that the majority eventually won a cloture vote to begin debate. It did not. It found another door.

Why did three attempts to begin Senate debate all fail in April 2010?

The majority fell short of sixty each time, losing 57-41 on April 26, 57-41 on April 27, and 56-42 on April 28, because minority senators withheld the votes needed to end debate on the motion to proceed. The Senate then proceeded through unanimous consent on April 28, a path requiring no roll call.

The April sequence established the pattern for the entire Senate phase: the majority could not reach sixty on its own, and progress depended on finding votes across the aisle or finding procedural paths around the shortfall. Readers who want the mechanics of the sixty-vote threshold explained at length can consult the companion explainer on the filibuster and cloture, which sets out how the modern Senate uses the cloture motion and why a determined minority can force the majority to negotiate. For this story, the essential point is simpler. Every subsequent turning point, from the substitute amendment to the conference report, would be decided at or near the sixty-vote line, and the senators who held the decisive votes knew it.

The failed April votes also shaped the majority’s strategy for the weeks ahead. Rather than forcing the underlying bill through a series of losing roll calls, the leadership turned to a substitute amendment that could carry the full Senate text, including the derivatives title produced by the Agriculture Committee. That substitute, known as the Dodd-Lincoln substitute, would become the real battleground, and the May votes on it would replay the April drama at higher stakes.

The motion to proceed deserves a brief explanation, because the April votes are often misunderstood. In the Senate, getting to a bill is itself a debatable question, and a single senator can insist on debating whether to debate. The motion to proceed is the leadership’s tool for forcing that question, and cloture on the motion is the tool for ending debate about it. The three April votes were therefore not votes on financial reform; they were votes on whether to consider financial reform. The distinction explains why the majority kept losing: senators who might eventually support the bill could still vote against proceeding to it while negotiations continued, using their cloture votes as bargaining chips for the amendment process to come.

Reid’s handling of the third failure showed the flexibility of Senate practice. After the 56-42 tally on April 28, the majority leader withdrew his motion to reconsider, which rendered the cloture motion moot, and the Senate proceeded to the measure by unanimous consent. Unanimous consent sounds like agreement, but in Senate practice it often means only that no senator chose to object at that moment, sometimes because the holdouts had extracted promises about the amendment process. The record does not disclose what understandings accompanied the April 28 consent, and this article does not speculate. What matters is the procedural fact: the Senate began debate without ever winning a cloture vote on the motion to proceed, and the sixty-vote problem was therefore deferred rather than solved. It would return in May, on the substitute, with higher stakes and the same arithmetic.

The effect of the April blockade was to establish the minority’s bargaining position before a word of the bill had been debated. By demonstrating that sixty votes were not available for even the preliminary step, the minority ensured that the amendment process would be negotiated rather than dictated. Every amendment adopted in May, and every amendment defeated, must be read against that background: the majority needed minority votes not only for cloture but for the orderly conduct of the floor, which gave individual senators influence far beyond their numbers. The Durbin amendment’s 64-33 margin and the Brown-Kaufman amendment’s 61-33 defeat both belong to the world the April votes created, a world in which nothing moved without a supermajority’s acquiescence.

The Substitute That Became the Bill

With the Senate finally on the bill, the majority’s vehicle for the floor fight was Amendment SA 3739, the Dodd-Lincoln substitute, proposed on April 29, 2010, in the nature of a substitute. The amendment carried the text of S. 3217, Senator Dodd’s bill, as the replacement language for H.R. 4173. In practical terms, the Senate struck the House-passed text and inserted the Senate’s own product, a routine maneuver with large consequences, because the Senate product included a derivatives title that the House bill had not contained in the same form. That title came from the Senate Agriculture Committee, chaired by Senator Blanche Lincoln of Arkansas, and it carried the provision that would become Section 716 of the enacted law.

Lincoln’s role deserves careful telling because the provision she pressed, the swaps push-out, reached the statute through an unusual path. Her vehicle was the Wall Street Transparency and Accountability Act of 2010, which the Agriculture Committee approved 13-8 on April 21, 2010, with one Republican, Senator Grassley of Iowa, joining the Democrats. The derivatives title was then folded into the Dodd-Lincoln substitute. There was no standalone Senate floor roll-call vote on the push-out provision itself. It entered the bill as part of SA 3739, which was agreed to by unanimous consent on May 20, 2010, after the successful cloture vote described below. Any account that assigns the Lincoln provision its own floor vote margin is inventing one; the record shows no such vote.

The terminology around this provision requires equal care. The swaps push-out is Section 716 of Dodd-Frank, titled “Prohibition against Federal Government bailouts of swaps entities.” It is the Lincoln provision. The Volcker Rule is Section 619, a separate provision entirely. The two are sometimes blurred in shorthand, but they do different things and they traveled through the process differently, and this article keeps them distinct throughout.

Section 716 as it stood in the substitute would have pushed certain swaps activities out of banks that benefited from federal support, forcing them into separately capitalized affiliates. The banks and their allies opposed it, Lincoln defended it, and the conference committee would later soften it substantially while keeping its core. The softening is part of the conference story told below. At this stage, the point is structural: a major title of the final statute reached the Senate floor inside a substitute amendment, never received its own roll-call vote, and survived because the substitute survived. That is how omnibus legislating works in a chamber where sixty votes are the price of everything, and it is why the May cloture votes on SA 3739 were the decisive votes of the Senate phase.

A substitute in the nature of a substitute is one of the Senate’s most powerful procedural devices. Instead of amending the House bill line by line, the Senate replaces its entire text with new language, debates and amends the replacement, and then passes the bill number with the new content. The device allowed the Senate to take H.R. 4173, a House product, and turn it into the vehicle for S. 3217, Dodd’s bill, without starting over. The text the Senate debated in May was therefore Dodd’s text, shaped by the Banking Committee he chaired, plus the Agriculture Committee’s derivatives title that Lincoln had built. The House number on the bill was, by May, little more than a shipping label.

Lincoln’s derivatives title illustrates how committee jurisdiction shapes substance. Derivatives reform fell to the Agriculture Committee because the Commodity Futures Trading Commission, which oversees futures markets, sits under that committee’s jurisdiction, a legacy of the days when futures meant agricultural commodities. Lincoln used that jurisdiction to write a sweeping title, the Wall Street Transparency and Accountability Act of 2010, and moved it through her committee 13-8 on April 21 with Grassley’s support. The 13-8 tally, with one Republican joining the Democrats, foreshadowed the bipartisan coalitions of the floor fights to come. Once folded into SA 3739, the title’s fate was tied to the substitute’s fate, which meant Lincoln’s provisions would live or die on the cloture votes of May 19 and 20 rather than on their own merits as a standalone measure.

The absence of a standalone floor vote on the push-out is worth dwelling on, because it affects how the provision’s democratic pedigree is judged. A provision that never faces its own roll call never forces senators to take a recorded position on it, which insulates it from the kind of direct accountability the Durbin amendment faced on May 13. Defenders of the practice note that the substitute itself faced the most heavily scrutinized votes of the entire process, the 57-42 and 60-40 cloture tallies, so the push-out was not smuggled in unnoticed. Critics note that voting for cloture on a thousand-page substitute is not the same as voting for any single title within it. Both observations are true, and the tension between them is characteristic of omnibus legislating under a supermajority rule. The article records the procedural fact without adjudicating the normative dispute: the push-out entered the bill inside SA 3739, SA 3739 was agreed to by unanimous consent after 60-40 cloture, and no senator ever cast a recorded vote for or against the push-out alone.

The distinction between Section 716 and Section 619, the Volcker Rule, matters for the same reason precision always matters in legislative history. The Volcker Rule restricted proprietary trading by banking entities; the Lincoln provision pushed swaps activities into separately capitalized affiliates. They addressed different activities, imposed different structures, and carried different section numbers. Conflating them, as some shorthand accounts do, misattributes the conference softening, the July 2012 effective date, and the two-year transition to the wrong provision. This article keeps the numbers straight throughout: 716 is Lincoln’s push-out, 619 is the Volcker Rule, and the softening described in the conference section belongs to 716.

Why did Dodd introduce a separate Senate bill instead of amending the House bill?

He gave the Senate’s committees the pen. S. 3217, introduced April 15, 2010, let the Banking Committee write the base text and the Agriculture Committee write the derivatives title, and the Senate substituted that product into H.R. 4173 that May. The bill number that reached the president was the House’s, but the words inside were, in substance, the Senate’s.

Three Numbers, One Bill: H.R. 4173, S. 3217, SA 3739

Readers encountering the passage record for the first time are often confused by the three identifiers, so the relationship among them is worth setting out plainly. H.R. 4173 was the House bill: introduced by Frank on December 2, 2009, passed by the House on December 11, and received by the Senate on January 20, 2010. S. 3217 was Dodd’s Senate bill: introduced on April 15, 2010, and developed through the Banking Committee he chaired. SA 3739 was the Dodd-Lincoln substitute amendment: proposed on April 29, 2010, carrying the S. 3217 text as a complete replacement for the H.R. 4173 text.

The Senate’s decision to work through the House number rather than passing S. 3217 as a freestanding bill follows standard practice when the House has acted first. Keeping a single legislative vehicle avoids the need to reconcile two separately passed bills at a later stage and preserves the House’s constitutional priority in the legislative sequence. The practical consequence was that everything the Senate did in May, the amendment fights, the cloture votes, the final passage tally, happened to H.R. 4173 as a vehicle, even though the substance being debated was Dodd’s.

On May 20, the mechanics became formal. The Senate struck all language after the enacting clause of H.R. 4173, substituted the language of S. 3217 as amended, and passed H.R. 4173 in lieu of S. 3217 by 59-39. The phrase in lieu of is the official marker of what occurred: the Senate chose the House-numbered vehicle over the Senate-numbered bill while adopting the Senate bill’s substance. Descriptions that say the Senate passed S. 3217 are therefore imprecise. The Senate passed H.R. 4173, containing S. 3217’s text, and S. 3217 as a separate bill never became law.

The vehicle question also determined the conference’s baseline. Because the Senate had substituted its text into H.R. 4173, the conferees reconciled the House-passed version of H.R. 4173 with the Senate-passed version of H.R. 4173: two versions of a single measure, not two different bills. The distinction is technical, but it defined the conference’s jurisdiction, which extended to every difference between the chambers’ texts and nothing else. The derivatives title’s journey through the three numbers illustrates the point. It began as an Agriculture Committee product, was folded into SA 3739, was agreed to as part of the substitute after the 60-40 cloture vote, became part of the Senate-passed H.R. 4173 on May 20, went to conference as Senate text, and emerged as Section 716 of the enrolled bill. Follow the numbers, and the provision’s path is fully legible; lose track of them, and the history dissolves into confusion.

How did the House’s December bill differ in fate from its June conference report?

The December bill passed with zero Republican ayes and was discarded as text: the Senate struck every word after the enacting clause and substituted its own product. The June conference report, built from the Senate’s substitute as reconciled with the House, drew three Republican ayes in the House and three in the Senate. The vehicle survived; the words did not.

What the Substitute Conceals

An amendment in the nature of a substitute is one of the Senate’s most powerful procedural devices, and SA 3739 shows why. When Senators Dodd and Lincoln proposed the substitute on April 29, 2010, they were not offering a tweak to the bill. They were offering a replacement text, carrying the Banking Committee’s product and the Agriculture Committee’s derivatives title, that would become the bill if adopted. The Senate’s subsequent votes were therefore not votes on individual provisions but votes on the package. A senator who supported the substitute’s core but opposed the Lincoln push-out had no way to express that division in a roll call. The only question put to the chamber was whether to end debate on the whole.

This is the procedural reason no standalone floor vote on the swaps push-out exists, and it is worth reflecting on what that means for democratic accountability. The most controversial provision of the Senate’s bill, the one that drew the fiercest industry opposition and the most intense lobbying, entered the legislation without any senator ever recording a yea or nay on it specifically. It was approved as part of a committee title, carried inside a substitute, adopted after a cloture vote that was about ending debate rather than endorsing contents, and then softened in conference. At each step, the provision advanced without a direct vote on its merits. The system worked as designed: committees write, substitutes package, cloture limits debate, conferences reconcile. But the design means that the provisions with the most significant economic consequences can travel the furthest with the fewest direct votes.

The contrast with the Durbin amendment sharpens the point. Durbin’s interchange provision received a standalone roll call, 64-33, with the yeas and nays of every senator recorded and attributable. The Lincoln push-out received no such vote. Both provisions became law. Both shaped the financial system for years. But only one of them can be traced to a moment when the Senate, as a body, said yes to it specifically. The other can be traced only to a committee vote, a substitute, and a conference. For readers who believe legislation should be traceable to votes, the substitute mechanism is the most challenging feature of this history. For readers who believe packaging is necessary to legislate at scale, it is the most essential.

The conference’s softening of Section 716 partially answers the accountability concern, because the conference did hold public sessions and the push-out’s final form was negotiated in the open. But the softening was itself a product of bargaining among conferees rather than a floor decision, and Lincoln’s certification that her provisions survived was a political judgment rather than a roll call. The substitute conceals, and the conference only partly reveals.

Amendments the Senate Adopted

Once the Senate was on the bill, senators offered amendments in large numbers, and the ledger of what was adopted and what failed is the clearest window into how the chamber shaped the text. The most consequential adopted amendment was also the one with the most bipartisan margin: the Durbin interchange amendment, SA 3989, offered by Senator Richard Durbin of Illinois.

The Durbin amendment directed the Federal Reserve to set debit-card interchange fees at levels deemed reasonable and proportional, and it exempted issuers with less than ten billion dollars in assets. On May 13, 2010, the Senate adopted it 64-33. The margin included seventeen Republicans and forty-seven Democrats, a coalition that crossed party lines more decisively than almost any other vote in the bill’s history. A floor statement entered into the Congressional Record summarized the result plainly: the amendment passed with sixty-four votes, seventeen from Republicans and forty-seven from Democrats, and it was accepted in conference. The provision survived the conference committee and appears in the enacted statute.

The Durbin vote is worth pausing over because it cuts against the simple story that the minority opposed everything. On this amendment, a substantial bloc of minority senators voted yes, and their votes were part of why the amendment’s margin was so comfortable. The amendment’s substance, regulating the fees that merchants pay on debit transactions, drew support from senators responsive to retailers and community banks, and the ten-billion-dollar exemption gave smaller issuers a reason not to fight it. Whatever the motives, the roll call is a matter of record: 64-33, with minority votes supplying a large share of the winning margin.

Other adopted amendments belong in the ledger as well. SA 3827, the Shelby-Dodd amendment concerning emergency lending and bailout authority, was adopted on May 5, a bipartisan product of negotiation between the ranking Republican and the chairman of the Banking Committee. It stands as a reminder that the amendment process was not purely a contest between the parties; some of the adopted changes were negotiated across the aisle before they ever reached the floor. The ledger table below collects the significant entries with their sponsors, outcomes, margins, and effects on the enacted text, so that the full pattern is visible in one place.

The substance of the Durbin amendment repays attention because its design explains its coalition. Debit interchange fees are paid by merchants to card-issuing banks each time a customer pays with a debit card, and retailers had complained for years that the fees were set without competitive pressure. By directing the Federal Reserve to limit the fees to amounts deemed reasonable and proportional, the amendment delivered a tangible benefit to merchants, a constituency with representation in every state. The exemption for issuers under ten billion dollars in assets protected community banks, which blunted the opposition that the banking industry might otherwise have mobilized. The result was a provision that united retailers’ allies in both parties while dividing the banking lobby’s friends, and the 64-33 tally reflected that alignment.

The seventeen Republican yes votes on Durbin deserve emphasis because they are the single largest bloc of minority support for any element of the bill. On the passage votes, minority support never exceeded three senators. On Durbin, it reached seventeen, which meant the amendment would have passed comfortably even without a single Democratic vote beyond the forty-seven it received. The amendment’s adoption was therefore not a case of the majority buying minority votes; it was a case of a bipartisan majority forming around a provision both parties’ members wanted. That distinction matters for the article’s thesis. The sixty-vote rule forced negotiation, but negotiation sometimes produced genuine agreement rather than mere transaction, and Durbin is the best evidence of it.

The Shelby-Dodd amendment, SA 3827, adopted May 5, belongs to a different category: the negotiated bipartisan product. Named for the Banking Committee’s ranking Republican and its chairman, the amendment addressed emergency lending authority, placing limits on the Federal Reserve’s power to extend credit in crises. Its adoption showed that the committee’s bipartisan channel remained open even while the floor fights raged, and it foreshadowed the conference’s willingness to accept provisions with cross-party parentage. The Congressional Record’s account of the July 15 final debate later described the Merkley-Levin Amendment No. 4101 as having been offered to the Dodd-Lincoln substitute as the basis of the provision adopted by the conference committee, a reminder that the substitute served as the chassis on which many members hung their priorities. The amendment process in the Senate operates by simple majority once debate is underway, which is why individual amendments could pass with margins like 64-33 while cloture on the underlying vehicle still required sixty. That asymmetry, simple majority for amendments and supermajority for the bill, is what made the amendment ledger the true record of the Senate’s substantive choices.

The Congressional Record’s account of the Durbin result adds a final layer of confirmation. A floor statement entered into the Record summarized the outcome in plain terms: the amendment passed with sixty-four votes, seventeen from Republicans and forty-seven from Democrats, and it was accepted in conference. The phrase accepted in conference carries a specific meaning in legislative practice. It means the conferees, presented with a Senate provision the House text had not contained, chose to retain it in the reconciled bill rather than dropping it or modifying it beyond recognition. For a floor amendment adopted over the objections of much of the banking industry, acceptance in conference was not automatic; it reflected the same bipartisan coalition that had produced the 64-33 margin, carried forward into the reconciliation stage.

The amendment ledger also serves the article’s findable artifact: a single table in which each significant floor and conference amendment is listed with who offered it, whether it was adopted, the margin, and what it changed in the enacted text. Readers who want to track each entry in their own notes can use VaultBook’s legislation study notebook, which is built for exactly that kind of provision-by-provision tracking.

The Amendment Ledger

The table below gathers the amendments that shaped the statute, on the floor and in conference. Each row names the sponsor, records the outcome and the margin where a roll call exists, and states what the amendment changed in the enacted text. The rows are ordered by the sequence in which the events occurred.

Amendment	Offered by	Outcome	Margin	Effect on enacted text
Bachus substitute amendment (House)	Rep. Bachus	Failed	175-251, Roll 966, Dec 11, 2009	Would have replaced the House bill with the minority’s alternative; its defeat confirmed the majority’s text as the House product
Dent motion to recommit with instructions (House)	Rep. Dent	Failed	190-232, Roll 967, Dec 11, 2009	The minority’s last attempt to alter the House bill before passage; rejected on party lines
SA 3739, Dodd-Lincoln substitute	Sens. Dodd and Lincoln	Adopted by unanimous consent after cloture invoked	60-40 cloture, May 20, 2010	Carried the S. 3217 text, including the Agriculture Committee derivatives title with the Section 716 swaps push-out, into H.R. 4173
SA 3827, Shelby-Dodd emergency lending amendment	Sens. Shelby and Dodd	Adopted	May 5, 2010; no roll-call margin in the verified record	Placed limits on emergency lending authority in the enacted text
SA 3989, Durbin debit-interchange amendment	Sen. Durbin	Adopted	64-33, May 13, 2010	Directed the Federal Reserve to set reasonable and proportional debit interchange fees, exempting issuers under ten billion dollars in assets; accepted in conference
SA 3789, Brown-Kaufman size-cap amendment	Sens. Brown and Kaufman	Failed	61-33, May 6, 2010	Not enacted; the more restrictive alternative to the final bill’s approach to large-bank size
Merkley-Levin Amendment No. 4101	Sens. Merkley and Levin	Adopted via the substitute and conference	Offered to SA 3739 during floor debate; no standalone margin	Described in the Congressional Record as the basis of the provision adopted by the conference committee
Lincoln swaps push-out, Agriculture title	Sen. Lincoln, via the Agriculture Committee	Folded into SA 3739; no standalone floor vote	None recorded	Enacted as Section 716, softened in conference with a July 2012 effective date, a two-year transition, and exemptions for hedging and bank-eligible swaps
Scott Brown pay-for objection	Sen. Scott Brown	Conferees agreed to revise	June 29, 2010	The nineteen billion dollar bank assessment was dropped; replaced with early termination of TARP spending authority and increased FDIC assessments
Bachus motion to recommit the conference report (House)	Rep. Bachus	Failed	198-229, Roll 412, June 30, 2010	The minority’s last attempt to alter the conference report before the House agreed to it 237-192

The ledger makes the pattern legible. Adopted amendments drew bipartisan coalitions, the failed restrictive alternative drew a bipartisan majority against it, the most structurally important provision never faced its own vote, and the last entry was not a floor amendment at all but a single senator’s objection that forced a completed conference to reopen. Each of these entries is examined in turn in the sections that follow.

A few patterns in the ledger deserve to be drawn out. First, the adopted amendments all drew bipartisan majorities, while the most restrictive failed amendment drew a bipartisan majority against it. The Senate of May 2010 was not a chamber divided into two disciplined blocs; it was a chamber in which coalitions formed and reformed around each provision, with the sixty-vote threshold disciplining the extremes. Second, the row with no recorded margin, the Lincoln push-out, is a finding in itself. The absence of a number where every other row has one is the documentary trace of the substitute procedure, and it reminds the reader that some of the statute’s most consequential provisions never faced a direct vote. Third, the final row is not an amendment at all but an objection, and its presence in an amendment ledger is the article’s quiet argument: in a sixty-vote Senate, a single senator’s letter can function as the most powerful amendment of all, rewriting a completed conference report without a roll call.

The Amendment That Failed

If the Durbin amendment showed what the Senate would adopt, the Brown-Kaufman amendment showed what it would not. SA 3789, offered by Senator Sherrod Brown of Ohio and Senator Ted Kaufman of Delaware, would have capped the size of large financial institutions by limiting bank liabilities and deposits as a share of the economy. Its sponsors framed it with the slogan that an institution too big to fail was too big to exist, and contemporary coverage described it neutrally as the more restrictive alternative to the bill’s approach: rather than regulating large banks more tightly, it would have made them smaller by law.

On May 6, 2010, the Senate defeated the amendment 61-33 on Record Vote 136. The coalition against it was bipartisan in the same way the Durbin coalition had been bipartisan for its amendment: a majority of both parties concluded that a hard statutory cap went further than the chamber was willing to go. Three Republicans joined twenty-nine Democrats and Senator Sanders in supporting the cap, which meant the amendment’s backers also crossed party lines, but thirty-three votes were far short of the sixty needed for adoption and far short of a simple majority as well. The amendment was later withdrawn on May 20 during the endgame proceedings, but the dispositive vote was the May 6 defeat, and the ledger records it as such.

The Brown-Kaufman vote matters to the passage story for two reasons. First, it answers the question of which amendments failed on the floor with the clearest possible example: the most far-reaching structural alternative offered in the Senate lost by twenty-eight votes. Second, it illuminates the later vote of Senator Feingold, who opposed the final bill for being too weak. The defeat of Brown-Kaufman was an early signal that the Senate would not impose the toughest available constraints on large banks, and senators who wanted those constraints took notice. The bill that emerged was the product of what sixty senators could agree to, and the Brown-Kaufman tally showed where the chamber’s limits lay.

It is worth stating plainly what the 61-33 margin means and what it does not. It means that a decisive majority of the Senate, including most Democrats, rejected a statutory size cap. It does not mean that the Senate rejected reform; the same chamber would pass the bill itself two weeks later. The distinction is the one this article keeps returning to: the Senate’s choices were bounded by the sixty-vote requirement, and within those bounds the chamber chose regulation over restructuring. Readers who prefer the restructuring approach can point to the Brown-Kaufman roll call as the moment their alternative lost. Readers who prefer the enacted approach can point to the same roll call as the moment the chamber settled the question.

Brown and Kaufman were an unlikely pair in seniority but a natural one in conviction. Brown, the senior senator from Ohio, had built his career on skepticism of financial concentration; Kaufman, appointed to fill the Delaware seat, had used his brief tenure to press for structural constraints on large banks. Their amendment translated a moral judgment, that institutions too big to fail should not exist, into a statutory formula: caps on liabilities and deposits as a share of the economy. The formula’s virtue was its clarity; its vice, in the eyes of its opponents, was its rigidity. A fixed cap would have forced divestitures regardless of economic conditions, and the amendment’s critics, in both parties, argued that the bill’s regulatory approach could achieve stability without the disruption.

The coalition for the cap is as instructive as the coalition against it. Three Republicans voted yes, which meant the amendment’s support, like the Durbin amendment’s, crossed party lines. Twenty-nine Democrats and Senator Sanders joined them. But thirty-three votes is not close to sixty, and it is not even close to fifty-one. The amendment’s defeat was decisive, not narrow, and its decisiveness is what made it a signal rather than a near miss. Senators who wanted structural downsizing learned on May 6 that the chamber would not go there; Feingold’s later votes suggest he learned the lesson and acted on it.

The amendment’s withdrawal on May 20, during the endgame, was a formality. Once the substitute had been agreed to and the Senate was moving to passage, the defeated amendment had no vehicle left, and its sponsors withdrew it rather than forcing a second recorded defeat. The ledger records the May 6 tally as dispositive, which is how the Congressional Record treats it as well. For readers keeping score of the Senate’s substantive choices, the Brown-Kaufman roll call is the moment the chamber defined the outer boundary of what it would do to large banks, and everything the conference later produced stayed inside that boundary.

The amendment’s framing deserves a last note because the language around it has sometimes been heated. Its sponsors’ slogan, that an institution too big to fail was too big to exist, was a moral claim about the financial system, and contemporary coverage described the proposal neutrally as the more restrictive alternative to the bill’s regulatory approach. This article follows that neutral description throughout. The 61-33 defeat does not show that the Senate favored large banks; it shows that a bipartisan majority preferred to regulate them rather than to break them up by statutory formula. Both positions were held in good faith, both were argued on the floor, and the roll call settled the question for the 111th Congress.

Sixty Votes, Twice in Two Days

The decisive Senate votes came on May 19 and May 20, 2010, and they concerned not the bill itself but the Dodd-Lincoln substitute, SA 3739. Because the substitute carried the entire Senate text, including the derivatives title, the cloture votes on it were the votes that determined whether the bill could move forward in its Senate form. The sequence is frequently misremembered, so it is set out here exactly as the record shows.

On May 19, the Senate voted on cloture on SA 3739, and cloture was not invoked: 57-42 on Record Vote 158. The majority was again three votes short, and a cloture motion on the bill itself was withdrawn by unanimous consent the same day. Contemporary press accounts confirmed the failure and treated the bill’s prospects as uncertain. The next day, May 20, the Senate voted again on reconsideration, and this time cloture was invoked 60-40 on Record Vote 160. The sixty-vote threshold was met exactly, with no vote to spare.

What changed between the failed May 19 vote and the successful May 20 vote?

Senator Scott Brown of Massachusetts, who had opposed cloture on May 19, switched to yes on May 20, and Senator Arlen Specter of Pennsylvania, absent on May 19, returned and voted yes. Those two changes moved the tally from 57-42 to 60-40, while Democratic Senators Cantwell and Feingold voted against cloture on the successful motion.

The arithmetic of the switch is the whole story of the Senate phase in miniature. Brown’s yes and Specter’s return supplied the margin; without either one, cloture would have failed again. The two Democrats who voted no on cloture, Cantwell of Washington and Feingold of Wisconsin, underscored that the majority’s own members were not unanimous, which made the minority votes load-bearing rather than decorative. SA 3739 was then agreed to by unanimous consent, and the Senate moved to final passage.

Later that same day, May 20, the Senate struck all after the enacting clause of H.R. 4173, substituted the language of S. 3217 as amended, and passed H.R. 4173 in lieu of S. 3217 by 59-39 on Record Vote 162. The vehicle that left the Senate was therefore H.R. 4173 in name but S. 3217 in substance, which is why histories of the statute sometimes describe the Senate as having passed its own bill. In form, the House bill number survived; in content, the Senate’s text replaced it. Feingold and Cantwell were the two Democrats who voted against passage, an early indication of the discontent that would later make Feingold the lone Democratic no vote on the conference report.

The May 20 passage vote is also the moment to correct a common error. Some summaries state that the Senate passed the bill 60-39 on May 19. The record shows otherwise: May 19 was a failed cloture vote, 57-42, on the substitute amendment, and the successful cloture vote was May 20, 60-40, also on the substitute. Final passage came later on May 20 at 59-39. The distinction between cloture on an amendment and passage of a bill is not a technicality here; it is the mechanism by which the Senate actually legislates, and getting the dates and margins right is what separates an accurate passage history from a garbled one.

May 19 deserves to be reconstructed in full, because the day’s failure is the most commonly misreported event in the bill’s history. The Senate convened with the Dodd-Lincoln substitute pending and a cloture motion filed on it. The roll call, Record Vote 158, came back 57-42: three votes short of the sixty needed. The same day, the leadership withdrew by unanimous consent a cloture motion it had filed on the bill itself, an acknowledgment that the substitute was the only vehicle that mattered. Contemporary press accounts treated the failure as a genuine setback, not a scheduling maneuver, and described the bill’s prospects in uncertain terms. The majority had now failed to invoke cloture four times in less than a month, counting the three April votes on the motion to proceed, and the pattern was becoming a story in itself.

The overnight turnaround between May 19 and May 20 is where individual senators became visible as individuals rather than as party members. Brown’s switch from no to yes supplied one vote. Specter’s return from absence supplied another. Neither change was mysterious: Brown had been negotiating over provisions he wanted, and his yes vote on cloture was the consideration for the concessions he would continue to extract through June. Specter’s absence on May 19 had been the more contingent event; his presence on May 20 restored a vote the majority had been counting on. The two changes together moved the tally from 57-42 to 60-40, exactly the threshold, with no margin for error.

One senator switched. One senator returned. Two senators voted no. The rest held. There was no surge, no wave, no last-minute stampede. There was a whip operation that found exactly the votes it needed and not one more, which is how sixty-vote thresholds are typically met when the underlying coalition is fragile.

The two Democratic no votes on the successful cloture motion, Cantwell and Feingold, are the detail that prevents any simple story about party discipline. A majority that cannot hold its own members on a cloture vote is a majority that must negotiate outward, and the May 20 tally is the proof. Sixty votes were assembled from fifty-eight Democrats and two Republicans, with two Democrats in opposition. The arithmetic made Brown and Specter, not any member of the leadership, the decisive actors of the day.

What followed the cloture vote was swift. SA 3739 was agreed to by unanimous consent, the objection-free path the Senate uses when the outcome is no longer in doubt. The Senate then struck all after the enacting clause of H.R. 4173, substituted the S. 3217 text as amended, and passed the bill 59-39 on Record Vote 162. The one-vote difference between the 60-40 cloture tally and the 59-39 passage tally reflected the difference between ending debate and adopting the measure; not every senator who votes to end debate votes for the bill, and not every senator present for one roll call is present for the next. Feingold and Cantwell voted no on passage, as they had on cloture, and their opposition would persist through July.

A final note on the record-keeping: the distinction between the May 19 and May 20 votes is the single most important factual discipline in this article. Summaries that collapse them into a single event get the date wrong, the margin wrong, and the motion wrong, all at once. The verified sequence is May 19, cloture not invoked, 57-42, on SA 3739; May 20, cloture invoked on reconsideration, 60-40, on SA 3739; May 20, passage, 59-39, of H.R. 4173 as amended. Any account that differs from that sequence is not supported by the Congressional Record.

What did the 59-39 passage vote decide that the 60-40 cloture vote had not?

Cloture ended debate on the substitute; passage enacted the bill. The 60-40 vote on May 20 overcame the filibuster on SA 3739. The 59-39 vote later that day struck the House text after the enacting clause, substituted the S. 3217 language as amended, and passed H.R. 4173 in lieu of S. 3217.

A Conference in Public

With each chamber having passed its own version, the bill went to a conference committee to reconcile the differences, and the conference that followed was unusual among modern conferences: it was public. The conferees held sessions on June 10, 15, 16, 17, 22, 23, and 24, 2010, and contemporary accounts described two weeks of official, publicly televised negotiations. Conference committees often do their real bargaining behind closed doors, with the public sessions serving as theater. In this case the public sessions were extensive enough, and the coverage detailed enough, that the bargaining itself left a visible record. Readers who want the general mechanics of how conferences reconcile House and Senate texts can consult the companion explainer on conference committees; what matters here is what this particular conference did with the text.

The conference completed its work in the early morning hours of June 25, 2010. The votes to approve the agreement were 20-11 among the House conferees and 7-5 among the Senate conferees, margins that reflected the partisan composition of the delegations. The agreement reconciled hundreds of differences between the chambers, but three sets of decisions are the ones the passage story requires.

First, the conference kept the Lincoln swaps push-out as Section 716 but softened it. The effective date was postponed to July 2012, later than the July 2011 date that applied to most of the derivatives title, and a discretionary two-year transition period was added. The push-out was made prospective, so banks could retain swaps executed before the end of the transition period. Exemptions were added for hedging and risk-mitigation activities and for swaps in bank-eligible securities such as interest-rate and currency swaps, and the text clarified that merely being a major swap participant would not trigger the push-out. Contemporary summaries put the practical effect plainly: banks could keep their interest-rate and foreign-exchange swaps and their hedging books, while credit-default swaps and commodity, energy, and equity swaps would have to move to separately capitalized affiliates within up to two years. Lincoln herself hailed the agreement as retaining her derivatives provisions, which was true of the core and generous about the edges.

Second, the conference dropped the pre-funded resolution-fund concepts that both chambers had considered. The House and Senate had each entertained the idea of a fund, paid for in advance by assessments on large firms, that would finance the orderly wind-down of failing institutions. The conference replaced those concepts with post-paid assessments, meaning the costs of any future resolution would be collected after the fact rather than banked in advance. The final bill therefore contained no pre-funded orderly liquidation fund, a point worth stating because later commentary sometimes assumes otherwise.

Third, the conference agreement carried a nineteen billion dollar assessment on large financial firms as the general pay-for for the bill. The assessment was designed to offset the bill’s cost as scored by the Congressional Budget Office, which put the conference agreement at roughly twenty billion dollars over ten years. The nineteen billion dollar number was not tied to any single program or fund; it was the price of making the bill’s budget arithmetic work. That pay-for would not survive the week, and its replacement is the subject of the next section. For the account of the consumer bureau title and what the conference left in regulators’ hands, the companion piece on the bureau’s creation and powers takes up the story where the passage history leaves it.

The public character of the conference is part of why the passage record is so complete. Televised sessions meant that the softening of Section 716, the abandonment of the pre-funded fund, and the adoption of the nineteen billion dollar pay-for all happened in view of the press and the public. When the conference finished on June 25, the agreement looked final. It was not.

Each element of the Section 716 softening reflected a specific negotiation, and the elements are worth enumerating because they show what softening meant in practice. Postponing the effective date to July 2012, a year later than the July 2011 date for most of the derivatives title, gave banks additional time to restructure. Adding a discretionary two-year transition period gave regulators room to extend the timeline further. Making the push-out prospective, so that swaps executed before the end of the transition period could be retained, protected existing books from forced liquidation. Exempting hedging and risk-mitigation activities preserved the core risk-management functions that even the provision’s supporters agreed banks should keep. Exempting swaps in bank-eligible securities, including interest-rate and currency swaps, removed the largest and most liquid markets from the provision’s reach. And clarifying that mere status as a major swap participant would not trigger the push-out narrowed the provision’s scope to the activities it was meant to capture. Taken together, the changes left the push-out’s architecture intact while reducing its practical bite, which is why Lincoln could hail the agreement as retaining her provisions while the banks could treat the outcome as manageable.

The abandonment of the pre-funded resolution fund was the conference’s other major structural decision. Both chambers had considered establishing a fund, financed in advance by assessments on large firms, to pay for future orderly liquidations. The conference replaced those concepts with post-paid assessments, under which the industry would be charged after a resolution rather than before. The choice reflected a judgment, shared by the conferees, that an advance fund was politically and practically difficult to sustain; it also removed one of the bill’s most visible potential costs to large firms. The final statute’s silence on a pre-funded fund is therefore not an oversight but a decision, and readers who assume the law created such a fund are misremembering the conference’s work.

The nineteen billion dollar pay-for that the first conference agreement carried needs the same careful treatment. The assessment on large financial firms was not earmarked for any program; it was a general offset for the bill’s budget cost, which the Congressional Budget Office had scored at roughly twenty billion dollars over ten years, a figure Representative Frank later cited. The pay-for existed because congressional budget rules required the bill’s costs to be offset, not because any title of the bill needed nineteen billion dollars in funding. Understanding the assessment as a budgetary plug rather than a programmatic investment is essential to understanding why it could be replaced so readily on June 29: plugs can be swapped for other plugs, and the conferees did exactly that.

The televised character of the conference amplified all of these decisions. Two weeks of public sessions meant that the Section 716 negotiations, the fund debate, and the pay-for discussions unfolded under press coverage, with outside groups and industry lobbyists able to track the bargaining in something close to real time. Whether the transparency changed the outcomes is unknowable from the record, but it changed the historical record itself, leaving a documentary trail far richer than the closed-door conferences that produced most major statutes. The votes that completed the work, 20-11 among House conferees and 7-5 among Senate conferees in the early hours of June 25, were therefore cast in public, on provisions the public had watched being negotiated.

Contemporary descriptions of the conference as two weeks of official, publicly televised negotiations capture something genuinely unusual in congressional practice. Most conference committees meet in public only for opening statements, then retreat to private bargaining, with the public record consisting of the final report and little else. Here the sessions themselves were the record: seven sitting days across fifteen calendar days, with the softening of the swaps push-out, the fate of the resolution fund, and the construction of the pay-for all debated where reporters and cameras could follow. The transparency did not prevent the June 29 reopening, which happened through a letter and a same-day reconvening rather than through extended public deliberation, but it did ensure that the reopening itself was immediately visible. A closed conference could have revised its report quietly; this one revised its report in the open, and the contrast between the June 25 completion and the June 29 revision became part of the public story of the bill.

What did the conference’s 20-11 and 7-5 closing votes signify?

They showed a conference divided along party lines but functional: the majorities in both delegations approved the June 25 report. The real fight came after, when Brown’s objection to the pay-for forced the June 29 reopening. The televised majorities settled the regulatory text; a single senator’s letter settled the funding.

The Reopening

On June 28, 2010, Senator Robert Byrd of West Virginia died. His death left the Senate with a vacancy that changed the arithmetic of the conference report’s prospects: the majority could no longer afford to lose a single vote and still reach sixty. Into that narrowed margin stepped Senator Scott Brown of Massachusetts, the same senator whose switch had supplied the decisive cloture vote on May 20. On June 29, Brown sent a letter to Chairmen Dodd and Frank stating that if the final version of the bill contained the higher taxes, meaning the nineteen billion dollar assessment, he would not support it. The letter’s operative sentence was blunt: if the final version of this bill contains these higher taxes, he would not support it. Senators Snowe and Collins of Maine, the other two Republicans who had voted for cloture in May, were also reported to have qualms about the fee.

What forced the conference to reconvene after June 25?

Senator Scott Brown’s June 29 letter warning that he would oppose the bill over the nineteen billion dollar bank assessment, combined with Senator Byrd’s death the day before, left the majority unable to lose any vote. The conferees reconvened the same day and produced a revised agreement without the assessment.

The conference that had completed its work in the early hours of June 25 was reopened on June 29, an extraordinary step. Conferees do not ordinarily revisit a finished agreement, and they did so here only because the votes required it. That evening, at 8:30, the revised conference report was filed as H. Rept. 111-517. The nineteen billion dollar assessment was gone. In its place, the conferees ended the spending authority of the Troubled Asset Relief Program early, later in that same month instead of in October, which yielded about eleven billion dollars, and increased deposit-insurance assessments levied by the FDIC on banks to cover the remainder.

How did one senator’s objection replace a nineteen billion dollar pay-for?

Brown’s letter made his vote contingent on dropping the assessment, and with Byrd’s seat vacant the majority had no vote to spare. The conferees therefore substituted two smaller offsets: ending TARP spending authority months early for about eleven billion dollars, plus higher FDIC assessments on banks, and filed the revised report that night.

The episode is the clearest possible illustration of the article’s thesis. A single senator, holding one of the sixty votes needed for cloture, objected to a single provision of a completed conference agreement, and the conference reopened and rewrote the pay-for to keep his vote. Nothing about the episode was accidental or personal; it was the structural consequence of the sixty-vote requirement operating on a majority that did not have sixty votes to spare. The revised report that emerged on June 29 was the text that would go to the floors of both chambers for final passage, and every vote from that point forward was a vote on the Brown-revised version.

The context of Byrd’s death on June 28 requires no embellishment. The Senate’s most senior member died three days after the conference finished, and his seat sat vacant. In a chamber operating at exactly sixty votes, a vacancy is not a sad footnote but a mathematical event: the majority’s margin for error, already zero on the May and July tallies, became negative in the sense that any single defection would now be fatal. The leadership understood this instantly, and so did Brown, whose letter of June 29 arrived with the timing of someone who had done the same arithmetic.

Brown’s letter to Dodd and Frank was brief and conditional. Its operative sentence warned that if the final version of the bill contained the higher taxes, he would not support it. The phrase higher taxes was doing political work, framing the nineteen billion dollar assessment as a tax increase rather than a regulatory offset, but the procedural meaning was plain: the assessment had to go, or his vote would go with it. Snowe and Collins were reported to share qualms about the fee, which meant the problem might have extended beyond a single senator, but Brown’s letter was the documented trigger, and the record treats it as such.

The two Maine senators’ reported unease is worth pausing over because it shows how thin the margin truly was. Collins and Snowe had supplied two of the three minority yes votes in May and would do so again in July; their support was never unconditional, and the nineteen billion dollar assessment tested it. Had their qualms hardened into opposition alongside Brown’s, the leadership would have faced a shortfall no procedural maneuver could fix. The conference’s decision to replace the pay-for therefore bought not only Brown’s vote but insurance against further erosion, a consideration the conferees could hardly have ignored with Byrd’s seat vacant and the calendar pressing toward the July recess.

The conferees’ response was as swift as the arithmetic demanded. On June 29, the same day as the letter, they reconvened, agreed to drop the assessment, and filed the revised report, H. Rept. 111-517, at 8:30 that evening. The replacement offsets were chosen for speed as much as for substance. Ending the Troubled Asset Relief Program’s spending authority early, later in June rather than in October, freed about eleven billion dollars in budget authority that could be redirected. Increasing the FDIC’s deposit-insurance assessments on banks covered the remainder. Neither offset required new legislative architecture; both repurposed existing authorities, which is why they could be assembled in a single day. The revised report was the text the chambers would vote on, and the nineteen billion dollar assessment passed into history as a provision that survived four days, from June 25 to June 29, before a single letter removed it.

The Conference Report’s Two Lives: June 25 and June 29

The conference committee produced two reports, and the differences between them are the most concentrated lesson in the statute’s history about who writes the law. The first report was completed in the early morning hours of June 25, 2010, after two weeks of televised sessions, and approved 20-11 by the House conferees and 7-5 by the Senate conferees. The second was filed on June 29 at 8:30 in the evening, after a one-day reconvening prompted by Senator Brown’s letter. The regulatory provisions were identical in both. The funding was not. Everything else the conference had decided, the Durbin amendment’s acceptance, the Lincoln push-out’s softening, the abandonment of the pre-funded resolution fund, survived the reopening untouched. Only the pay-for changed, and the pay-for changed because one vote required it.

This selectivity is itself revealing. The conferees did not use the reopening to revisit substantive compromises, even though the same leverage that forced the funding change could theoretically have forced others. The reason is that Brown’s demand was specific: his letter objected to the higher taxes, meaning the nineteen billion dollar assessment, and conditioned his support on their removal. He did not demand changes to the derivatives title or the debit-fee provisions. The leadership, operating with no margin, granted exactly what was demanded and nothing more. The episode shows how precisely pivotal voters can target their interventions. Brown did not rewrite the bill. He repriced it.

The two reports also illustrate the difference between the conference’s public phase and its private phase. The June 25 report was the product of open negotiation, with the press characterizing the sessions as two weeks of official, publicly televised conference negotiations. The June 29 revision was the product of urgent, closed bargaining, conducted in a single day under the pressure of a fixed House vote on June 30. The public phase produced the statute’s regulatory architecture. The private phase produced its funding. Both phases left their marks on the text, but only the first was visible while it happened. The second is visible only in retrospect, in the difference between the two filed reports and in the letter that forced the difference.

For readers tracking the money, the substitution’s mechanics are worth spelling out. The original nineteen billion dollar assessment would have fallen on large financial firms as a dedicated levy, sized to offset the bill’s scored cost to the budget. Its replacement had two parts with different incidence. Ending TARP’s spending authority early, later that month instead of in October, freed about eleven billion dollars by canceling authority the government had not used. Increasing FDIC deposit-insurance assessments shifted the remaining cost onto banks through higher premiums for the insurance that protected their depositors. The first part was a bookkeeping acceleration. The second part was a real cost imposed on a defined set of institutions. Together they replaced a tax-like assessment with a combination of canceled spending authority and higher insurance premiums, a different political bargain even if the budgetary bottom line was comparable.

What did the conference’s first round decide that the reopening left alone?

Everything regulatory. The June 29 revision touched only the pay-for: the televised decisions on the Durbin amendment, the softened Lincoln push-out, and the dropped pre-funded resolution fund all survived intact. Brown’s letter demanded removal of the assessment, not revision of the regulatory titles, and the conferees granted exactly what was demanded.

An Unamendable Choice: The Conference Report on the Floor

A conference report occupies a peculiar place in congressional procedure: it is the only text both chambers will ever vote on, and neither chamber may amend it. The rule against amending conference reports is what gives conferees their power, because the chambers must take the agreement or leave it, and it is what made the June 29 revision so consequential. Once H. Rept. 111-517 was filed, every subsequent vote was a binary choice on the Brown-revised text, with no opportunity to restore the nineteen billion dollar assessment, unsoften Section 716, or revive the pre-funded fund. Senators and representatives who disliked individual provisions had to weigh them against the whole package, which is why Grassley’s no vote and Feingold’s no vote are best read as judgments that the package, on balance, failed their tests.

The House’s handling of the report illustrated the majority’s control of the binary choice. H. Res. 1490, the rule governing debate on the report, passed 234-189 on Roll 410 with no Republican votes, setting the terms under which the House would consider the unamendable text. The Bachus motion to recommit, the minority’s last procedural weapon, failed 198-229 on Roll 412. The motion to recommit is worth understanding: it is the minority’s right to offer final instructions before passage, and its defeat by thirty-one votes showed that the majority’s coalition, though not bipartisan, was disciplined. The final 237-192 tally on Roll 413 was therefore the product of a process designed to produce exactly one outcome, a yes or no on the conferees’ work.

The Senate’s binary choice operated under tighter constraints. Cloture on the conference report required sixty votes, and the 60-38 morning tally on Record Vote 206 supplied exactly that. The afternoon’s 60-39 agreement on Record Vote 208 was the substantive decision, but it was the cloture vote that determined whether the decision could be made at all. In a chamber where any senator can extend debate indefinitely, the unamendable report is not exempt from the filibuster, which means the conferees’ power is bounded by the same sixty-vote arithmetic that bounded everything else. The June 29 revision was the conferees’ response to that boundary: they rewrote the pay-for because the alternative was a failed cloture vote and no bill at all.

The unamendable character of the report also explains why the Brown episode could not have happened on the floor. Had the nineteen billion dollar assessment still been in the report on July 15, Brown could have voted no, but he could not have amended the assessment out; the Senate would have faced the same binary choice, and the bill might have failed. By forcing the change in conference, where the text could still be rewritten, Brown achieved what no floor amendment could have achieved after filing. The episode is a lesson in the timing of influence: the power to shape a conference report belongs to those who act before it is filed, and Brown acted on June 29, hours before the filing deadline made the text final.

Final Passage in Both Chambers

The House acted first on the revised conference report. On June 30, 2010, at 6:54 in the evening, the House agreed to the conference report 237-192 on Roll Call 413. The party breakdown told the story of a bill that had gained a little minority support since December without gaining much: 234 Democrats voted yes and 19 voted no, while 3 Republicans voted yes and 173 voted no, with 4 members not voting. The three Republican ayes were Representative Joseph Cao of Louisiana, Representative Mike Castle of Delaware, and Representative Walter Jones of North Carolina. The record is specific on this point because a common error names a different third Republican; Representative Judy Biggert of Illinois voted no on the conference report, and the Clerk’s individual vote record confirms Cao, Castle, and Jones as the three.

The day’s preliminary votes showed the majority’s control of the floor. The rule for considering the conference report, H. Res. 1490, passed 234-189 on Roll 410 with no Republican votes, and a motion to recommit offered by Representative Bachus failed 198-229 on Roll 412. The House had moved from zero Republican votes in December to three in June, a shift that registered the conference’s concessions without changing the bill’s fundamentally partisan character in the lower chamber.

The Senate took up the conference report on July 15, 2010. That morning, the Senate invoked cloture 60-38 on Record Vote 206, and in the afternoon it agreed to the conference report 60-39 on Record Vote 208. The three Republican ayes were Senators Scott Brown of Massachusetts, Susan Collins of Maine, and Olympia Snowe of Maine, the same trio that had supplied the decisive cloture votes in May. The lone Democratic no vote was Senator Russ Feingold of Wisconsin. Senator Grassley of Iowa, who had voted for the Senate bill in May and whose Agriculture Committee vote had helped advance the derivatives title, voted against the conference report, citing the derivatives language and the offsets. Senator Crapo of Idaho did not vote, and the seat of the late Senator Byrd remained vacant.

Why did the majority party need minority votes if it held the Senate majority?

The majority did not hold sixty reliable votes on its own. Cloture required sixty, the April attempts failed at 57-41 and 56-42, the May 19 attempt failed at 57-42, and the May 20 and July 15 successes reached exactly sixty only with minority senators voting yes. Without those votes, debate could not be ended.

Feingold’s no vote requires its own telling, because it is the vote that completes the article’s opening promise: the senator from the majority party who voted against the bill for being too weak. Feingold had also voted against the Senate bill on May 20, one of two Democrats to do so, and his stated reason never wavered. He considered the measure insufficiently restrictive on the banking industry. His floor statement put the standard plainly: after thirty years of giving in to the wishes of Wall Street lobbyists, Congress needed to finally enact tough reforms, but the key reforms were not included in the bill. The test for the legislation was a simple one: whether it would prevent another financial crisis. As the bill stood, he said, it fails that test. Contemporary coverage summarized his position without embellishment: he voted against the measure because it was not restrictive enough. The defeat of the Brown-Kaufman size cap in May had been the preview; Feingold’s July no vote was the conclusion.

With the Senate’s 60-39 vote on July 15, the bill was cleared for the President. On July 21, 2010, President Barack Obama signed it into law as Public Law 111-203, cited in the Statutes at Large as 124 Stat. 1376. The five stages were complete: House passage in December with no minority votes, a Senate spring in which every decisive vote required minority support, a public conference in June, a reopened conference at one senator’s demand, and final passage in July on near-party lines in the House and a sixty-vote coalition in the Senate.

The House’s June 30 proceedings followed the choreography of a majority confident of its numbers. The rule for considering the conference report, H. Res. 1490, passed 234-189 on Roll 410 with no Republican support, which showed that the minority’s opposition extended even to the procedural question of whether the report should be debated. The Bachus motion to recommit failed 198-229 on Roll 412, a wider margin than the final tally, repeating the December pattern in which the minority’s alternatives lost by more than the bill itself won. Then, at 6:54 p.m., the House agreed to the conference report 237-192 on Roll 413.

The party breakdown of Roll 413 repays a close look. Democrats voted 234 to 19 in favor, a substantial improvement in party unity over December’s 27 defections. Republicans voted 3 to 173 in favor, which is to say the minority remained overwhelmingly opposed while conceding three votes. The three Republican ayes, Cao of Louisiana, Castle of Delaware, and Jones of North Carolina, were not a bloc; they were three individuals whose districts and judgments led them to break with their party. The Clerk’s individual vote record is the authority for their names, and it is also the authority for the correction this article observes: Biggert of Illinois, sometimes misidentified as the third Republican yes, voted no. Precision about these names matters because the three votes are the entire House-side evidence for the claim that the minority participated in the final product.

The Senate’s July 15 was a study in the same sixty-vote arithmetic that had governed May. Cloture was invoked in the morning, 60-38 on Record Vote 206, and the report was agreed to in the afternoon, 60-39 on Record Vote 208. The Republican ayes were Brown, Collins, and Snowe, the trio whose votes had been load-bearing since May. Feingold’s no vote made him the only Democrat in opposition, a distinction he had effectively held since May 20. Grassley’s no vote is the subtler data point: a senator who had voted for the bill in May and advanced the derivatives title in committee rejected the conference product over the revised derivatives language and the new offsets. His trajectory shows that Senate support was conditional on content, not tribal, and that the conference’s revisions had costs as well as benefits. Crapo’s absence and the vacant Byrd seat complete the picture of a chamber voting at the absolute limit of the possible.

The one-vote difference between the morning’s 60-38 cloture tally and the afternoon’s 60-39 passage tally is a small arithmetic lesson in Senate attendance. Cloture and passage are separate questions, put to the chamber hours apart, and the membership present for each can differ. A senator who supports ending debate may oppose the underlying measure, and a senator absent in the morning may be present in the afternoon, or the reverse. The July 15 pair shows the threshold clearing twice with slightly different coalitions, which is all the rules require: sixty to end debate, a majority of those present to adopt the report, with the political reality that anything short of sixty on the report itself would have invited a filibuster of the enrollment. The leadership needed sixty twice, and twice it assembled exactly that.

Feingold’s July no vote was the culmination of a position he had held without wavering since May. His floor statement’s standard, whether the legislation would prevent another financial crisis, was a test the bill could not pass in his judgment because the key reforms were missing. The statement’s bluntness, that it fails that test, is unusual in Senate debate, where members more often praise bills they oppose for their good intentions. Feingold offered no such praise. His vote is the article’s promised exhibit: a majority-party senator who opposed the bill for being too weak, and whose opposition was substantive rather than procedural. The Brown-Kaufman defeat had shown him in May that the Senate would not go further; his July vote was his answer to that showing.

The Arithmetic of Every Roll Call

The passage story can be retold as a sequence of numbers, and the retelling is useful because the numbers are the argument. On December 11, 2009, the House defeated the Bachus substitute 175-251 on Roll 966, defeated the Dent motion to recommit 190-232 on Roll 967, and passed the bill 223-202 on Roll 968. The three tallies trace a majority that could reject the minority’s alternatives by wide margins while passing its own bill by a narrower one, the signature of a chamber where twenty-seven majority-party defections narrowed but did not threaten the outcome.

The Senate’s April sequence reads 57-41, 57-41, 56-42 on Record Votes 124, 126, and 127. The repetition of 57-41 on the first two votes shows a minority holding its line across days; the slip to 56-42 on the third shows the majority losing ground rather than gaining it. None of the three reached sixty, and the chamber’s response, proceeding by unanimous consent, is the procedural equivalent of changing the subject: rather than winning the vote, the majority found a path that required no vote.

May’s numbers tell a denser story. The Shelby-Dodd amendment’s adoption on May 5 has no recorded margin in this account’s sources beyond its adoption, which itself illustrates how voice votes and unanimous consent can move provisions without roll calls. The Brown-Kaufman defeat, 61-33 on Record Vote 136, is the mirror image: a recorded, decisive rejection. The Durbin adoption, 64-33, is the high-water mark of bipartisanship in the entire process. Then the cloture sequence: 57-42 on May 19, 60-40 on May 20, both on SA 3739, followed by 59-39 passage on Record Vote 162. The progression from 57 to 60 to 59 across three roll calls in two days is the quantitative fingerprint of the pivotal-voter dynamic: two senators changed the outcome, and the outcome changed the statute.

June’s numbers belong to the conference: 20-11 and 7-5 on June 25, the committee’s internal majorities, followed by the refiling of H. Rept. 111-517 on June 29, a number rather than a tally, marking the moment the agreement changed without a recorded committee vote on the changes. The House’s June 30 sequence reads 234-189 on the rule, 198-229 on the recommit motion, and 237-192 on the report, Roll Calls 410, 412, and 413. The Senate’s July 15 reads 60-38 on cloture and 60-39 on the report, Record Votes 206 and 208. The final Senate numbers return to the threshold that governed everything: sixty, reached exactly, twice in one day.

Read together, the roll calls show a majority that never once had votes to spare in the Senate and rarely needed spares in the House. That is the entire structural argument in numerical form, and it is why the article dwells on margins that other histories round away.

The party-unity trends across the seven months are visible in the numbers as well. House Democrats went from twenty-seven defections in December to nineteen on the conference report, a gain in cohesion that reflected the conference’s concessions and the leadership’s whipping. House Republicans went from zero yes votes to three, a small but symbolically significant movement. In the Senate, the majority’s internal divisions were the constant: two Democrats opposed cloture on May 20, two opposed passage the same day, and one opposed the conference report in July, while the minority’s yes votes held steady at the three senators whose support had been negotiated. The direction of movement favored the bill’s supporters in the House and held even in the Senate, which is why the final tallies, 237-192 and 60-39, look more comfortable than the spring’s knife-edge votes without ever escaping the sixty-vote constraint.

The frequency with which the Senate landed on exactly sixty is itself a finding. The May 20 cloture vote was 60-40, the July 15 cloture vote was 60-38, and the July 15 passage vote was 60-39. Three decisive roll calls, three tallies at or within two votes of the threshold, none with room to spare. Majorities that command sixty-five or seventy votes can absorb defections; this majority could absorb none, which is why Brown’s switch, Specter’s return, and the Collins and Snowe votes were not merely helpful but necessary. The closeness of the tallies is the quantitative signature of the four-vote statute: a majority operating at the absolute limit of its strength, purchasing each additional vote at the price the seller demanded.

What the Record Does Not Show

An honest passage history reports its silences alongside its tallies, and this record has several. The most consequential silence is the missing roll call on the swaps push-out. Because the provision entered the bill inside the Dodd-Lincoln substitute, no senator ever recorded a yes or no on Section 716 as such. The provision’s supporters can point to the 60-40 cloture vote and the unanimous-consent agreement to SA 3739 as its democratic warrant; its critics can reply that neither action tested the provision on its own. The record does not resolve that dispute, and this article does not pretend otherwise.

A second silence surrounds the unanimous-consent agreement of April 28. The Senate proceeded to the bill without objection, but the record does not disclose what understandings, if any, the holdouts extracted in exchange for withholding their objections. Unanimous consent agreements are often accompanied by side understandings about the amendment process, and the orderly conduct of the May amendment fights suggests some framework was in place, but the specifics are not in the sources this article relies on, and the article does not invent them.

A third set of caveats concerns sourcing. The zero-Republican-vote figure for December 11 rests on unanimous contemporary press reporting rather than a directly retrieved Clerk individual-vote page; the consensus is strong enough to state the figure, but the sourcing is worth disclosing. The 64-33 Durbin margin is confirmed by multiple contemporary outlets and corroborated by a Congressional Record floor statement quoting the 64-vote figure with its party breakdown. The Statutes at Large citation, 124 Stat. 1376, is the standard citation confirmed through legal sources rather than a direct government pull. None of these caveats weakens the account’s core claims, but stating them is part of the article’s commitment to show its work.

What the silences share is a common lesson: procedural choices determine not only outcomes but also what the historical record preserves. A standalone vote creates a number; a substitute procedure creates a gap where a number would have been. The gaps are themselves evidence of how the Senate legislates, and a passage history that noted only the numbers would miss half the story.

One more matter of the record is worth naming because retellings of this history commonly get it wrong. There was no successful 60-39 cloture vote on May 19; the May 19 cloture vote failed 57-42 and the successful cloture vote came on May 20. The House conference-report Republicans were Cao, Castle, and Walter Jones, not the names sometimes attached to that vote. And the swaps push-out never received a standalone Senate floor vote. These corrections are disclosed here rather than buried because the article’s authority depends on them: a passage history that silently repeated the errors would be worse than no history at all. Readers who check the roll-call numbers cited here against the primary sources will find them as stated, and where the retellings and the record disagreed, the record won.

Two No Votes from the Majority

The majority’s internal divisions are easy to overlook in a story dominated by the sixty-vote threshold, but the record preserves them with the same precision as everything else. On May 20, 2010, when the Senate invoked cloture on the Dodd-Lincoln substitute 60-40, two Democratic senators voted no: Maria Cantwell of Washington and Russ Feingold of Wisconsin. Later that day, when the Senate passed the bill 59-39, the same two Democrats voted no. On July 15, when the Senate agreed to the conference report 60-39, Feingold voted no as the only Democrat in opposition. Three roll calls, three no votes from Feingold, two from Cantwell, all from the majority’s own side.

Feingold’s rationale is a matter of record, and this article has quoted it: the bill was not restrictive enough on the banking industry, the key reforms were missing, and the legislation failed his test of whether it would prevent another financial crisis. His position was consistent across May and July, and it was substantive rather than procedural. He did not object to the process; he objected to the product. The defeat of the Brown-Kaufman amendment on May 6 had shown him that the Senate would not impose the toughest available constraints, and his subsequent votes were the logical consequence. A senator who believed the bill was too weak had only one tool, the no vote, and Feingold used it three times.

Cantwell’s votes are recorded with the same precision but without an accompanying explanation in the sources this article relies on. The record documents her no votes on cloture and on passage without attributing a stated rationale, and this article does not supply one. The votes themselves are the finding: on the two most important Senate roll calls of May 20, the majority could not hold all of its members, which is why the sixty-vote threshold had to be met with minority help. Whatever Cantwell’s reasons, the arithmetic consequence was identical to Feingold’s. Each Democratic no vote raised the price of the minority votes needed to replace it.

The two no votes also illuminate the article’s structural claim from an unexpected angle. In a sixty-vote chamber, defections from the majority do not merely embarrass the leadership; they transfer bargaining power to the minority, because every lost majority vote must be replaced from across the aisle. Cantwell’s and Feingold’s opposition therefore increased Brown’s, Collins’s, and Snowe’s influence, since their yes votes became the replacements. The purists of the majority, by withholding their support, empowered the marginal voters of the minority to demand more. That dynamic is not a criticism of any participant; it is the mechanical consequence of the threshold, and it operated the same way regardless of what motivated each dissenter.

Feingold’s July 15 distinction, the lone Democratic no on the conference report, deserves a final emphasis. In a vote decided 60-39, a single additional Democratic defection would not have changed the outcome, but it would have changed the story, and the fact that no second Democrat joined him shows the limits of the too-weak critique within the caucus. Feingold stood alone, by his own account on principle, and the bill passed without him. The episode completes the article’s opening promise: the majority-party senator who voted against the finished product for being too weak, documented across three roll calls and explained in his own words.

On July 15, their paths diverged. Feingold voted against the conference report, becoming the only Democratic senator to do so. Cantwell supported it. The divergence is instructive. Between May and July, the conference had softened some provisions and rewritten the pay-for, but it had not fundamentally altered the bill’s regulatory architecture. What changed was the political context: the bill was now a completed conference report, the alternative to passage was the collapse of seven months of work, and the majority needed every vote it could hold. Cantwell evidently concluded that the conference product, whatever its shortcomings, was worth supporting. Feingold concluded that it was not.

There is a final irony in the Cantwell-Feingold story. The two senators who opposed the bill from the left in May were the reason the majority’s margin was so thin in July. Had either of them been persuadable on the conference report, Brown’s leverage over the pay-for would have been reduced, because the leadership would have had a vote to spare. Their opposition, or in Feingold’s case his continued opposition, is part of what made the three Republicans indispensable. The left flank’s dissent empowered the center-right’s demands. That is the sixty-vote Senate in operation: every senator’s position affects every other senator’s leverage, and the final text reflects the equilibrium of all of them.

The Four-Vote Statute

The passage record invites a structural conclusion, and the record is detailed enough to state it precisely. The four-vote statute: because sixty votes were required and the majority did not have them by itself at the decisive moments, the final content of Dodd-Frank was set by a very small number of senators whose individual objections could reopen a completed conference, which is a structural feature of modern lawmaking rather than an accident of 2010.

The claim is deliberately narrow, and each of its elements is documented above. Sixty votes were required: the April cloture failures at 57-41, 57-41, and 56-42, the May 19 failure at 57-42, and the May 20 and July 15 successes at exactly sixty establish the threshold as the binding constraint of the entire Senate phase. The majority did not have sixty by itself: the April and May 19 shortfalls, the two Democratic no votes on the May 20 cloture motion, and Feingold’s final no vote show a majority that needed help at every turn. A very small number of senators set the final content: Brown’s switch supplied the May 20 cloture margin, Brown’s letter forced the June 29 reopening, and the Collins and Snowe votes were load-bearing on both the May and July tallies. Individual objections could reopen a completed conference: the June 29 refiling after the June 25 completion is the proof, with H. Rept. 111-517 filed at 8:30 that evening carrying a different pay-for than the agreement the conferees had approved four days earlier.

The counter-reading deserves a fair hearing before it is set aside. The House vote of December 11, 2009, with zero Republican ayes, supports the view that one party wrote the law, and nothing in the Senate history erases that tally. But a statute is not written by one chamber, and the Senate record shows the minority extracting specific changes at each stage: the Durbin amendment’s 64-33 margin included seventeen Republican votes, the Section 716 softening was negotiated to hold the conference coalition together, and the entire pay-for was rewritten at Brown’s demand. A bill written entirely by one party does not get its funding title replaced at the insistence of a single minority senator after the conference has finished. The House record and the Senate record are both real, and the accurate summary is that the House wrote a partisan bill and the Senate’s sixty-vote rule forced it to buy the votes that made it law.

This dynamic was not unique to financial reform in that Congress. The health care overhaul enacted the same year traveled its own sixty-vote path, and readers interested in the parallel can consult the companion passage history of the Affordable Care Act, which shows the same structural logic operating on different substance. The comparison is instructive precisely because the two bills differed so much in content while sharing the same procedural constraint. In both cases, the need for sixty shaped the text more than any single policy argument did, and in both cases the senators who held the sixtieth vote held disproportionate influence over the outcome.

None of this is a criticism of the participants. Brown used his position to remove a tax he opposed; Lincoln used hers to keep the derivatives title she had built in committee; Durbin used his to attach a provision with broad bipartisan appeal; Feingold used his vote to register the judgment that the bill was too weak. Each acted within the rules, and the rules gave pivotal voters exactly the influence the rules were designed to give them. The structural feature is the filibuster itself, or more precisely the sixty-vote cloture threshold as it operated in 2010, and the accident of 2010 was only which senators happened to hold the pivotal seats. The lesson for readers of legislative history is general: when the threshold is sixty and the majority holds fewer, the text will be set at the margin by the voters the majority must attract, and the provisions those voters demand will be visible in the statute for anyone who knows where the votes were.

The name four-vote statute refers to the margin by which the majority fell short, and then to the margin by which it succeeded. In April, the shortfalls were three, three, and four votes. On May 19, the shortfall was three. On May 20 and July 15, the majority reached sixty exactly, with no vote to spare. The statute was therefore written, at every decisive moment, within a band of about four votes, and the senators inside that band, Brown, Specter on the day of his return, Collins, Snowe, and on the other side Cantwell and Feingold, were the effective authors of the final text to a degree far beyond their numbers. Four votes is not a metaphor; it is the measured distance between the majority’s strength and the threshold the rules imposed.

The provisions those votes bought are itemized in the ledger, but the list is worth restating as a single argument. Brown’s May 20 yes bought the substitute’s survival, and his June 29 letter bought the pay-for’s replacement. The Collins and Snowe votes, load-bearing in May and July, bought whatever concessions the leadership made to keep them, concessions the record reflects in the conference’s softening of contested provisions. The Durbin coalition’s seventeen Republican votes bought the interchange provision’s comfortable margin and its acceptance in conference. Feingold’s and Cantwell’s no votes bought nothing, which is itself a finding: in a sixty-vote chamber, opposition from within the majority is absorbed by votes purchased from the minority, and the purists of the majority party end up with less influence than the marginal voters of the minority. That is the structural irony the article’s central claim describes, and it is visible in every roll call.

The counter-reading, that one party wrote the law, is best engaged on its strongest ground, which is the House record. The December 11 tally, with zero Republican votes, is a fact, and it supports the inference that the House majority neither needed nor sought minority input. But the inference stops at the chamber door. A statute requires the Senate’s assent, and the Senate’s assent required minority votes, and those votes were purchased with substantive changes. The accurate synthesis is sequential: the House wrote a partisan bill, and the Senate’s rules forced the majority to pay for its passage with provisions the House had not chosen. Neither half of that sentence is dispensable, and histories that quote only one half are incomplete.

The parallel with the health care overhaul of the same Congress, whose passage history is told in the companion account of the Affordable Care Act’s passage, reinforces the structural reading. Two bills of very different substance, moving through the same Senate under the same sixty-vote rule, both had their contents set at the margin by the senators whose votes were needed. The recurrence of the pattern across different policies suggests that the pattern belongs to the rules rather than to the issues. Change the threshold, and the authorship changes; keep the threshold, and the pivotal voters will always be the effective drafters of the last compromises. That is why the article presents the four-vote dynamic as a structural feature of modern lawmaking rather than an accident of 2010: the names on the pivotal votes change from Congress to Congress, but the logic of the threshold does not.

A final word on neutrality is in order, because the article’s thesis could be mistaken for a partisan argument. It is not. The account reports both parties’ votes at every stage, attributes every characterization to its source, and treats the participants’ strategies as rational responses to the rules rather than as virtues or vices. Brown’s position, Lincoln’s committee work, Durbin’s coalition-building, Feingold’s dissent, and Grassley’s conditional support are all described as what they were: moves within a procedural game whose rules were fixed. Readers of any political persuasion can accept the structural conclusion without endorsing any participant’s goals, because the conclusion is about arithmetic, not about ideology. Sixty was required, sixty was barely assembled, and the text shows the fingerprints of the assembly.

What Each Pivotal Vote Purchased

The series thesis for this article is that procedure determined text, and Dodd-Frank offers an unusually well-documented record of which provisions each pivotal vote bought. It is worth making that mapping explicit, because it is the feature of this legislative history that separates it from most others. In the typical major enactment, the bargaining that sets the final language happens in private, and historians reconstruct it from memoirs and interviews. Here the bargaining happened on the Senate floor, in televised conference sessions, and in roll calls with published margins, and the provisions can be matched to the votes that secured them. What follows is that matching, vote by vote, in the order the pivotal moments occurred.

The April 28 unanimous-consent agreement purchased the debate itself. Without it, after three failed cloture votes, the bill would have remained stuck at the threshold, and the majority would have faced the choice of abandoning the effort or spending weeks more on a fourth attempt. The minority’s agreement to proceed, given without a recorded vote, is the least visible pivotal moment in the history and one of the most important. Everything that followed, the amendments, the substitute, the conference, depended on the Senate getting to the bill at all.

The May 20 cloture vote on SA 3739, 60-40, purchased the substitute and everything inside it. Brown’s switch and Specter’s presence were the two votes that moved the tally from 57-42 to 60-40, and in exchange the bill’s managers accepted a substitute whose derivatives title they had not written and whose fee provisions they had not sought. The Durbin amendment, adopted 64-33 a week earlier, had already been folded into the package the substitute carried. The price of the sixtieth vote was therefore not a single concession but the entire Senate product: the substitute as a whole, with Lincoln’s title, Durbin’s fees, and the rejected Brown-Kaufman alternative all part of the record the conference would inherit.

The May 6 defeat of Brown-Kaufman, 61-33, purchased the statute’s shape by negation. The Senate’s refusal to cap bank size defined the law as a regulatory statute rather than a structural one, and every provision that survived must be read against the more aggressive alternative the chamber turned down. The thirty-three senators who voted yes bought nothing in the text, but they bought a marker: the vote stands as the Senate’s considered judgment that the crisis was a failure of practices rather than of scale, and that judgment constrained everything the conference could do.

The June 29 conference revision purchased the final sixty votes at the price of the pay-for. Brown’s letter exchanged his support, and by extension the support of Snowe and Collins, for the removal of the nineteen billion dollar assessment and its replacement with the early end of TARP and higher FDIC premiums. The conference’s first-round decisions, the televised softening of the push-out, the acceptance of Durbin, the abandonment of the pre-funded resolution fund, had been purchased with the expectation that the original pay-for would hold. When it did not, the funding was rewritten in a day, and the bill that went to the floors was different from the one the conference had completed on June 25.

The July 15 final votes purchased enactment at the price of Feingold. The majority accepted the loss of the one Democrat who judged the bill too weak in order to hold the three Republicans whose votes made sixty. That exchange, one Democrat lost and three Republicans gained, is the arithmetic of the sixty-vote Senate rendered as text: the bill had to be strong enough to keep fifty-seven Democrats and moderate enough to attract three Republicans, and the point where those two constraints met was the statute.

Why did Senator Grassley support the Senate bill in May but oppose the conference report in July?

He backed the process, then rejected its product. Grassley voted for the Senate bill in May and helped the Agriculture Committee approve the Lincoln derivatives title 13-8 in April. But he opposed the July 15 conference report over its derivatives language and offsets. The conference’s compromises cost the report a supporter it had held in the spring.

The Record Versus the Memory

Legislative histories decay in predictable ways, and Dodd-Frank’s has decayed in four specific ways that the verified record corrects. Each correction matters because each one, left uncorrected, distorts the procedure-determines-text thesis this article advances.

The first decay concerns the May cloture votes. Summaries sometimes report that the Senate passed the bill on May 19, 2010 by 60-39, or that cloture succeeded on that date. The record shows the opposite: on May 19, cloture on the Dodd-Lincoln substitute failed 57-42, and the successful cloture vote came on May 20, 60-40, on reconsideration. The confusion is understandable, since the two days blur together in memory, but the distinction is substantive. The May 19 failure is what made Brown’s switch and Specter’s return the decisive events of May 20. A history that merges the two days into a single successful vote erases the very mechanism, the overnight three-vote swing, that demonstrates the pivotal voters’ power.

The second decay concerns the House conference report’s Republican supporters. Casual accounts sometimes name Representative Judy Biggert of Illinois among the Republican ayes. The House Clerk’s individual vote record for Roll 413 shows she voted no. The three Republican ayes were Cao of Louisiana, Castle of Delaware, and Walter Jones of North Carolina. Getting the names right is not pedantry. The point of naming them is to show that the revised bill drew minority support the original had lacked, and misattributing that support muddies the evidence.

The third decay merges the Volcker Rule with the Lincoln push-out into a single provision, sometimes described as a “Volcker-Lincoln” section 716. They are separate. Section 716 is the Lincoln swaps push-out, titled “Prohibition against Federal Government bailouts of swaps entities.” Section 619 is the Volcker Rule, restricting proprietary trading and certain fund investments. The conflation obscures the distinct legislative paths the two provisions traveled: the push-out came through the Agriculture Committee’s derivatives title without a standalone floor vote, while the Volcker Rule came through the Banking Committee’s work and the substitute. Collapsing them into one erases the Agriculture Committee’s authorship and the conference’s separate treatment of each.

The fourth decay misdescribes the nineteen billion dollar assessment as funding for a pre-funded orderly liquidation fund. No such fund survived in the final bill. The House’s and Senate’s pre-funded resolution-fund concepts were dropped in the conference’s first round, in favor of post-paid assessments. The nineteen billion dollar figure was the general pay-for, sized to offset the bill’s scored cost to the budget. The distinction matters because Brown’s objection is otherwise misread as a fight about resolution policy. It was a fight about who paid for the legislation, and the conference’s response, ending TARP early and raising FDIC premiums, was a different answer to the same budgetary question.

These four corrections share a moral. The statute’s history is unusually well documented, which means its misrememberings are checkable rather than merely arguable. Every claim in this article traces to a roll call, a committee vote, a filed report, or contemporary press quoting the participants. That density of documentation is what makes the procedure-determines-text thesis more than an interpretation. It is a reading of a record that anyone can verify.

The Two Citations: H.R. 4173 and Public Law 111-203

Every federal statute carries two identities, and Dodd-Frank’s pair tells the story of its passage in miniature. As a bill, it was H.R. 4173 in the 111th Congress, introduced by Representative Frank on December 2, 2009. The “H.R.” prefix marks it as a House bill, and the number marks its place in the sequence of measures introduced in that chamber. As a law, it is Public Law 111-203, signed by President Obama on July 21, 2010, and cited in the Statutes at Large as 124 Stat. 1376. The “111” marks the Congress, the “203” marks its place in the sequence of public laws that Congress enacted, and the Statutes at Large citation gives its permanent location in the official compilation of federal law.

The distance between those two citations is the distance this article has traced. H.R. 4173 began as Frank’s nine-day House sprint, a Democratic bill passed without a single Republican vote. Public Law 111-203 ended as a Senate-negotiated statute, passed with three Republican votes in each chamber and opposed by one Democrat. The bill number survived the journey. The text did not. When researchers cite 124 Stat. 1376, they are citing the product of the conference committee’s two reports, the Senate’s substitute, the floor amendments, and the pivotal votes that priced each of them, not the bill the House passed in December.

The Statutes at Large citation also records the statute’s scale. Beginning at page 1376 of volume 124 and running to page 2223, the law is among the longest financial enactments in American history. Its length is a measure of its ambition and of the legislative process that produced it: every title represents a negotiation, every section a compromise, every compromise a vote. The citations are dry, but they encode the history. H.R. 4173 is where the story started. Public Law 111-203 is where the votes brought it.

Sponsors and Two Hundred Thirty-One Days

The statute bears the names of Senator Christopher Dodd of Connecticut and Representative Barney Frank of Massachusetts. Dodd sponsored S. 3217, introduced on April 15, 2010, and chaired the Senate Banking Committee through the bill’s Senate phase. Frank sponsored H.R. 4173, introduced on December 2, 2009, and chaired the House Financial Services Committee. The naming convention follows the standard practice of crediting the lead sponsors in each chamber, and the two chairmen shepherded the bill through every stage described here, from the House markup through the televised conference to the signing ceremony. For the full substance of what their bill contained, the companion complete guide to the act takes up where this passage history ends.

The chairmanships were not ceremonial details. Dodd’s Banking Committee produced S. 3217 and managed the Senate floor; Frank’s Financial Services Committee produced H.R. 4173 and managed the House floor. The committee jurisdictions explain why the bill’s two halves look different: the Banking Committee’s product reflected Senate priorities and Senate compromises, including the derivatives title that arrived via Agriculture, while the Financial Services product reflected the House majority’s preferences. When the conference reconciled the two, it was reconciling the work of two chairmen with different chambers, different rules, and different coalitions behind them. The statute’s double name is therefore more than a courtesy; it records the dual parentage that the passage history documents.

The Office of Management and Budget issued a Statement of Administration Policy on December 8, 2009, on H.R. 4173, styled “The Wall Street Reform and Consumer Protection Act of 2009 (Rep. Frank, D-Massachusetts),” putting the administration on record on the House bill three days before passage. In the Senate, the Banking Committee’s chairman, Dodd, sponsored S. 3217 and managed the floor, while the Agriculture Committee’s chairman, Lincoln, supplied the derivatives title that became the floor’s most contested provisions. In the House, the Financial Services Committee’s chairman, Frank, sponsored H.R. 4173 and managed its nine-day sprint to passage.

The signature on July 21, 2010, closed a span of two hundred thirty-one days that had begun with Frank’s afternoon introduction on December 2, 2009. Between those two timestamps lay every procedural regime the article has described: House majority rule, Senate cloture fights, amendment combat under a substitute, public conference bargaining, a reopened conference, and final passage on unamendable reports. The two chairmen whose names the statute bears were present for all of it, and the roll-call numbers they left behind are the reason the passage can be reconstructed vote by vote rather than remembered as a blur.

The pace also reflects the majority’s management of risk. The nine-day House sprint minimized the time opponents had to organize. The deliberate Senate buildup maximized the time the leadership had to count votes, and the record shows the counting was still incomplete on May 19, when cloture failed 57-42. The two-week televised conference maximized public legitimacy for the compromises it produced. And the one-day June 29 revision minimized the time anyone had to unravel the deal before the House voted on June 30. Each phase’s tempo served its purpose, and the purposes were all about votes: getting them, keeping them, and spending them before they expired.

The calendar of the bill’s journey runs from December 2, 2009, to July 21, 2010: two hundred thirty-one days, about seven and a half months, roughly thirty-three weeks. The stages fit inside that span with room to spare. December belonged to the House. January was quiet, with the Senate receiving the bill on the twentieth. April brought the three failed cloture votes and the unanimous-consent breakthrough. May brought the amendment fights and the 60-40 cloture victory. June brought the public conference, its completion, its reopening, and the revised report. July brought final passage and the President’s signature. A bill that began as a House chairman’s draft in early December became Public Law 111-203 in late July, and at every stage the votes that moved it forward were counted, recorded, and preserved.

That preservation is what makes this passage history possible. The roll calls are numbered and public, the conference sessions were televised, the objection letter was a matter of record, and the refiling is timestamped to the half hour. Few major statutes of the modern era left so complete a trail from introduction to signing, and the trail shows, step by step, how procedure determined text. The House passed a partisan bill. The Senate could not move it without minority votes. The minority’s price was written into the amendments, the softened provisions, and the replaced pay-for. The conference finished, reopened, and finished again. And a Democratic senator voted no because the result was too weak, which is perhaps the most telling detail of all: in a process governed by the search for sixty, the final text disappointed the purists of both parties, and it passed anyway.

The chairmen’s roles extended beyond sponsorship into the management of each stage. Dodd, chairing Banking, was the Senate’s floor manager for the bill and the co-namesake of the substitute that carried it. Frank, chairing Financial Services, managed the House’s two floor actions and co-chaired the conference. The naming of the statute for the two chairmen follows the congressional custom of crediting the members who carried the legislation, and in this case the custom fits the facts: no other members were as continuously involved from introduction to signing.

The two hundred thirty-one days can be summarized as a calendar of thresholds. December: the House, needing a simple majority, passed the bill without minority help. January: the bill waited while the Senate organized itself. April: the Senate, needing sixty, failed three times to begin debate and then proceeded by consent. May: the Senate amended the bill, defeated its most restrictive alternative, and passed it with sixty votes assembled from both parties. June: the conference reconciled the texts in public, finished, reopened at a single senator’s demand, and filed a revised report. July: both chambers passed the revised report, and the President signed it. Each month’s entry is a roll call, each roll call is numbered, and the numbers are the article’s evidence for its claim. Procedure determined text, the pivotal votes are documented, and the documentation is complete.

Frequently Asked Questions

Q: How did Dodd-Frank pass Congress?

Dodd-Frank passed in five stages. The House passed H.R. 4173 on December 11, 2009, by 223-202 with no Republican votes. The Senate received the bill in January 2010; after three failed cloture votes on the motion to proceed in April, the chamber proceeded by unanimous consent on April 28. In May the Senate adopted floor amendments, defeated the Brown-Kaufman size cap 61-33, invoked cloture 60-40 on the Dodd-Lincoln substitute on May 20, and passed the bill 59-39 the same day. A public conference committee reconciled the two versions in June, completed its work June 25, reopened June 29 at Senator Scott Brown’s demand, and filed a revised report. The House agreed to the conference report June 30 by 237-192, the Senate agreed July 15 by 60-39, and President Obama signed it July 21, 2010, as Public Law 111-203.

Q: Did any Republicans vote for Dodd-Frank?

In the House vote of December 11, 2009, no Republican voted for the bill; the 223-202 tally included zero minority votes. Minority support appeared in the Senate, where the sixty-vote cloture requirement made it necessary. The Durbin debit-interchange amendment passed 64-33 on May 13, 2010, with seventeen Republican votes. Cloture on the Dodd-Lincoln substitute succeeded 60-40 on May 20 with Senator Scott Brown’s yes vote. On the conference report, three House Republicans voted yes on June 30, 2010: Joseph Cao of Louisiana, Mike Castle of Delaware, and Walter Jones of North Carolina. In the Senate on July 15, 2010, three Republicans voted yes: Scott Brown of Massachusetts, Susan Collins of Maine, and Olympia Snowe of Maine. Republicans thus voted for the bill at nearly every Senate stage, though not a single House Republican supported the original December passage.

Q: Why did Russ Feingold vote against Dodd-Frank?

Senator Russ Feingold of Wisconsin, a Democrat, voted against the bill because he considered it too weak, not too strong. He voted no on the Senate bill on May 20, 2010, one of two Democrats to do so, and voted no again on the conference report on July 15, 2010, as the only Democratic senator in opposition. His stated reason was that the measure was not restrictive enough on the banking industry. In his floor statement he said that after thirty years of deferring to Wall Street lobbyists, Congress needed tough reforms, but the key reforms were missing from the bill. He framed the test plainly: whether the legislation would prevent another financial crisis, and as the bill stood, he said, it fails that test. The earlier defeat of the Brown-Kaufman size-cap amendment had already signaled that the Senate would not impose the toughest available constraints.

Q: What happened when the chambers reconciled the Dodd-Frank text?

A conference committee met in public, televised sessions on June 10, 15, 16, 17, 22, 23, and 24, 2010, and completed its work in the early morning of June 25, approving the agreement 20-11 among House conferees and 7-5 among Senate conferees. The conferees kept the Lincoln swaps push-out as Section 716 but softened it with a July 2012 effective date, a two-year transition, and exemptions for hedging and bank-eligible swaps. They dropped the pre-funded resolution-fund concepts both chambers had considered in favor of post-paid assessments. They included a nineteen billion dollar assessment on large firms as the bill’s pay-for. After Senator Scott Brown objected to the assessment, the conference reopened on June 29 and replaced it with early termination of TARP authority plus higher FDIC assessments, filing the revised report that evening.

Q: Which Dodd-Frank amendments failed on the Senate floor?

The most prominent failed amendment was SA 3789, the Brown-Kaufman size cap offered by Senators Sherrod Brown of Ohio and Ted Kaufman of Delaware, which would have limited bank liabilities and deposits as a share of the economy. The Senate defeated it 61-33 on May 6, 2010, with three Republicans joining twenty-nine Democrats and Senator Sanders in support; it was later withdrawn during the May 20 endgame. The amendment was described at the time as the more restrictive alternative to the bill’s approach of regulating large banks rather than shrinking them by statute. Its defeat showed where the chamber’s limits lay: a decisive bipartisan majority rejected a hard size cap two weeks before passing the bill itself. The sixty-vote requirement bounded the Senate’s choices, and within those bounds the chamber chose tighter regulation over structural downsizing.

Q: What is the Durbin amendment in Dodd-Frank?

The Durbin amendment, SA 3989, was offered by Senator Richard Durbin of Illinois and adopted by the Senate 64-33 on May 13, 2010. It directed the Federal Reserve to set debit-card interchange fees, the fees merchants pay when customers use debit cards, at levels deemed reasonable and proportional, and it exempted card issuers with less than ten billion dollars in assets. The winning coalition was bipartisan: seventeen Republicans and forty-seven Democrats voted yes, making it one of the most cross-party votes of the entire debate. The amendment was accepted in conference and appears in the enacted statute. It is the clearest counterexample to the claim that the minority opposed every part of the bill; on this provision, minority senators supplied a large share of the winning margin.

Q: How long did Dodd-Frank take to pass?

Two hundred thirty-one days passed between introduction and signing: H.R. 4173 was introduced on December 2, 2009, and President Obama signed it on July 21, 2010, about seven and a half months. The House passed it nine days after introduction, on December 11, 2009. The Senate phase stretched across the spring: three failed cloture votes in late April, floor amendments in early and mid-May, cloture and passage on May 20. The conference committee worked through most of June, finishing June 25, reopening June 29, and filing its revised report that evening. Final passage came June 30 in the House and July 15 in the Senate. The pace reflected the Senate’s sixty-vote requirement, which forced weeks of negotiation that a simple-majority chamber would not have needed.

Q: Who were the sponsors Dodd-Frank is named after?

The statute is named for Senator Christopher Dodd of Connecticut and Representative Barney Frank of Massachusetts, the lead sponsors in their respective chambers. Dodd sponsored S. 3217, introduced April 15, 2010, and chaired the Senate Banking Committee during the bill’s Senate phase. Frank sponsored H.R. 4173, introduced December 2, 2009, and chaired the House Financial Services Committee. The naming follows the standard congressional practice of crediting the principal sponsors in each chamber. Both chairmen shepherded the measure through every stage: House markup and passage, the Senate floor fight, the televised conference committee, and the revised conference report that became Public Law 111-203 when President Obama signed it on July 21, 2010.

Q: What was S. 3217 and how did it relate to H.R. 4173?

S. 3217 was Senator Christopher Dodd’s financial reform bill, introduced in the Senate on April 15, 2010. When the Senate took up the House-passed H.R. 4173, it did not vote on the House text as written. Instead, on May 20, 2010, the Senate struck all language after the enacting clause of H.R. 4173 and substituted the text of S. 3217 as amended, then passed H.R. 4173 in lieu of S. 3217 by 59-39. The House bill number therefore survived as the legislative vehicle while the Senate’s substance replaced the House language. This is why histories sometimes describe the Senate as having passed its own bill: in form the vehicle was H.R. 4173, but in content it was the Dodd bill, including the Agriculture Committee’s derivatives title that became Section 716.

Q: What was SA 3739 and why did it matter?

SA 3739 was the Dodd-Lincoln substitute amendment, proposed April 29, 2010, in the nature of a substitute, meaning it proposed to replace the entire text of the bill. It carried the language of S. 3217, including the Senate Agriculture Committee’s derivatives title with the Lincoln swaps push-out provision. Because the substitute contained the whole Senate product, the cloture votes on it were the decisive votes of the Senate phase: cloture failed 57-42 on May 19, 2010, and succeeded 60-40 on May 20 upon reconsideration. After the successful vote, the substitute was agreed to by unanimous consent, and the Senate then passed the amended bill the same day. Without SA 3739, the derivatives title would not have reached the conference committee, and Section 716 would not be in the statute.

Q: How did the swaps push-out reach the statute without a Senate floor vote?

The swaps push-out, enacted as Section 716, originated in the Senate Agriculture Committee’s derivatives title, the Wall Street Transparency and Accountability Act of 2010, which the committee approved 13-8 on April 21, 2010, under Chairwoman Blanche Lincoln of Arkansas. That title was folded into the Dodd-Lincoln substitute amendment, SA 3739, rather than being offered as a standalone floor amendment. When the Senate invoked cloture on SA 3739 by 60-40 on May 20, 2010, and then agreed to the substitute by unanimous consent, the push-out entered the bill without ever receiving its own roll-call vote. There is no recorded Senate floor margin for the provision itself, and accounts that assign it one are mistaken. The conference committee later softened it, but the core survived into the enacted law.

Q: What is the difference between Section 716 and Section 619?

Section 716 is the Lincoln swaps push-out provision, titled “Prohibition against Federal Government bailouts of swaps entities.” It required certain swaps activities to be conducted in separately capitalized affiliates rather than inside banks with federal support, subject to the exemptions and transition periods added in conference. Section 619 is the Volcker Rule, a separate provision restricting proprietary trading and certain fund investments by banking entities. The two are sometimes blurred in shorthand summaries, but they do different things, carry different section numbers, and traveled through the legislative process as distinct provisions. Keeping them distinct matters because the conference softening described in this article, the July 2012 effective date and the two-year transition, applied to Section 716, not to Section 619.

Q: Why was the Dodd-Frank conference committee televised?

The conference reconciling the House and Senate versions held its sessions in public on June 10, 15, 16, 17, 22, 23, and 24, 2010, and contemporary accounts described two weeks of official, publicly televised negotiations. Conference committees often conduct their real bargaining privately, but in this case the public sessions were extensive enough that the bargaining left a visible record: the softening of Section 716, the abandonment of the pre-funded resolution fund, and the adoption of the nineteen billion dollar pay-for all occurred in view of the press and public. The transparency is one reason the passage record is unusually complete. When the conference finished on June 25 and then reopened on June 29 at Senator Scott Brown’s demand, both events were documented in the open.

Q: Who voted against cloture on May 20 even though they belonged to the majority?

On May 20, 2010, when the Senate invoked cloture 60-40 on the Dodd-Lincoln substitute amendment SA 3739, two Democratic senators voted no: Maria Cantwell of Washington and Russ Feingold of Wisconsin. Their opposition meant the sixty-vote threshold was reached only because minority senators voted yes, with Senator Scott Brown switching to yes and Senator Arlen Specter returning from absence to vote yes. Cantwell and Feingold also voted against final passage of the bill later that day, when it passed 59-39. Feingold would go on to cast the lone Democratic no vote on the conference report on July 15, consistent with his stated view that the bill was not restrictive enough on the banking industry.

Q: Why did Senator Byrd’s death matter to the conference outcome?

Senator Robert Byrd of West Virginia died on June 28, 2010, three days after the conference committee completed its work. His death left a Senate vacancy that reduced the majority’s margin for error to zero: with sixty votes required for cloture on the conference report, the majority could no longer afford to lose a single vote. The very next day, June 29, Senator Scott Brown of Massachusetts warned in a letter that he would oppose the bill if it retained the nineteen billion dollar bank assessment. Because the vacancy meant Brown’s vote could not be replaced, his objection was decisive, and the conferees reconvened the same day to rewrite the pay-for. Without Byrd’s death, the majority might have absorbed Brown’s defection; with the seat vacant, it could not.

Q: What happened to the pre-funded resolution fund proposals?

Both the House and the Senate considered creating a pre-funded resolution fund, financed in advance by assessments on large financial firms, that would pay for the orderly wind-down of failing institutions. The conference committee dropped both versions. In their place, the final bill provided for post-paid assessments, meaning the costs of any future resolution would be collected from the industry after the fact rather than banked in advance. The enacted statute therefore contains no pre-funded orderly liquidation fund. The nineteen billion dollar assessment that the first conference agreement carried was a general pay-for to offset the bill’s budget cost, not funding for such a fund, and even that assessment was replaced on June 29 with early termination of TARP authority and higher FDIC assessments.

Q: What was the Shelby-Dodd amendment SA 3827?

SA 3827 was an amendment offered by Senators Richard Shelby and Christopher Dodd addressing bailout-related concerns in the Senate bill, and it was adopted on May 5, 2010. It is included in this article’s amendment ledger for one reason: readers tracking amendment numbers can easily confuse it with Senator Lincoln’s derivatives provisions, and the two must be kept separate. The Lincoln swaps push-out entered the bill through the Agriculture Committee’s derivatives title folded into the Dodd-Lincoln substitute SA 3739, and no standalone floor vote was ever taken on it. SA 3827 was an entirely different measure. Keeping the numbers straight matters because the legislative history of the derivatives title is frequently misattributed to the wrong amendment.

Q: Who changed their votes between the May 19 and May 20 cloture votes?

The three-vote swing that carried cloture on the Dodd-Lincoln substitute had identifiable authors. Senator Scott Brown of Massachusetts voted no on May 19 and yes on May 20, supplying the critical switch. Senator Arlen Specter of Pennsylvania was absent on May 19 and voted yes on May 20. On the other side, Senators Maria Cantwell of Washington and Russ Feingold of Wisconsin, both Democrats, voted no on cloture on May 20. The result moved from 57-42 against invoking cloture to 60-40 in favor, exactly the threshold required. Brown’s switch foreshadowed his later role: the same senator whose vote unlocked the substitute in May would force the conference to reopen in June over the bill’s funding.

Q: Which House Republicans voted for the conference report?

Three House Republicans voted yes on the conference report on June 30, 2010: Joseph Cao of Louisiana, Mike Castle of Delaware, and Walter Jones of North Carolina. The overall tally was 237-192 on Roll 413, with 234 Democrats voting yes, 19 Democrats voting no, 173 Republicans voting no, and 4 members not voting. The three Republican ayes are worth naming precisely because later accounts sometimes misidentify them: Representative Judy Biggert of Illinois is occasionally listed as a supporter, but the House Clerk’s individual vote record shows she voted no. The contrast with December 11, 2009, when zero Republicans supported the original House bill, reflects how much the Senate process and the conference had changed the legislation.

Q: What did the June 29, 2010 conference revision change about the bill’s funding?

The revision replaced the conference agreement’s nineteen billion dollar assessment on large financial firms, which had served as the general pay-for offsetting the bill’s scored cost. After Senator Scott Brown’s June 29 letter stating he would not support a final bill containing those higher taxes, and with Senator Byrd’s death leaving the majority unable to spare a vote, conferees reconvened and dropped the assessment. They replaced it with two measures: ending the Troubled Asset Relief Program’s spending authority early, later that month instead of in October, yielding about eleven billion dollars, and increasing the deposit-insurance assessments that banks paid to the Federal Deposit Insurance Corporation to cover the remainder. The revised conference report, H. Rept. 111-517, was filed that evening at 8:30.