Early Implementation: 1933–1956

Soon after Glass–Steagall's passage in 1933, various parties sought to find weaknesses in its “wall of separation” between commercial and investment banking. During the law's first two decades, judges, bureaucrats, and legislators all sought to define the act's practical applications in terms that prevented easy evasion.Footnote ²⁶ Divergences in official institutional perspectives in this period were small.

One early challenge to the law's efficacy was rooted in its central language, which prohibited anyone “primarily engaged” in securities activities from also engaging in commercial banking. In Board of Governors v. Agnew (1946), the Supreme Court was asked to review the Federal Reserve Board's decision to remove as directors of a national bank several men who were indisputably involved in securities activities.Footnote ²⁷ First, the Court had to decide whether Federal Reserve decisions made under Glass–Steagall were subject to judicial review, which the Board of Governors disputed. For a unanimous Court, Justice Douglas decided this momentous question in the affirmative without offering much justification, ensuring that bureaucratic decisions about Glass–Steagall would receive judicial scrutiny.Footnote ²⁸ Second, the Court had to decide whether “primarily engaged” would be given a broad or narrow meaning. As the Fed read it, “primarily” meant “substantial,” whereas the removed directors argued it should mean “principal,” i.e., “greater than fifty percent.”Footnote ²⁹ The Court ruled that the Board's broad reading was both reasonable and fully consonant with the purposes of the statute, and therefore due deference.Footnote ³⁰ In doing so, they allowed regulators to arrest an early trend toward drift.

A challenging development in the early years of Glass–Steagall was the migration of many bankers to the bank holding company (BHC) corporate form, in large part because it was not covered by the main provisions of Glass–Steagall or, indeed, by much of any regulatory law.Footnote ³¹ By bringing bank ownership under the same corporate umbrella as companies pursuing non-banking activities, a BHC could effectively join control of banking and non-banking activities, thus undermining Glass–Steagall's purpose. Congress was well aware that this shift had the potential to undermine the separation between commercial and investment banking, and bills addressing it were introduced into almost every session of Congress after 1938, when President Roosevelt specifically called for regulation of BHCs. Major efforts in 1938 (Glass-McAdoo), 1947 (offered by Fed Chairman Marriner Eccles), and 1949–50 (Maybank-Robertson) all died in committee, opposed by the American Bankers Association (ABA) and its sympathetic regulator, the OCC. By the early 1950s, the Federal Reserve was sounding a consistent refrain to pass strong regulation of all BHCs, essentially sounding the alarm about policy drift.Footnote ³² A critical source of disagreement that doomed these bills in committee was the question of whether regulation should extend to BHCs controlling just one bank.Footnote ³³ With President Eisenhower's support, a bill finally emerged in 1956 and Congress passed the BHC Act by an overwhelming margin, establishing Federal Reserve regulation and prohibiting BHCs from managing or controlling non-banking assets other than those directly related to their core banking business. This only partially addressed the BHC threat, however: in order to win easy passage, the act included a slew of exemptions, most crucially excluding BHCs controlling only one bank, a source of disappointment for the President and the Federal Reserve.Footnote ³⁴ At the time of passage, only 46 BHCs were subjected to regulation, while 116 (mostly very small) BHCs were exempted.Footnote ³⁵ Consequential as it was, BHC Act supporters secured a consensus only by retaining many exemptions, demonstrating how legislation intended to promote the general good is often greased with side payments to affected interests in order to secure passage, thus undermining the very rationale that motivated the reform.Footnote ³⁶ The price of this consensus was relative weakness, and by 1958 the Fed would be lobbying Congress again to close the Act's loopholes, especially the one-bank exemption.Footnote ³⁷

The language of the Bank Holding Company Act also left the door open to reinterpretation when various parties would later seek to reorient the broad Glass–Steagall institution. Key language allowed banks to conduct those activities “so closely related to banking as to be a proper incident thereto” without further defining just what these activities might be.Footnote ³⁸ Regulators would be charged with making initial determinations about which activities were permissible, giving them significant power to direct the evolution of the law in practice. Of course, their decisions could also then be contested in court by parties who could articulate why their interests were impaired. This open-textured language thus set the stage for a long struggle over the direction of Glass–Steagall's institutional development.

Fig. 1.

Bank Holding Company Corporate Organization (Each Unit's Regulator Noted in Parentheses)

Reproduced from General Accounting Office, GAO/T-GGD-95-121, “Modernization of the Financial Services Regulatory System. Statement of James L. Bothwell, Director, Financial Institutions and Market Issues” (1995): 11.

Protecting Separation under the BHC Act, 1956–1971

The next period of Glass–Steagall (and now BHC Act) development was characterized by the same dynamic of seeking to find ways to adjust to innovations at the least possible political cost, with only small divergences in institutional perspectives opening up. Bureaucrats and legislators remained basically committed to the original aims of Glass–Steagall, but they nevertheless sought to accommodate various constituents as they reacted to new developments. Where courts saw this accommodation as going against their conception of Glass–Steagall's institutional purpose, they intervened to protect it.

The major challenges during this time were dictated by the compromise of the BHC Act, which had left many loopholes intentionally and also contained enough ambiguity to allow others to form. For example, in 1957, commercial banks sought the ability to profit from conducting brokerage transactions for their customers. Such off-balance-sheet activities could potentially enhance a bank's profitability and, if kept small, would not pose a major threat to Glass–Steagall's fundamental differentiation between commercial and investment banks. The OCC allowed the shift, although the regulator still insisted that such services be offered only to existing customers rather than marketed to the public in order to ensure that they remain modest in scope.Footnote ³⁹ Such incremental decisions may seem inconsequential, but they may accumulate until they represent a substantial institutional change. Although the OCC would not provide a comprehensive rationale for such actions until much later, this early concession to banks foreshadowed the formation of a distinct institutional perspective for commercial banks' regulators.

Courts sometimes arrested these small steps toward expanded powers for commercial banks, showing their willingness to stand up for the institutional conception of Glass–Steagall as a strict wall of separation. In Baker, Watts & Co. v. Saxon (D.D.C. 1966), a federal district judge invalidated the Comptroller's decision to allow a bank to underwrite and trade state bonds not backed by the taxing power.Footnote ⁴⁰ Glass–Steagall itself contained an exception for “obligations of the United States, or general obligations of any State or of any political subdivision thereof,” but the court interpreted “general obligations” narrowly, to include only those debts directly backed by the state's power of taxation, thus preventing this intentionally included loophole from growing. On one level, this case can be seen as a narrow matter of statutory interpretation, but on another, it represents an early clash between the major ways of thinking with respect to Glass–Steagall's purposes.

In 1965, aiming to strengthen the institution of Glass–Steagall-plus-BHC-as-strict-wall-of-separation, Congress took aim at many of the loopholes left in place by the 1956 Act. House Banking Committee Chairman Wright Patman (D-TX) first introduced a narrow bill repealing one particular exemption for testamentary trusts, which had allowed the blossoming of a miniature empire of unregulated firms in Florida under the auspices of the estate of Alfred du Pont. But Rep. Charles E. Bennett (D-FL), encouraged by the Fed, offered a comprehensive amendment to level the playing field, which would have repealed all of the most important exemptions, including for one-bank BHCs. This broader bill passed the House in spite of opposition by Patman, who argued that every exemption deserved individual consideration before being repealed.Footnote ⁴¹ Although the Fed was adamantly in favor of ending the one-bank exemption, doing so was strongly opposed by the Independent Bankers Association of America, the ABA, and the OCC (which said “imaginary possibilities of abuse” were no reason for action and worried about Fed jurisdictional imperialism). To secure passage, the Senate Banking Committee dropped the one-bank exemption before sending the bill to the floor.Footnote ⁴² The House assented to the Senate amendments by voice vote, and the bill was signed into law by a fairly uninterested President Johnson on July 1, 1966, having closed many important loopholes but leaving the most important one untouched. Most legislators were apparently still committed to an institutional vision of Glass–Steagall as the separation of commercial and investment banking, making them willing to support loophole-closing upkeep. Their commitment was not so strong, however, as to make them willing to spend much political capital on the matter, and so concessions to vocal interests limited the scope of their responses.

Legislators' political calculations seem to have been fairly short-sighted, as rapidly growing numbers of BHCs took advantage of the one-bank loophole to escape Fed supervision.Footnote ⁴³ There were 550 one-bank BHCs in 1965, 691 by year-end 1968, and 810 by February 1969, controlling 40 percent of all commercial bank deposits.Footnote ⁴⁴ President Nixon urged Congress to stop this “disturbing trend in the past year toward erosion of the traditional separation of powers between [banks and commerce].” Chairman Patman now abandoned his defense of the one-bank loophole, and indeed now derided the Administration for being too lax. After struggling over who should have regulatory jurisdiction and the power to define the scope of permitted activities (and therefore who would be the first responder to future industry attempts to circumvent or redefine the law), Patman's Fed-favoring bill ultimately won out in the House. The Senate passed an amended version of the bill, but not before adding several exemptions for very small BHCs and those in which banking was only a small part of the company's overall operations, as well as clarifying that “traditional banking practices” were not meant to be targeted.Footnote ⁴⁵ The ABA intensely lobbied House conferees to accept these bank-friendly Senate amendments and mostly succeeded. Once again, loopholes that served well-positioned interests were included as long as they seemed not to threaten the overall efficacy of the law.

Importantly, there were some unresolved disputes between the House and Senate on how the Act's language would operate. Although Congress had considered including a “laundry list” of prohibited activities, in the end they failed to agree on one and instead legislated the rather vague standard that any allowed non-bank activities must result in public benefits.Footnote ⁴⁶ In practice, this left the Fed to exercise its own discretion to determine what “public benefits” would mean, as courts would be unable to discern a clear congressional intent capable of trumping bureaucratic interpretations embedded in a reasonable framing. Congress clearly has the formal power to explicitly dictate the course of a law's institutional transformation, but doing so may require a degree of political consensus and detailed policy knowledge that seldom exists. Given the political costs of resisting drift, the responsibility for doing so is likely to be shunted onto regulators—with the added political bonus of plausible blame-avoidance if anything should go so wrong as to attract critical public attention.Footnote ⁴⁷

The next major developments for Glass–Steagall and the BHC Act came in the federal courts. The judiciary's centrality was guaranteed by the unanimous holding in Association of Data Processing Service Organizations v. Camp (1970), which established the precedent that aggrieved competitors were entitled to standing to challenge regulatory decisions in spite of the absence of any explicit statutory provision providing it.Footnote ⁴⁸ Thenceforth, the chief opponents of new powers for commercial banks—their potential competitors, especially among non-deposit-taking investment banks—would be able to bring legal challenges. As Macey emphasizes, securities firms were naturally the most dedicated defenders of understanding Glass–Steagall as a strict wall of separation because of the protection from competition the Act afforded them, and they would vigorously advocate this perspective in court over the coming years.

The first major case in which they did so was Investment Company Institute v. Camp (1971).Footnote ⁴⁹ The ICI, a securities industry trade group, contended that the Comptroller's decision permitting banks to offer customers “investment funds” violated Glass–Steagall's ban on commercial banks conducting securities activities. The District Court ruled against the Comptroller, but the Court of Appeals reversed, deferring to the Comptroller. Writing for a Supreme Court majority, Justice Stewart acknowledged the need for deference, but found that, whatever the Comptroller might say, the investment funds in question were so nearly indistinguishable from mutual funds as to represent violations of Glass–Steagall's clear purpose—which he accepted was well understood as the ICI framed it, as a strict separation.Footnote ⁵⁰ Deference was inappropriate when the Comptroller had offered little rationale for its holding (with appellate counsel's “post hoc rationalizations” deemed too little, too late) and when Glass–Steagall's legislative history (discussed at great length) clearly indicated that mutual fund activity was meant to be prohibited.Footnote ⁵¹ In light of these facts, Justice Stewart concluded that the Comptroller's narrow reading of the term “securities” would vitiate the enacting Congress's will and was thus impermissible.Footnote ⁵² Justice Blackmun dissented, complaining that the majority's opinion was “based more on what is … desirable national banking policy than on what is a necessary judicial construction of the Glass–Steagall Act … Policy considerations are for the Congress and not for this Court.”Footnote ⁵³ Blackmun thus found textualism inadequate to settle the ambiguities, and preferred to defer to the Comptroller's expertise as to how this particular decision would affect the functioning of the law.

ICI v. Camp vividly illustrates why disagreements about the institutional nature of Glass–Steagall had such large stakes. Whereas Stewart and the majority accepted the idea that Glass–Steagall served the public interest by strictly preventing commercial banks from offering any services that resemble investment banking, the OCC and Blackmun in dissent took a narrower view of the law that limited its powers to the bare minimum required by the text.Footnote ⁵⁴ In doing so, the OCC insisted that it was serving the public's true interests by allowing banks to enhance their profitability while taking on little risk—thereby increasing the stability of the banking system. Whether Glass–Steagall's “wall of separation” was an institution worth defending from the pressures of a changing world would be a question regulators and courts were forced to confront with increasing frequency.

Competing Institutional Visions Clash: 1970s to 1980s

The next phase of Glass–Steagall's development features a growing conflict between institutional conceptions of the law. Regulators were less and less committed to cabining commercial bank powers, instead conceiving of their primary institutional responsibility as safeguarding their own supervisory powers in the larger context of a rapidly changing financial services industry. As bureaucratic decisions butted up against the limits imposed by statutory text, Congress was too internally torn to dictate the direction of change. As a result, judges were confronted by a rising tide of security industry complaints and found that they were the ones charged with figuring out what sorts of conversion or drift were consistent with the law.

Congress had taken care in 1970 to give administrative responsibility for Glass–Steagall and the BHC Acts to the Fed rather than dividing it among the three banking agencies, apparently in hopes of keeping implementation strict. It came as a surprise to legislators, then, that the Fed from 1970 through 1974 turned out to be fairly lenient in allowing non-bank activities, leading the BHC to become far and away the dominant corporate form for commercial banking during the period. In 1972, the Board amended the regulations governing which actions were “closely related to banking,” enlarging the permissible scope of activities for BHCs.Footnote ⁵⁵ The Fed's move away from a strong “wall of separation” was somewhat equivocal; in 1974, with an economic downturn, the Fed return adopted a “go slow” policy and multiplied its denials of applications for new activities.Footnote ⁵⁶ Meanwhile, the OCC continued to facilitate a slow but steady opening of new activities to commercial banks—for example, in 1974 interpreting Glass–Steagall to let banks advertise computer-assisted stock purchasing services, although they could still only be offered to those with checking accounts at the bank.Footnote ⁵⁷

Pressures on commercial banking significantly intensified during the 1970s because of a confluence of economic conditions. Commercial banks were prohibited from offering customers interest on their demand deposits and subject to caps on interest rates for savings accounts by Regulation Q (which also came from the Banking Act of 1933). For many years, these caps were not all that onerous, as the rates they required remained competitive with other savings vehicles such as Treasury bills. As annual inflation rates rose to unprecedented heights—sometimes into the low double-digits—because of oil shocks and a hard-to-break wage-price spiral, the burden imposed by Regulation Q became much more significant, however. Customers keeping their money in demand deposits or savings accounts would essentially be forced to pay a large tax as they watched inflation outpace the interest they were offered by banks.Footnote ⁵⁸ In response to this shortcoming of commercial banks, the new money market mutual fund industry emerged to offer customers fairly liquid savings vehicles that could keep up with inflation (and eventually take advantage of the high Treasury yields caused by the Fed's attempt to crush inflation). Banks lost business to these upstarts, and at the same time saw the real value of their fixed-rate loans diminish because of inflation. Businesses also found new ways to raise capital through market offerings, especially commercial paper (short-term corporate bonds).

This phenomenon of commercial banks being cut out of the loop, called “disintermediation,” meant that the well-developed regulatory system for commercial banking covered a shrinking proportion of the nation's assets, a trend which pleased the securities industry but worried banking regulators charged with overseeing the stability of the nation's financial system.Footnote ⁵⁹ The Chairman of the FDIC from 1981 through 1985, William Isaac, would later write, “[B]y confining banks to a narrow range of products and services of declining profitability, Glass–Steagall threatens the long-term health and survival of banks as the fulcrum of our regulatory system.”Footnote ⁶⁰

Quite naturally, commercial banks' regulators were determined to resist this process. They believed it would be highly damaging to the economy if the firms they regulated were forced to wither in the face of financial innovation, and decided to convert the statutory tools at their disposal to the purpose of shoring up commercial banks' viability. Understanding their actions requires us to appreciate this broader institutional perspective, rather than simply insisting that regulators were caught up in the intellectual movement in support of deregulation.Footnote ⁶¹ Though there is little doubt that they were influenced by that movement, their motivations in changing their orientation toward Glass–Steagall were considerably more complex.

To meet the challenges of the new financial scene, the Fed hoped to expand the menu of services that BHCs could offer, and the OCC hoped to give national banks the broadest powers their statutes could support. By allowing banks access to profitable activities, they hoped to strengthen the overall safety and soundness of their regulated institutions, increasing their ability to withstand difficult economic times. This goal was reflected in regulators' joint emphasis on capital adequacy requirements, and was also made explicit in 1987 when the Fed announced it would require affiliated banks to channel revenue toward any troubled banking affiliate, thus making the expanded scope of banks' activity a source of systemic strength.Footnote ⁶² Regulators were thus largely motivated by a concern that still animates banking policy debates today: the desire to avoid taxpayer-funded bailouts of large financial institutions. Embodying this position, Comptroller of the Currency C. T. Conover delivered a speech in 1984 entitled “Strength = New Powers + Firm Supervision.”Footnote ⁶³

Another contemporaneous challenge is worth noting, although a full treatment is beyond the scope of this article. Financial markets and large corporations in the 1970s were becoming increasingly globalized. Because other important markets did not have any law equivalent to Glass–Steagall, banks that were solely commercial in their U.S. incarnation were simultaneously full-service operations in foreign markets, providing their overseas customers with a full range of financial services. For example, J. P. Morgan, one of America's leading commercial banks, included a London office reputed to be among the boldest innovators in the U.K. securities market. While this arrangement might not normally have caused any special problems for U.S. regulators, innovation in financial derivatives in the 1970s, 1980s, and 1990s gave J. P. Morgan and other firms new products that did not fit clearly into either the “commercial” or “investment” category. Once found to be profitable in London, these products could be brought back to the United States, where regulators would have little guidance for how to treat them. To the extent that Glass–Steagall was thought of as restricting commercial banks to well-understood, staid lines of business, the rapid importation of complex, hard-to-understand derivatives clearly posed a threat. On the other hand, to the extent that regulators were glad to allow novelties in commercial banking if they offered stability, derivatives seemed like an attractive addition both for the profits they offered and because of the belief that they could allow firms to manage risk more effectively than ever before.Footnote ⁶⁴

Keeping in mind the context of a dynamic marketplace exerting pressures on old ways of doing business and of regulating, those who remained committed to the institutional conception of Glass–Steagall as a strict separation could nevertheless fairly argue that the changes implemented by bank regulators' statutory interpretations in the 1980s represented drift. Those who saw Glass–Steagall's wall as self-evidently in the public interest, and moves to weaken it as harmful actions undertaken on behalf of banks, could allege corrosive capture with some plausibility. Explaining the OCC's stance, Comptroller Conover was not coy: “if a bank can make a strong case that a proposed activity is legal, our inclination is to approve it.”Footnote ⁶⁵ The OCC abandoned limitations on banks' brokerage services and opened the door to bank subsidiaries marketing insurance.Footnote ⁶⁶ Acting in parallel, the Fed allowed bank subsidiaries to market and underwrite all types of government securities, provide full brokerage services, and to trade or underwrite revenue-bonds and mortgage-backed securities.Footnote ⁶⁷ Affiliates were given full power to deal in commercial paper.Footnote ⁶⁸ Whether viewed through a lens of conversion or drift, bureaucrats had become prime movers in transforming Glass–Steagall.

There is clearly a case for understanding the regulators' attitudes as shaped by a kind of “cultural capture,” in which government officials come to strongly identify with the beliefs and interests of the firms they regulate.Footnote ⁶⁹ In seeking to enhance banks' profitability, regulators had come to adopt many of commercial banks' interests as largely coincident with their own. If one believes that their highest duty should have been realizing the Glass–Steagall Act's separation, thereby faithfully honoring the vision of the statute's framers of the 1930s, then it is fair to say that the regulators were negligent. On the other hand, if we think of these regulators as owing a deeper fidelity to protecting the institution of a functional and well-regulated banking sector through administration of the larger edifice of banking laws, their actions can make sense without any cynicism. Regulators thought they could maintain the vitality of the law's deeper spirit by allowing Glass–Steagall's requirements to bend, whereas they believed that a literalistic fidelity to the law would lead to a disastrous break in the whole system. (And, ironically, the case for remaining true to the strict-separation vision of Glass–Steagall was advanced almost exclusively by investment banks, making it hard to believe that regulators' protectiveness of commercial banking interests was self-evidently contrary to public interest.)

Legislators were best situated to negotiate the competition between institutional perspectives, since they could offer fresh statutory language settling how much continuing relevancy the half-century-old Glass–Steagall Act should retain. While some legislators showed interest in this task, especially when changes directly provoked their important constituents, no critical mass coalesced during this period to provide direction one way or the other. Instead, other banking policies took priority. The House passed a bill in 1980 to restrict BHCs from acting as insurance agents, but the Senate never responded. The two large banking bills that did pass, the Depository Institutions Deregulation and Monetary Control Act of 1980 and the Garn-St. Germain Depository Institutions Act of 1982, contained few provisions relevant to Glass–Steagall in spite of Senate Banking Committee Chairman Jake Garn's (R-UT) desire to address the issue.Footnote ⁷⁰

Signs soon emerged that Congress was considering abandoning the defense of Glass–Steagall's central tenets and turning toward large-scale “reform”—in other words, legislators began to embrace the frame of conversion favored by regulators. In 1983, Ronald Reagan proposed granting banks the ability to conduct real estate operations, but it was not an administration priority and came to naught. The House considered legislation in 1984 to close remaining loopholes in the 1956 legislation as the Senate worked at cross-purposes, passing a bill to allow BHCs to form subsidiaries dealing in mortgage-backed securities and municipal bonds. No legislation resulted.Footnote ⁷¹

Even as regulators altered their focus and Congress failed to react decisively to keep up with this change, the statutory text itself continued to act as a significant constraining force. The bureaucrats themselves were not immune to textualist misgivings about offering expanded powers without changing the laws. Several decisions made by the Federal Reserve split the Board of Governors, 5–2, with rare dissents from Chairman Paul Volcker, whose objections were based less on policy disagreement than on a lack of legal authority to approve the changes under existing law. Volcker wrote: “As the Board as a whole has repeatedly urged, the plain and desirable remedy to this legal and substantive morass is a fresh congressional mandate.”Footnote ⁷² Volcker and other Fed officials also testified in numerous congressional committees throughout the early 1980s, practically begging legislators to address “the significant erosion, at the margin, of the commerce-banking barrier” and other developing legal rifts.Footnote ⁷³ Central to these pleas was the argument that the Fed was incapable of reaching optimal policy without new statutory language. As it was, the Fed could provide legally creative (and sometimes dubious) patchwork fixes subject to judicial second-guessing, providing (at best) uncertain resolution.Footnote ⁷⁴

With regulators having largely abandoned the framing of Glass–Steagall as the necessary provider of total separation, and Congress having become somewhat ambivalent, the task of ensuring the law's continuing relevancy in the 1980s fell mostly to the courts. Judges were confronted with two starkly opposed institutional visions assigning very different roles to Glass–Steagall. Given the esoteric nature of the policy area, it is probably fair to suppose that judges had no particular policy preference for either vision. Instead, as they sought to determine statutory meaning, they balanced a generic textualist desire to preserve respect for the laws on the books with their impulse to defer to expert regulators' calls for flexibility when reasonable.

In Board of Governors v. ICI (1981), the Supreme Court deferred to the Fed's determination that banks could affiliate with closed-end investment companies, distinguishing the case from ICI v. Camp (which seemed to require the opposite result) by noting that the regulator this time offered an extensive, convincing rationale for its action and that affiliates, rather than banks themselves, would be empowered.Footnote ⁷⁵ In other words, the Fed offered an account of conversion that seemed compelling.

Continuing the trend toward deference, the D.C. Circuit ruled in favor of the Fed's decision to allow a state member-bank to deal commercial paper in A.G. Becker Inc. v. Board of Governors (1982).Footnote ⁷⁶ The Fed had decided to interpret Glass–Steagall's prohibition on banks marketing “securities” as not covering commercial paper, arguing that these instruments functioned enough like loans to fit into the core functions of banking.Footnote ⁷⁷ Showing sympathy for a regulator trying to make the best of difficult circumstances, the majority in Becker emphasized the deference due to the Fed's opinion because of the agency's expertise, the lack of an explicit definition of “notes and securities” in the statute (“we cannot assume that Congress intended the term to comprise a set of rigid and unchanging categories”), and the thoroughness and validity of the board's reasoning.Footnote ⁷⁸ The court thus endorsed the Fed's functional analysis, agreeing that commercial paper was more like normal bank lending than securities speculation.Footnote ⁷⁹ A vigorous textualist dissent denounced the majority's holding as contrary to ICI v. Camp and Glass–Steagall's simple language, which left no room for basing regulatory decisions on functional equivalencies.Footnote ⁸⁰ Statutory language that had seemed clear for many decades was thus unsettled by a novel reading backed by a new institutional vision.

The overall trend toward deference to agencies' functional interpretations was reinforced by Chevron v. NRDC (June 1984), the landmark case involving the regulation of factory pollution.Footnote ⁸¹ Justice Stevens, writing for a unanimous Court, set out a two-pronged test for agency deference fated to become canonical: first, if the intent of Congress is clear in the statute's language, this is always definitive for courts; second, if ambiguity exists, an agency's reasonable and permissible constructions of the statute deserve judicial deference.Footnote ⁸² Chevron belonged to a long-term trend toward deference, but its independent causal influence would grow as courts worked to sort out the banking laws in the years to come.

When regulators became too aggressive and appeared to simply ignore the statute's plain meaning, however, courts were still willing to take a stand. A string of Glass–Steagall cases in the years after Chevron probed the limits of deference. Two different strands of cases are particularly important. The first had to do with the Federal Reserve's ability to define what constituted a bank for the purposes of the BHC Act. The second returns to the issue of commercial paper debated in A.G. Becker.

In First Bancorp. v. Board of Governors (Feb. 1984), the Tenth Circuit showed that deference was not absolute. In this case, change was driven by loophole-exploiting private firms acting without regulators' blessing, and the Fed reverted to its earlier role as protector of Glass–Steagall. But the court ruled that attempts to arrest institutional drift could not be halted without some statutory sanction.Footnote ⁸³ “Banks” under the BHC Act were defined as institutions offering both demand deposits and commercial loans, and the early 1980s saw a rise in so-called non-bank banks, institutions that offered only one of these services and thus circumvented regulation. Pernicious loophole or not, the court found that the Fed's sudden decision to treat as a bank any entity that offered both commercial loans and negotiable orders of withdrawal (NOW) was clearly inconsistent with the unambiguous statutory definition of what was to be considered a bank.Footnote ⁸⁴ To reach its preferred outcome, the Fed had treated NOWs as creating a “legal right of withdraw on demand” and thus ignored their specific legal terms that made such withdrawals “negotiable.” Such a functional flexibility in interpreting the law might have been sensible, but the court insisted it illegitimately created an unpredictable legal environment for regulated actors doing their best to comply.

The Fed did not relent, promulgating a new official definition of a bank affecting both NOW accounts and commercial loans, this time offering more extensive justification. It quickly found itself back in front of a Tenth Circuit panel in Dimension Financial v. Board of Governors (Sept. 1984), where it argued that non-bank banks enjoyed an unfair competitive advantage by escaping Fed regulation.Footnote ⁸⁵ In reviewing the Board's change in the definition of commercial loans, the court declared that there had been no changes in common usage, congressional directive, or regulatory usage regarding the meaning of “commercial loan” since 1970 apart from the Fed's new regulation. It then chastised the Fed for making changes “with no reference to the actual meaning,” but rather “purely … to stop changes in the business of providing financial services—thus to bring a large number of additional enterprises under its jurisdiction.” The judges further scolded this “device to freeze the changes in the business of financial services until the Board could persuade Congress to act.”Footnote ⁸⁶ The court reasoned that the institutions the Fed now sought to target had perfectly legally exploited “permitted exceptions” and could not suddenly be treated as if they had impermissibly evaded the BHC Act. If anyone was going to fix this “problem, if it is that,” it would have to be legislators, who had proven themselves perfectly capable of tending to the BHC Act in the past.Footnote ⁸⁷ Statutory text thus enabled industry-driven drift, even when regulators opposed it.

A unanimous Supreme Court affirmed in Board of Governors v. Dimension Financial Corporation (Jan. 1986), following the lower court's logic closely.Footnote ⁸⁸ Chief Justice Burger found that Chevron's first prong applied, since the statutory language creating the loophole was unambiguous and therefore did not leave the Federal Reserve latitude to close it at will.Footnote ⁸⁹ The Board, in this case seeking to combat change, justified its position by arguing that “a statute must be read with a view to the ‘policy of the legislation as a whole’ and cannot be read to negate the plain purpose of the legislation.”Footnote ⁹⁰ The Court, however, declined to entertain this purposive logic, insisting that such imperfections in the statute were Congress's to fix.Footnote ⁹¹

At stake in the Dimension cases was whether Congress could be cut out of policymaking in an area dominated by experts. If the courts had assented to the Fed's policy changes, they would have effectively given the banking agencies unlimited scope in deciding how and when to allow drift or conversion, thereby depriving Congress of ultimate authority to shape the functioning of particular statutes. As with Justice Blackmun's 1971 ICI v. Camp dissent, the Court apparently found that the language at issue was clear enough to resist reinterpretation, even in the face of contrary political will. It also showed faith in Congress's ability to provide direction for statutes facing new pressures, even when those statutes were old and rusty. Subsequent cases, however, would expose this faith as limited.

The second strand of cases, which began in 1982 with the D.C. Circuit's above-described A.G. Becker decision, temporally overlaps with the first. In its June 1984 case known as Becker, the Supreme Court overturned the D.C. Circuit's decision to treat marketing commercial paper as permissible for commercial banks.Footnote ⁹² Justice Blackmun, writing for the majority, insisted that commercial paper “falls within the plain language of the act,” and therefore should be treated as a security for the law's purposes.Footnote ⁹³ The Fed's turn in this case toward bank safety and soundness was deemed illegitimate, as it required disregarding statutory language.Footnote ⁹⁴

Becker shares with Board of Governors v. Dimension Financial (to be decided by the Supreme Court nineteen months later) an insistence that the Fed is not allowed to treat itself as ultimate arbiter of how to “make the statute as a whole work.” Instead, Congress must be the source of this guidance, with its inaction representing an “adherence” to previous statutory language.Footnote ⁹⁵ But not everyone on the Court dismissed the Fed's articulation of its institutional perspective out of hand. Foreshadowing things to come, Justice O'Connor, joined by Justices Brennan and Stevens, dissented in Becker. Because there was no explicit definition of “securities” provided in the statute, the dissenters would have deferred to the Board's determinations on grounds of expertise, in line with Chevron (handed down three days earlier).

In the wake of Becker, courts attempted to figure out what to do with the Comptroller's similar decision to allow banks to establish individual retirement accounts (IRAs). ICI v. Conover (N.D. Cal., Aug. 1984) aligned with Becker, insisting that the Comptroller was charged with administering unchanging statutory language and not with creating its own discretionary law.Footnote ⁹⁶ This was to be the high-water mark for the textualist position embodied in Becker, though. In a parallel action, also called ICI v. Conover (D.D.C., Nov. 1984), the ruling instead followed Chevron in finding the Comptroller's IRA decision worthy of deference, and the same result obtained in another nearly identical action, ICI v. Clarke (D.Conn., Jan. 1986).Footnote ⁹⁷ The Second Circuit affirmed the deferential result in a quick per curiam opinion in May 1986, and a unanimous D.C. Circuit followed suit later in the month.Footnote ⁹⁸ Noting the extensive justification provided by the Comptroller in making his ruling, the court distinguished ICI v. Camp and determined that the key statutory language and legislative history was ambiguous, thus warranting Chevron deference (though nothing had really changed since 1971).Footnote ⁹⁹ Rather than seeing statutory deficiencies as being Congress's to correct, the court wrote that “Congress has charged the Comptroller with making” complex economic judgments “in the first instance,” and that it was none of the court's business to interfere with that delegation.Footnote ¹⁰⁰ Shorty thereafter, the Ninth Circuit overturned the outlying Northern District of California, declaring that “whether [the Comptroller] performs his task wisely is not for us to decide.”Footnote ¹⁰¹ Much to the dismay of the investment bankers seeking to keep Glass–Steagall's wall of separation robust, the Supreme Court denied certiorari in November 1986.Footnote ¹⁰²

Finally, capping this strand of cases, the D.C. Circuit took up another case called Securities Industry Association (SIA) v. Board of Governors (Dec. 1986).Footnote ¹⁰³ A unanimous three-judge panel challenged Becker head-on, upholding a permissive commercial paper ruling made by the Fed on remand from Becker that circumvented the Supreme Court's specific complaints by relying on a different section of Glass–Steagall.Footnote ¹⁰⁴ Noting that the Board had now “comprehensively addressed the language, history and purposes of the Act,” the court found the action met Chevron's deferential standard: the interpretation could be reconciled with the statute's text and was reasonable.Footnote ¹⁰⁵ Once again, this time perhaps more surprisingly, the Supreme Court denied cert. in June 1987, choosing to remain silent as its own precedent was distinguished into oblivion.Footnote ¹⁰⁶ Quietly but effectively, then, the Court allowed Chevron deference to become the controlling rule in banking regulation.

Why did the Court ultimately decide to accept the institutional perspective of conversion advanced by regulators? It seems likely that as the 1980s wore on and congressional attention failed to be converted into congressional action, textualist insistence on the legislature's ability to provide guidance no longer seemed grounded in a realistic understanding of the policymaking process. Holding the line until Congress acts, even as doing so frustrates bureaucrats' efforts to make the overall system of regulation function sensibly, became a difficult proposition to defend. In the persistent absence of renewed congressional action, judges' limited capacity and resources made acquiescence to the banking regulators the court's best hope for achieving a consistent, sensible legal environment for banks. In such a fast-changing context, achieving the coherence aspect of the rule of law was in tension with remaining faithful to a fixed understanding of the law's text. Though investment banks protested to what they characterized as drift, given the larger context the Court decided that the long-settled term “securities” could be reinterpreted on behalf of regulators' legitimate purposes.

Although judges might hope that letting the executive branch have its way would secure some modicum of coherence, the resolution that bureaucrats could provide was in many ways problematic, being both uncertain and jurisdictionally messy. With the recent commercial paper holdings allowing banks to act so much like securities firms, the Securities and Exchange Commission (SEC) asserted its right to regulate them and convinced a D.C. District Court judge of its power to do so in ABA v. SEC.Footnote ¹⁰⁷ A unanimous D.C. Circuit panel reversed the ruling in November 1986, stating that although “the original administrative construction of the Glass–Steagall Act has been dismantled piecemeal over the last fifty years,” this did not justify the SEC's functionalist argument for jurisdiction, which had no basis in the text of laws that made jurisdiction a function of “the species of financial institution.” Since its attempt to adapt to new institutional developments had no basis at all in statutory text, whether the SEC had a sound conversion story of its own struck the court as beside the point.Footnote ¹⁰⁸ In her opinion for the court, Judge Wald sympathized with the SEC's argument from functional need, but nevertheless asserted that “it is not the job of courts to play the game of ‘what if’ in predicting hypothetical congressional reactions to hypothetical circumstances,” especially when the statutory language itself was so clear.Footnote ¹⁰⁹ Showing worries for courts' capacities, Wald wrote that allowing such untethered improvisation “would embroil courts in constant controversies as to whether the context that had changed was really relevant.”Footnote ¹¹⁰ Here, in other words, interpretive creativity stemming from a compelling institutional framing would not be allowed to circumvent clear text—with the unfortunate consequence of opening a gap in regulatory supervision that would grow in the coming years.

Transformation of the Glass–Steagall regime continued with regulator encouragement now—for example, with the FDIC's decision to grant the state banks it regulated the ability to have securities affiliates. The D.C. Circuit upheld this decision in ICI v. FDIC (Apr. 1987), finding that Glass–Steagall's language was clearly meant to give regulators the power to make such a choice, and therefore Chevron's first-prong analysis was all that was necessary to decide the question.Footnote ¹¹¹

Institutional Transformation Consummated: Late 1980s to 1999

Beginning in the late 1980s, many legislators came to share regulators' belief that converting the institutional pieces of the Glass–Steagall regime to modern finance required significant reshuffling. But their attempts to deliver reform ran aground against the original vision of Glass–Steagall as a strict separation of banking and commerce, defended by the securities industry just as vigorously in Congress as in court. Ironically, during this period, commercial bankers' successes at winning regulatory accommodation from regulators lessened the urgency of their push for broad legislation, as they could afford to oppose any amendments that failed to match the steadily improving treatment they were already receiving under reinterpreted existing law.

Congress seemed poised to provide Glass–Steagall leadership in 1987, when it brought non-bank banks into the BHC regulatory framework in the (mostly unrelated) Competitive Equality Banking Act. It simultaneously signaled its intent to take larger action against regulator-driven change by imposing a one-year moratorium on regulatory approval of BHC applications to engage in real estate, insurance, or securities underwriting activities, giving legislators time to draft a comprehensive proposal. That came in 1988 in the form of the bipartisan Proxmire-Garn bill, negotiated by Senate Banking Committee Chairman William Proxmire (D-WI) and Ranking Member Garn, which would allow banks to market securities subject to Fed supervision and “firewalls” prohibiting risky arrangements, as well as conferring a limited ability to market insurance. With ABA and insurance industry support, the bill sailed through the Senate, 94–2.Footnote ¹¹² It seemed that embracing conversion had become the ascendant, bipartisan consensus. In the House, however, Glass–Steagall reform became entangled with other priorities. Banking Committee Chairman Fernand St. Germain (D-RI) insisted on including strict consumer protections and an expansion of the Community Reinvestment Act (CRA); House Energy and Commerce Committee Chairman John Dingell (D-MI), a champion of the securities industry, wanted to sharply limit bank affiliates' permitted activities (no doubt continuing to frame the issue in terms of preventing socially harmful drift).Footnote ¹¹³ No compromise was reached, and the congressional session ended without House action.

At this point, new Fed Chairman Alan Greenspan wrote and repeatedly testified to urge action, with a thinly veiled message that the Fed would act on its own if Congress would not. The outgoing Proxmire (headed to retirement deprived of a career-capping achievement) supported Greenspan's message, but St. Germain was emphatic that he would not be dictated to by the Fed, saying he would feel free to overturn Fed actions—a hollow threat, since what the Fed was demanding was congressional action.Footnote ¹¹⁴

With the fallout from the savings and loan crisis dominating the Banking Committees' attention in the years to come, St. Germain never made good on his promise. The giant Financial Institutions Reform, Recovery, and Enforcement Act of 1989 failed to address banking powers except to empower banks to acquire distressed thrifts.Footnote ¹¹⁵ The Fed and OCC continued apace in their subordination of commercial-investment bank separation to other goals. In November 1989, the Fed allowed banks' subsidiaries to derive as much as 10 percent (up from 5 percent) of their revenue from their otherwise ineligible securities activities.Footnote ¹¹⁶ This had an immediate impact: total bank-ineligible securities underwritten by these subsidiaries nearly doubled from $36.2 billion in the third quarter of 1989 to $68.7 billion in the fourth quarter, and by 1991 these firms accounted for 10 percent of the market for certain types of financial products, including asset-backed securities.Footnote ¹¹⁷ By 1995, they would represent nearly 15 percent of all broker-dealer assets, meaning some of the most important firms in the securities business were being regulated as if they were purely commercial banks—a problem the GAO diagnosed as endemic to congressional inaction, and not eradicable by the regulators themselves given existing statutory language.Footnote ¹¹⁸

In 1991, at the urging of President Bush, Congress took up Glass–Steagall reform again alongside an overhaul of the nation's acutely distressed deposit insurance system. By this point, even the securities industry became cautious supporters of reform, wanting reciprocal access to banking activities instead of defending an increasingly illusory strict separation.Footnote ¹¹⁹ “Resisting drift” had become an obsolete institutional framing. But embracing conversion would be useful to them only if it enabled banks to affiliate with non-bank commercial enterprises, since most controlled non-financial subsidiaries.Footnote ¹²⁰ A broad compromise passed by the House Banking Committee successfully fended off most amendments that would have threatened this nascent coalition, including Rep. Jim Leach's attempt to prevent affiliations between banking and commerce. At that point, the bill had to face the demands of four other committees, where opposing interests once again bogged it down.Footnote ¹²¹ In the end, Glass–Steagall reform was thrown overboard that year to secure a narrow consensus on deposit insurance reform.Footnote ¹²²

The Republican takeover of 1994 gave Glass–Steagall reformers new hope, but it was to prove unfounded. While favoring conversion may have become the dominant view, determining just what shape changes ought to take remained quite contentious. Regulatory changes already in place had lessened the urgency felt by many commercial bankers, who were doing much of what they wanted without having had to change the underlying statutes. Some members of Congress, including new House Banking Committee Chairman Leach, were still determined to “bridge the gap between law and reality,” but once again other priorities intervened.Footnote ¹²³ Although Leach's committee easily passed his reform bill in 1995, resistance from the Commerce Committee and disinterest in the Senate doomed the effort once again. In an attempt to salvage the effort, reform was joined to various regulatory relief provisions, along with a permanent moratorium on OCC expansion of bank powers. This last provision would prevent the commercial banks' favored regulator from directing the course of future change, and, therefore, it soured commercial bankers on the entire package. The compromise soon unraveled.Footnote ¹²⁴

Commercial bankers could afford to reject legislative action in part because they expected the Supreme Court to ratify OCC decisions confirming some bank insurance powers, opening a whole new frontier for new business lines to support their profitability.Footnote ¹²⁵ They were not disappointed: unanimous decisions in Nationsbank v. Variable Annuity Life Insurance Co. (1995) and Barnett Bank v. Nelson (1996) deferred to the OCC, paving the way for BHC affiliates to sell insurance.Footnote ¹²⁶ The ruling allowed banks to greatly expand their insurance operations and thus further weakened their demands for legislation.

Back in the legislative arena, in 1996 Chairman Leach continued to press for Glass–Steagall reform, but was ultimately compelled to table it in favor of other priorities again.

As other bills gained momentum, he tried to fill one of the regulatory gaps that had opened up by attaching a provision subjecting banks that marketed insurance to state licensing requirements, but the newly empowered banks saw the provision as undercutting their gains from regulators and courts and therefore opposed it, and Leach's efforts stalled once more.Footnote ¹²⁷

After these legislative failures, the Fed acted on its own volition to allow BHCs to expand securities operations in affiliates, changing the cap on bank affiliates' securities revenue from 10 percent to 25 percent. Similarly, the OCC began allowing bank subsidiaries to perform activities previously restricted to affiliates.Footnote ¹²⁸ These changes meant that the legislative status quo now advantaged commercial banks at the expense of the securities and insurance industries. Converting existing institutions to acknowledge the new reality now became a priority for these industries, which found that commercial banks were restricted from entering their core businesses only where the original statute was crystal-clear, while they were still barred from engaging in all commercial banking activities.Footnote ¹²⁹ The special-interest bargain Macey had countenanced when he wrote his article in 1984 was now in tatters, in large part thanks to the courts' unwillingness to support it under pressure.

In 1997, legislative efforts resumed yet again. Chairman Leach's goal was to repair the now woefully fragmented, bank-favoring regulatory edifice, advancing the goal of having a fair, coherent, and capable regulation even as the wall between commercial and investment banking was chipped away. Reform would provide an “equalitarian” framework, while also still maintaining the separation between banking and commerce.Footnote ¹³⁰ Early in the year, a merger between Bankers Trust and the brokerage Alex. Brown had been approved, representing the biggest and most blatant breach of Glass–Steagall yet. In May, former Fed Chairman Volcker told Congress that legislation was overdue, with the resolution provided by courts and regulators being uncertain and inadequate. Members of Congress other than Leach finally began to feel jealous of their own institutional prerogative, as well. Conference Chairman John Boehner (R-OH) declared, “Congress wants to preserve our right to write laws, which frankly has been overshadowed by the OCC and other regulators.”Footnote ¹³¹ The legislative wrangling during 1997 bogged down yet again, however, as there were heated intraparty disagreements about whether to allow a mix of banking and commerce (which Leach vehemently opposed). A compromise brokered by the Republican leadership in March 1998 stalled yet again when the ABA deemed the compromise unpalatable. Leach sought to lure commercial banks back to the table by including a provision to allow limited securities powers to bank subsidiaries (rather than just to BHC affiliates), and hoped to gain Democratic support by tying in an extension of the CRA.Footnote ¹³²

At this point, the political landscape was altered by the announcement on April 6, 1998, of the merger of Citicorp and Travelers Group, at that point the biggest stock swap in history. Both companies' shares surged as the intended merger was greeted with enthusiasm, despite then-existing laws requiring (in no uncertain terms) that the merged company divest itself of insurance and underwriting activities—which accounted for about half of Travelers' revenue. The Fed approved the merger, giving a five-year window for divestment—though most expected that congressional action would finally arrive before then.Footnote ¹³³

With more pressure on Congress than ever before, Leach's reform bill was reintroduced in the House in May 1998. Leach successfully stripped the provisions that would allow mixing banking and commerce, passing an amendment 229–193, and the final bill squeaked by 214–213, backed by about 2/3 of Republicans and 1/3 of Democrats after many favors were called in.Footnote ¹³⁴ Senate Banking Committee Chairman Alphonse D'Amato (R-NY) reluctantly agreed to try to get a bill through his chamber, but this attempt ran into Sen. Phil Gramm's (R-TX) intransigent opposition to provisions relating to the CRA.Footnote ¹³⁵ Meanwhile, Treasury Secretary Robert Rubin promised a veto from President Clinton if the bill vested supervisory powers wholly in the Fed rather than the Treasury Department, indicating what many called a “turf war.”Footnote ¹³⁶ Despite the breadth of the interests favoring compromise, 1998 ended without Senate action.

In 1999, the new Congress picked up where the old one left off. Finally, this was to be the year when comprehensive legislative conversion would triumph. Gramm was now chairman of the Senate Banking Committee after D'Amato's November defeat, and he directed a Senate bill that would allow non-banking subsidiaries for banks with assets less than $1 billion and was unsurprisingly uncompromising on the CRA. With its controversial provisions, it passed the Senate floor on a party-line vote, 55–44. A broader coalition was forged in the House by grafting elements of a Commerce Committee effort onto Leach's Banking Committee bill and, in spite of fierce floor battles over peripheral issues, the House bill won the administration's backing and strong bipartisan support, passing 343–86 on July 1.Footnote ¹³⁷ Conference committee negotiations featured intense conflicts, especially over the CRA, where Gramm doggedly held out until he won concessions. The Fed and Treasury eventually announced a truce in their jurisdictional disputes, agreeing on a structure that gave the Treasury much of what it wanted on supervisory powers for bank subsidiaries.Footnote ¹³⁸ After more than a month of wrangling, a compromise emerged that won passage in both houses on November 4, 363–57 and 90–8 respectively. A week later, President Clinton signed into law the Financial Services Modernization Act, better known as Gramm-Leach-Bliley (GLB).Footnote ¹³⁹

Most sources regard GLB simply as a repeal of Glass–Steagall, and understandably so. A new corporate form was born—the Financial Services Holding Corporation—which explicitly allowed commercial banks, investment banks, and insurance companies to reside under the same corporate umbrella. As noted at the beginning of this article, Rep. Leach's office could not resist frosting their celebratory cake with “Glass–Steagall, R.I.P., 1933–1999.”Footnote ¹⁴⁰ But as we have seen, although eulogies may well have been in order, “repeal” of the law's substance had mostly been accomplished by a long process of bureaucratic and judicial reinterpretation. The regulators' broad institutional focus on “banking regulation” as a whole, which had already done so much to change de facto law, now finally matched the law in the statute books. This vision did not, however, represent a wholesale deregulation of the banking industry. Rep. Leach thought of GLB as “procompetitive, but not anti-regulatory” reform, and was pleased to have ensured that the new regime would not allow the intermixing of banking with other kinds of commerce.Footnote ¹⁴¹ For those who adhered to the institutional vision of Glass–Steagall as providing a needed and useful barrier between commercial and investment banking, GLB represented the consummation of the long process of drift, or corrosive capture, with commercial bankers' interests having penetrated policymakers across the political spectrum. But the preceding narrative should make clear that the regulators and legislative reformers had come to be genuinely committed to their very different institutional perspective, leading them to believe the new institutional landscape would better serve the public's interest in a stable banking sector.