The Federal Deregulation of Insurance

Essay - Volume 97 - Issue 1

The efforts to get the federal government out of the business of regulating insurance have been comprehensive but not entire. The project offers two insights about deregulation and how to do it. The first insight is comparative. Given that courts, Congress, and agencies have all tried to undo the federal regulation of insurers, the higher quality of deregulation done by regulators themselves, as opposed to the other branches of government, is informative and makes out a story of comparative advantage when it comes to regulation. The second insight serves as a reminder of the stickiness of globalization. Because the federal government has made commitments to the European Union, it cannot entirely remove itself from the oversight of insurance, much as the policymakers in power today might wish to do so. Both insights make insurance a case study about the way that deregulation can work more generally.

Deregulation has been prioritized by the Trump Administration from the very beginning. It was touted in some of the first executive orders issued by the President, less than two weeks after his inauguration.[2] It began to be realized by Congress’s unprecedented use of the Congressional Review Act (CRA) to rescind signature rules promulgated in the last year of the Obama Administration, beginning with seven House votes in the first week of February 2017.[3]

The pursuit of deregulation has continued with a string of presidential appointees signaling their intentions to cut rules and reduce enforcement.[4] The head of the Consumer Financial Protection Bureau has vowed to stop “pushing the envelope” when it comes to enforcement and has reassigned enforcement resources to education and advocacy roles, and he is not alone in making these sorts of choices.[5] The Treasury Department has issued a series of white papers designed to reduce regulatory burdens; these white papers are becoming a customary way to set forth a deregulatory roadmap for agencies and departments.[6]

Perhaps the most dramatic example of deregulation has been the effort to abandon the federal role in the supervision of insurance companies. That oversight, less than a decade old and cautious even during the zenith of the Obama Administration, has become the subject of deregulatory attentions from all three branches of government. The goal, at least domestically, has not been to reduce the burdens of federal regulations of the industry but to eliminate them entirely.[7] Despite these ambitions, the retreat of federal oversight of the insurance industry during the Trump Administration has been comprehensive but not entire.

The experience so far of federal deregulation of insurance offers two different insights. The first concerns the right way to do deregulation domestically, for federal insurance regulation has been challenged by all parts of the government. The insurance experience suggests it is better done by agencies than by courts or Congress. The second serves as a reminder of the increasing importance of international relations in domestic regulation. It is only the fact that the United States has committed itself to obligations to the European Union that has ensured that the federal regulatory role in insurance will remain at least somewhat relevant during the Trump Administration. In many ways, the two tools the federal government has to regulate insurance—the domestic one made dormant and the international one still important—are quite different. Papers could be written about either; this Essay considers both because the changing federal role in insurance is important in its own right. Federal insurance supervision can also serve as an example of how to do domestic deregulation and the consequences of regulatory globalization.

Although states have traditionally had the exclusive power to regulate insurance, the federal government was given a role in 2010 through legislation responding to the financial crisis, which the collapse of the insurance giant AIG had exacerbated.[8] The federal government received the power to designate some insurance firms as “systemically important financial institutions” (SIFIs).[9] It then designated the three largest American insurance firms, and in so doing began to roll out a message to the industry about what sort of prudence it expected, as well as what kind of size would amount to systemic significance. In addition, a newly created Federal Insurance Office (FIO) was given the power to represent the United States before international organizations of insurance and financial regulators, centralizing in the federal government an international role that had previously been left to the insurance commission of the states and federal territories.[10]

SIFI designation has been attacked by all three branches of government, beginning in 2016. However, the federal role in insurance regulation will remain relevant as long as the Administration lives up to the so-called “covered agreement” concluded with European insurance regulators in the waning days of the Obama Administration and signed by the Secretary of the Treasury on September 22, 2017.[11]

The administrative state is experiencing a deregulatory moment, even as global entanglements make comprehensive deregulation all the more difficult given our commitments to our foreign trading partners. Understanding the possibilities of deregulation in this context matters not just for insurance but also for the deregulatory project as a whole.

I. Deregulation: A How-To Guide

The Trump Administration has found deregulation to be uneven going. The efforts so far, particularly the environmental ones, have been characterized by Ricky Revesz as “sloppy and careless, [because] they’ve shown significant disrespect for rule of law and courts have called them on it.”[12] On the other hand, those CRA resolutions, at a minimum, have made deregulation—at least comparatively—one of the Administration’s most prominent success stories.[13] The problem for the Administration’s non-CRA deregulatory efforts has frequently been one of procedural execution, with agencies failing to identify the rational basis for deregulation, or staying Obama-era rules without much justification. Courts have frequently been unwilling to excuse these failures or to take on the deregulation mantle themselves, while Congress has been distracted with other priorities and partisan divisions.[14]

To make sense of how a government might pursue an insurance-deregulatory agenda, it is useful to review the legal constraints on deregulation. The standards for deregulation by agencies are modest (or at least they should be since the seminal Chevron decision’s deference holding captured the imagination of federal judges), by courts procedural, and by Congress essentially nonexistent.[15] These constraints shape what deregulation can look like and, this Essay argues, affect the quality of deregulatory decision-making. Legal constraints alone will not tell the whole story—despite its legal flexibility, Congress has been unable to realize many of its deregulatory ambitions. But they are of great consequence when it comes to deregulatory policymaking by courts and agencies.

A. Deregulation by Agencies

For agencies inclined to deregulate, the choice to do so is evaluated on the same standard as the decision to regulate in the first place. As Cass Sunstein has put it, “Courts should not treat deregulation substantially differently from regulation.”[16]

This means that the most important constraint on deregulation is that a deregulatory removal of a rule or order must be implemented with the same degree of process that accompanied the promulgation of the rule or order.[17] Moreover, the basis for deregulation cannot be justified solely with reference to the political preferences or taste for repeal of the appointees who run the agency; instead, there must be a case made in the deregulatory order that the public interest will, defined quite broadly, be served by the administrative action.[18] As we will see, these procedural and modest substantive requirements have meant that the quality of the deregulatory decisions by the Financial Stability Oversight Council (FSOC or the Council) have been well explained, especially in comparison to the efforts of Congress and the courts.

These hurdles apply only when agencies seek to actively remove rules or orders from the books. Removing rules or orders is the most effective and lasting form of deregulation, but it is not the only kind of deregulation that agencies can pursue.

Agencies can suspend the operation of a rule if they can establish that good cause requires it—a difficult standard to meet for a high-profile rollback, without any sign of an emergency, and one that the Trump administration has sometimes tried, often unsuccessfully.[19] Another, more temporary form of deregulation can be handled by agency enforcement divisions. The choice not to bring enforcement actions to curb violations of existing rules that could have been brought is all but unreviewable.[20] For insurance regulation, this means that if the government chooses not to designate any more insurance companies as significant SIFIs, there is no risk that anyone can force it to do so with a lawsuit.

However, there are limits to the effectiveness of deregulation through nonenforcement. The Council could change its mind and ramp up the designations; the choice to restart enforcement, and against whom, is just as unreviewable as the choice not to enforce.[21] Moreover, ceasing new designations leaves the old designations on the books, with all the messages to industry that they send.

Deregulation, ideally, would be something premeditated, rather than pursued at haste. J.B. Ruhl and James Salzman think that the decision about how to stop regulating should be part of the initial calculation about when to regulate; their work on regulatory exit was the subject of the Duke Law Journal symposium on administrative law in 2018.[22] Ideally when regulating, Ruhl and Salzman have argued, “Government should also ask how it will exit when it realizes it (1) has accomplished Goal X, (2) is not achieving Goal X, or (3) has regulated more than necessary to achieve Goal X.”[23] Arguably, this exit potential is a part of federal insurance regulation. Dodd-Frank did provide for an exit from SIFI designations and required regular reviews of designations to see if they remained appropriate.[24] So did the government’s implementing regulations—it used those annual reviews to de-designate first GE Capital, and then AIG, two of the four firms (three of which were insurers) on the SIFI list.[25]

By contrast, poorly planned deregulation is fraught with the potential for mistakes. As Hannah Wiseman has observed, “[A]gencies facing immediate pressure to cut budgets, regulations, and programs will not have the time to conduct detailed comparisons of their most effective programs or . . . decide on a particular metric for effectiveness.”[26] One of the institutions involved with insurance supervision, the Treasury Department’s Office of Financial Research, has been characterized as facing this kind of chaos.[27]

It is by no means obvious that the nascent federal scheme to regulate insurance was devised with much attention as to how to reduce the federal government’s involvement—the point was to give the federal government a role, not to end it. But by creating potential for exit and institutionalizing lookback review, the federal government’s insurance regulations were created with the possibility of deregulation in mind.

In section II(C)(1), we will see how, exactly, that lookback process has been used to withdraw the Council from the field of insurance regulation.

B. Deregulation by Congress

Michael Van Alstine has written about congressional deregulation—he calls it “statutory deregulation” and argues that it is the way to deregulate most aggressively.[28] As Van Alstine has put it, congressional deregulation can result in “a dismantling of regulatory oversight over broad sectors of commerce.”[29] Moreover, as a democratically elected body without the accountability worries that agencies pose, there are few legal constraints to what a deregulatory Congress can do, conditional on its ability to pass any sort of legislation at all. Congress must establish that its deregulatory bills do not violate the separation of powers and that there is a rational basis for its decision—two tests that are famously easy to meet.[30] But there is no procedural review or requirement of a commensurate level of process when the Legislature acts, as opposed to when agencies do.

Despite the straightforwardness of the constraints on deregulation, Congress has only rarely pushed deregulatory legislation forward. There have been many false starts; during the Obama Administration, the Legislature debated, and the House occasionally passed, bills designed to throw some sand in the gears of the regulatory state. There was the Regulatory Accountability Act, which would have introduced cost-benefit, formal adjudication and other procedural requirements to rulemakings;[31] the Separation of Powers Restoration Act, which would have ended judicial deference to agency interpretations of law;[32] and Regulations from the Executive in Need of Scrutiny Act, which would have empowered Congress to issue resolutions of disapproval for any agency rule and required congressional approval of major agency rules before those regulations could be implemented.[33] Each of these statutes would have affected every agency’s efforts to make policy; none of them would have been at risk of a judicial finding of unconstitutionality, and yet the Legislature has been unable to present them as bills passed by both houses and ready for the President’s signature.

Congress’s central role in the deregulatory story being told by this Administration lies not in its normal order but in the order created by a deregulatory fast-track, one that can be used only when the stars have aligned and both the presidency and Legislature are in the same hands, after taking over for a President of a different party. In such a case, the Legislature can and has utilized the CRA[34] to facilitate congressional reversals of administrative regulations.[35] The CRA permits Congress to undo regulations through a fast-track joint resolution of disapproval.[36] Even more dramatically, it “salts the earth”—“[n]o substantially similar rule can be subsequently adopted by regulators once Congress has reversed a rule under the CRA.”[37] However, it applies only to rules adopted within the last sixty days that Congress was in session, meaning that the window for a CRA resolution is brief, though various congressmen have been mulling over ways to expand this window.[38]

During the first four months of the Trump Administration, CRA resolutions of disapproval were passed fifteen times,[39] undoing an array of rules ranging from the Interior Department’s antipollution stream protection rule[40] to the Securities and Exchange Commission’s resource extraction rule, which required mineral companies to report payments made to foreign governments.[41] The Administration also rolled back rules related to internet privacy, drug testing for unemployment compensation, and other areas.[42] The turn to the CRA was unprecedented; the statute had only been utilized once before, when the Bush Administration undid a late Clinton Administration rule on ergonomics in the workplace.[43]

The CRA was not used to effectuate insurance deregulation—neither the Council’s insurance-relevant rules nor its designations were promulgated during the final sixty days of the Obama Administration. The covered agreement with the European Union discussed in subpart III(B) may have arguably been eligible for CRA treatment, as it was passed during the final week of the Obama Administration (and was also transmitted to Congress then as well), despite it not obviously counting as a “rule” subject to CRA supervision. But the Administration eventually signed the covered agreement, and the CRA is unlikely ever to be used in any case where the incoming Administration agrees with the previous one on the merits of a regulatory policy.

Instead, the reliance on the CRA underscores how the limited external constraints on congressional action are less relevant than internal constraints imposed by the legislative process—the filibuster and other delaying tactics available to the Senate, the problems of scheduling votes where legislative time is scarce and priorities compete, and the need to harmonize legislation in two houses.[44] It all means that while deregulation may be easily achieved by Congress in theory, with little check on deregulatory powers, it is, in fact, no easy legislative task.

C. Deregulation by Courts

There has been little written about court-driven deregulation as such, but there is little doubt that the judicial deregulatory role is critical, given that courts act as a final check on agency initiatives. This means that when agencies lose court cases, the effect is deregulatory (unless they lose court cases when they are trying to deregulate)—a policy does not go into effect or a regulatory action is reversed. As Van Alstine has observed, this sort of “less conspicuous” deregulation can be accomplished by courts invoking constitutional or administrative law.[45] Courts review agency regulations pursuant to the Administrative Procedure Act (APA), which was passed, as George Shepherd has argued, with the support of regulated industry because it was thought it would provide a check on the growth of the New Deal state.[46]

The judiciary has taken up the charge of the APA; judges reverse agency actions approximately one-third of the time, making judicial review a real hurdle.[47] This check on regulation has proven vibrant, despite the fact that, doctrinally, judicial review is supposed to be deferential to findings of fact and reasonable constructions of law.[48] Courts do not defer on the required procedures that agencies must follow, however, and are as likely to reverse agencies on the matters on which they are supposed to be deferential as on the matters where they are not.[49]

There are limits to what courts can do, however. The traditional remedy upon reversal is to return the rule or order to the agency with instruction to try again if it chooses, meaning that a judicial order voiding a regulation is ordinarily accompanied by an invitation to the agency to regulate in a similar, but more procedurally complete way.[50] As Nicholas Parrillo has shown, courts rarely used their contempt powers to bring agencies into line, even after repeated violations of the law, meaning that courts, unlike Congress through its CRA review, are very unlikely to salt the earth with a deregulatory injunction.[51] Judicial review, as the APA’s name suggests, is principally for procedural mistakes (as well as legal ones), though courts have reserved for themselves the right to take a hard look at the substantive basis for the agency rule.[52]

Deregulation through courts is also a narrower process than when done through Congress. Courts must wait for a plaintiff to put a regulation before them and then may only pass on that particular regulation, rather than on a broader regulatory program involving a variety of rules and orders. In section II(C)(2), we will see how one court interpreted its role to reverse an order extending extra supervision to an insurance company; it focused on procedural failures and a disagreement on statutory interpretation to do so, but it did not, and probably could not, conclude that the rules underlying the designation decision were themselves defective—it could only reject those rules as applied to one insurer.

II. The Retreat from Regulating Big Insurers

SIFI designation of insurance companies is overseen by the FSOC, a committee of financial regulators created by the Dodd-Frank Wall Street Reform Act to consider broad risks to the American financial system. Designated firms are subject to extra supervision by the Federal Reserve Board; the hope is to ensure that the SIFI is well-capitalized and not engaged in overly risky business schemes.[53]

The Council has always been an idiosyncratic regulatory entity—a creation of financial crisis legislation that jammed agency heads together to ensure that the broad view of the financial economy was being considered by these heads at some point, instead of creating a new consolidated financial market regulator to take this view. The new Administration has tried to get it out of the business of regulating insurance in two ways. First, the Council itself has de-designated designated insurers—a deregulation strategy—and hinted that it will not designate more insurers—a nonenforcement strategy. Second, the Treasury Department has recently proposed that the Council should be reformed in a way that would have it do less oversight and operate more as a think tank for its member regulators.[54]

Congress has also tried to cut the Council out of insurance oversight. The Financial CHOICE Act, which passed the House on June 8, 2017, would completely eliminate the Council’s designation power.[55]

Congress’s effort to remove powers from the Council must be considered in light of the way the Council used those powers—and how the Council itself, and the courts, have cut back on insurance regulation, in addition to how Congress might do so itself. In particular, of the three insurance companies and one commercial lender who were designated as systemically significant, none remain subject to the Fed’s oversight as of this writing.[56] Two of the insurance companies—AIG and Prudential—were de-designated by the Council itself, along with the commercial lender—GE Capital.[57] MetLife got out of its designation by contesting it in court.[58]

The sample is not large, but it permits a qualitative comparison of the way deregulation can be done by an agency and the way it can be done by a court or the Legislature. The Council considered the substantive basis of its designation decision in its decisions to de-designate and illustrated how the de-designated firms had transformed themselves into less risky entities. The court focused on what it perceived as procedural missteps in the initial regulation, while the Legislature has sought, somewhat bluntly, to end the designation power altogether.

More generally, the de-designation process illustrates how courts and agencies deregulate: courts deregulate procedurally and agencies substantively, while Congress simply removes regulatory powers altogether, or did in its bills designed to deregulate insurance. Normatively, there are reasons to prefer deregulation on substantive grounds over regulation that is stopped for procedural reasons. But, to be sure, this substantive advantage for agency deregulation in the effort to remove the federal government from insurance oversight is not immutable. Agencies could slow future regulators’ efforts by promulgating internal procedural hurdles. Congress can deregulate in any way it wants to—bluntly, or with a scalpel—by piling on procedural hurdles or by imposing substantive exemptions. Even courts, through reference to hard-look review, can reverse agency regulatory efforts on substantive grounds; they are not tied only to procedure, even though procedural missteps are the usual focus. The argument in this Essay is only that the various parts of the government may have certain comparative advantages when it comes to deregulation, and that insurance deregulation exemplifies them.

A. How the Regulation of Individual Firms Could Constitute Regulation of the Entire Sector

The prospect that nonbanks could be designated as systemically significant and subjected to oversight by the Fed was something that worried a number of very large financial companies.[59] GE Capital and AIG were unable to avoid initial designation, but once designated both firms transformed their operations and their balance sheets in an effort to escape the process.[60] Both firms shrank, reduced their reliance on short-term sources of funding, sold off businesses, and reduced their leverage, or ratio of total assets to equity.[61] Prudential did so much less.

The insurance industry responded to the prospect that any insurance company might be designated with alarm, but it was not alone. When the Treasury Department’s Office of Financial Research issued a white paper suggesting that certain large asset managers could, and perhaps should, be subject to designation, the response from that industry was intense.[62]

Designation, and the attendant oversight by the Fed, was viewed by these firms as a penalty, illustrating that SIFI designation put the federal government in the business of insurance regulation, a function of its penalty power. Penalties are often part of functional administrative schemes, and there are reasons to believe that a regime that penalizes some nonbanks for growing too large and interconnected is a useful way to regulate both those institutions and the rest of the insurance sector. It can make sense to subject specific firms to intensive oversight because of the particular risks posed by the firms. In the run-up to the financial crisis, AIG—a then very large American insurance company overseen by a welter of state and federal regulators (AIG’s primary federal regulator was the now-shuttered Office of Thrift Supervision, involved because one of AIG’s subsidiaries held a thrift charter)—got involved in two different businesses that both posed substantial run risk.[63] In one, its financial products subsidiary sold insurance contracts on corporate credit (so-called credit default swaps) and failed to properly hedge them.[64] In the other, it engaged in securities lending, which meant that it was loaning out securities it held in its treasury in exchange for a payment and cash collateral, both of which it could reinvest.[65]

When the financial crisis hit, AIG was obligated to post more collateral to satisfy its counterparties than it could make good on its credit insurance; at the same time, its securities-lending funding dried up, as many of the borrowers of its securities were the kind of financial institutions that got into trouble during the crisis.[66] AIG required a massive bailout, and recognizing the way that its runnable businesses could interact to topple the firm would have been difficult.[67] But uncovering these sorts of connections are the very particulars that a team of supervisors tasked to a single large firm might be able to suss out. Targeted, whole-firm supervision of a conglomerate like AIG might offer regulators a full picture of the risks being taken by an insurer otherwise supervised on a piecemeal basis.

But more broadly, making examples of three insurers provided guidance to the rest of the industry about how the federal government thought about what constituted systemic risk in the provision of insurance services.[68] Moreover, regulating the insurance industry by looking very closely at three insurance companies arguably also constitutes cost-effective supervision. There is evidence, for example, that the industry as a whole has reduced its holdings of risky assets since the designations—but federal regulators did not have to place supervisors in each and every insurance company in the country to bring about this result.[69]

Moreover, the designation process did not only discipline nonbank firms; it also disciplined regulators who were failing to regulate their industry to a degree that the Council thought appropriate—those regulators risked having their regulatory-oversight responsibilities transferred to the Federal Reserve through a SIFI designation, a loss of turf that few agencies would appreciate.[70] That prospect may have improved state regulation of insurer solvency.[71]

Still, the use of SIFI designations on insurance companies, even for AIG, was controversial. They marked the first federal foray into insurance regulation, which, under the McCarran-Ferguson Act, had been exclusively reserved to the states.[72] Dodd-Frank created a statutory basis for federal involvement in insurance, but the insurance industry and state regulators have expressed opposition to the federal role ever since the passage of the statute.[73] Their fear was that any role for the national government, no matter how narrowly tailored, would constitute a first step in the federalization of insurance oversight.

Critics relatedly—and as it turned out, incorrectly—also worried that the designation process would be a “Hotel California,” from which no designated nonbank would ever be able to leave.[74] Some observers also suspected that the Council took its designation cues from an international process. The global Financial Stability Board (FSB), a group of regulators from all over the world that American agencies had joined, designated the three American insurance companies as globally risky before the American agencies performed their own analysis.[75] The chair of the Senate Banking Committee had wondered whether the Council’s designation decisions were “sufficiently open, objective, data-driven, and free from the influence of outside organizations”—by which he meant the FSB.[76]

The quantum of discretion afforded designation decisions by the Council has occasioned widespread debate. Jeremy Kress has praised the flexibility of the process;[77] Daniel Schwarcz and I have found the administrative component of designation to be a reasonable and sensible response to risk in insurance.[78] On the other hand, Christina Skinner has written that there ought to be more levels of designation than the binary choice the Council has to make.[79] To her, the fearsome prospect of designation, paired with the alternative of doing nothing, resembles regulatory regimes that are often criticized for being too blunt.[80] The Department of Agriculture’s meat inspection service, for example, can pull its inspectors from a plant when they find violations, or it can do little else, and it has often been criticized for that all-or-nothing dynamic.[81] Skinner has proposed a sort of graduated oversight, including an information-gathering intermediate step called SIFI Lite.[82] The Treasury Department itself has suggested that the Council might be better served identifying activities that should be regulated for risk, rather than firms.[83] The suggestion would transform the Council from being a regulator to one making recommendations to its member regulators about what they should supervise—it would become something of a think tank.[84]

Other commentators have suggested that the Council should prepare a cost–benefit analysis accompanying any designation decision. The district court that reversed the Council’s designation of MetLife insisted, quite implausibly, that the Council’s governing statute required a cost–benefit analysis, a claim that is surprisingly common when it comes to the judicial supervision of financial regulators.[85] Cost–benefit analyses are controversial, however, because of the difficulty of quantifying the benefits of a regulation designed to avoid a financial crisis.[86] Such benefits are extremely hard to forecast; they involve macroeconomic employment and growth projections subject to a wide array of outcomes depending on the assumptions made.[87]

B. The Council’s Process

As a group of agencies, the Council is best understood as a regulator made up of regulators. The Council is chaired by the Treasury Secretary and includes nine other federal financial regulators, all of whom are the heads or chairs of their respective agencies.[88] It also includes, in a non-voting capacity, some non-federal regulators—a state banking supervisor, a state insurance commissioner, and a state securities administrator, as chosen by the state regulators in each issue area.[89] While this state participation is relatively unique, only federal regulators get to vote on designations.[90]

As for the statutory guidance governing those regulations, Congress promulgated the sort of multi-factor balancing test that former Justice Antonin Scalia used to bemoan.[91] Congress gave the Council eleven factors to apply to designations, one of which included “any other risk-related factors that the Council deems appropriate.”[92] The Council, in turn, added content to that broad mandate by promulgating a regulation indicating that it would principally focus on size, substitutability, interconnectedness, leverage, liquidity, and existing regulatory scrutiny when deciding whether to designate a firm as systemically important.[93]

Those six factors were applied to the three designated insurance companies through a three-step process. The first test was straightforward and quantitative and focused mainly on size. Nonbank financial companies with more than $50 billion in assets pass this stage.[94] The initial stage is followed by two qualitative stages: one in which the Council assesses the riskiness of particular institutions that pass the quantitative threshold using a broad array of publicly available data, and a second, if necessary, where the institution can present evidence to the Council designed to persuade it not to designate.[95]

The MetLife designation exemplifies the process. After passing the first and second stages of FSOC review, MetLife was informed that it was being considered for designation and offered an opportunity to respond. The firm submitted over 21,000 pages of materials to the Council and made its case to regulators in person as well as through briefing.[96] After considering these materials and the argument of the firm and recommendations of Council staff, the Council, by a vote of 9–1, with the federal insurance expert dissenting, officially designated MetLife.[97]

In reaching its conclusion, the Council discussed each of the ten statutory factors in Dodd-Frank and the six categories contained in its final rule. The Council observed that MetLife was the largest insurance provider in the United States and was “significantly interconnected to insurance companies and other financial firms through its products and capital markets activities.”[98] These activities, including securities lending and the funding of agreement-backed notes, created liabilities “that increase the potential for asset liquidations by MetLife in the event of its material financial distress.”[99] The Council posited that “MetLife’s complexity, intra-firm connections, and potential difficulty to resolve” could aggravate the risk that financial distress at the company could impair financial market functioning.[100] Additionally, while acknowledging that MetLife’s operating insurers were subject to state insurance regulation, the Council noted that this regulation was focused predominantly on protecting policyholders rather than on safety and soundness; by regulating the firm on a state-by-state basis, the existing scheme made it difficult to impose and monitor firm requirements about capital buffers—liquidity requirements designed to ensure that shocks would not be fatal.[101]

C. De-Designation by the Council and the Courts

Insurance deregulation has been done much more coherently by the Council than by the government’s other deregulators. It offers some evidence that if deregulation must be pursued, agencies have the ability to focus on substance more than courts do, and they enjoy an expertise advantage over Congress, which, at least in financial regulation, tends to spend its time considering blunt and broad deregulatory bills, which it then fails to pass.

1. De-Designation by the Council.—We can see how the Council deregulates through its decisions de-designating AIG, Prudential, and GE Capital. In the Council’s decision that de-designated GE Capital on June 28, 2016, it emphasized that the firm had transformed itself in an effort to reduce its potential to create systemic risk. As the Council observed, “GE Capital has fundamentally changed its business.”[102] In particular, it had engaged in “a series of divestitures, a transformation of its funding model, and a corporate reorganization,” all of which made the company substantially smaller.[103] It was a “much less significant participant in financial markets and the economy” and, in addition, had “shifted away from short-term funding, and reduced its interconnectedness with large financial institutions.”[104] The Council also observed that GE Capital “no longer owns any U.S. depository institutions and does not provide financing to consumers or small business customers in the United States.”[105] All told, the de-designation order was twenty-three pages long. It cited statutory authority and the Council’s own regulation. In addition to considering the shrunken size of the changed firm, it evaluated its contagiousness and emphasized that the risk that other firms would fail if GE Capital failed was one of the drivers of the decision to designate, and to de-designate.

For these reasons, the Council removed its regulatory oversight (which, again, it exercised through the Fed) over a firm that had changed and become less systemically risky. The interconnectedness of the firm particularly mattered to the Council’s analysis; the concern was not whether it would be bad for the economy if GE Capital failed but whether other firms would be more likely to fail if GE Capital failed. One of the reasons for the financial crisis bailout of AIG was the fact that so many of its counterparties were financial firms depending on it to stand behind its credit default swaps, which can be understood as a form of insurance written on the prospect of a corporate bond default.[106] The Council wanted to make sure that GE Capital’s failure would not start a chain reaction. Hence the concern about whether other parties depended upon GE Capital and whether other lenders could pick up the firm’s slack.

To be sure, the Council also considered the firm’s financing. GE Capital was designated in part because of its size, but also because it relied on short-term debt that could disappear quickly in difficult times. The idea is that a firm that depends on rolling over debt every month might be more subject to market fluctuations than a firm that would not have to worry about finding financing until next year. By reducing its reliance on short-term debt, GE Capital managed to reassure the Council that this was not a risk.

For AIG, the Council relied on comparisons between the firm and its peers. The comparisons indicated that in changing its business model, the firm had become much smaller than the biggest banks, and smaller even than an undesignated insurance company, Berkshire Hathaway, in total assets.[107] Moreover, the company held less debt than the large banks subject to extra Fed supervision; this debt was larger than that issued by its insurance peers Prudential and MetLife but, once again, was lower than the undesignated Berkshire Hathaway.[108] After, among other things, exiting business in ways that made that company less interconnected with the rest of the financial system and reducing “the amounts of its total debt outstanding, short-term debt, derivatives, securities lending, repurchase agreements, and total assets,” in some cases significantly as FSOC put it, AIG also won de-designation.[109] The Council observed that the firm had reduced its securities lending and its leverage—to the point that assets were 5.4 times larger than equity, making the firm leveraged, but substantially less leveraged than the largest banks.[110] Finally, the Council emphasized that only $9.4 billion in credit default swaps—the derivatives that had brought AIG lower during the financial crisis—had been written on the firm’s credit.[111] That was far lower than the $25.6 billion outstanding referencing Berkshire Hathaway’s credit, to say nothing of the $20.5 billion referencing Goldman Sachs and the $20.9 billion referencing Morgan Stanley.[112] As the Council indicated,

capital markets exposures to AIG have decreased, and the company has sold certain businesses in which it held dominant market shares, rendering the company less interconnected with other financial institutions and smaller in scope and size. . . . While this liquidity risk is material, the Council’s analysis . . . indicates that the level of forced asset sales by AIG in the event of its material financial distress may be lower than previously contemplated . . . .[113]

In sum, the Council noted, AIG was a “notably different” business than the one that was initially designated.[114] It accordingly ordered that the firm’s designation as a SIFI be revoked.

Prudential’s de-designation order, promulgated on October 16, 2018, was sixty-six pages long, the lengthiest such FSOC order yet, and once again, it was analytically coherent, applying the standards identified by the Council as the ones on which designation turns in depth.[115] The challenge for the Council was that its de-designation decision was, for the first time, made for a firm that had, in most relevant senses, gotten bigger than it was when it was first identified as systemically important.[116] If, as a result, the order is less persuasive than the Council’s two other de-designation decisions, it is recognizably in the same genre.

The decision included a qualitative and quantitative analysis, focusing on a finding that any distress at the firm would be unlikely to be transmitted to the broader financial markets.[117] In particular, the Council considered three ways that trouble at Prudential could be contagious.[118] For fire sales of assets brought on by the potential disappearance of sources of funding (the “Asset Liquidation Transmission Channel”), the Council concluded, for example, that “there is not a significant risk that a forced asset liquidation by Prudential would disrupt trading in key markets or cause significant losses or funding problems for other firms with similar holdings” and that Prudential’s fire-sale risk was “manageable” by the firm itself.[119] For exposure risk (the “Exposure Transmission Channel”), the Council concluded that despite the fact that unlike other designated firms Prudential had not shrunk, and instead had increased its securities lending and derivatives commitments, few other firms were exposed to risks in a way that would threaten their viability.[120] For replaceability risk (the “Critical Function or Service Transmission Channel”), the Council concluded that despite the fact that Prudential is the country’s largest insurer, other firms could take its place if it imploded, given that the insurance markets were “highly competitive.”[121] The order reviewed all the assets held by Prudential, often in comparison to the similar asset holdings of other institutions.[122] It, for example, reported that Prudential was the 16th and 11th riskiest firm when it came to fire sales, assuming two different sorts of shocks.[123] In all, though the order is wanting in some particulars, the Council made an elaborate and lengthy case for

In all three de-designation cases, the Council examined the balance sheets of the companies in detail and assessed the role of the firms in the broader financial system. As we have seen, the critical criterion for the Council was contagion—the regulators wanted to establish that if the firms failed, it would be unlikely to create a financial panic with broader implications for the domestic economy.

There are both process and structural advantages to the Council’s de-designation role.

On the process front, although the Council works differently than other agencies, like all of them, it is required by the APA to consider evidence, apply that evidence to the governing legal standards, and issue a reasoned decision, reviewable by the courts.[124] The Council hears from applicants for de-designation and has issued a roadmap for how these decisions will be made; that roadmap commits the Council to a process that it must follow.[125]

Structurally, the Council is comprised of the heads of a diverse array of federal financial regulators. De-designation decisions require two-thirds of these regulators to agree on the decision, along with the Secretary of the Treasury.[126] Although the heads of all the agencies with roles on the Council are political appointees, their appointments are often staggered, with terms of four to six years, making it possible for holdovers from previous administrations to be serving alongside new appointees (as was the case at the beginning of the Trump Administration, when the chairs of the Fed and the FDIC, the directors of the Federal Housing Finance Administration and the CFPB, and the independent insurance member had been appointed by the previous President).[127] As time passes, an Administration can be more sure of getting sympathetic appointees on the Council, but even then, the members of the Council will head independent agencies, largely insulated from political control, and will have different perspectives and priorities on questions like insurance regulation. This sort of diversity of opinion is not in and of itself a guarantee of wise policymaking, but scholars ranging from Eric Posner and Cass Sunstein to Condorcet have argued that different perspectives can improve decision-making.[128] At the very least, the evidence on insurance deregulation suggests that there is something to the required process and structure of decision-making.

2. De-Designation by the Courts.—The Council’s substantive version of deregulation can be compared to the procedural focus of the court that undid MetLife’s designation, which was less compelling.

MetLife filed suit to undo its designation by the Council in the U.S. District Court for the District of Columbia. That court reversed the Council’s designation on two grounds. The first emphasized that the Council had failed to apply the factors it had previously indicated were critical to the decision to designate to MetLife in a manner the court could credit. Rather than assessing the riskiness of MetLife, the court concluded that two of the Council’s criteria were ignored without adequate explanation. The court complained that the Council had “abandoned the Guidance and refused to evaluate MetLife’s vulnerability to material financial distress,” and that its designation “hardly adhered to any standard when it came to assessing MetLife’s threat to U.S. financial stability.”[129] It meant that the designation decision itself was “fatally flawed.”[130]

The second ground for reversal turned on the Council’s failure to do an adequate cost–benefit analysis in making its designation decision.[131]

Getting too far into the weeds of whether the Council applied its standards consistently is unnecessary. But it is fair to say that the court relied on a gotcha form of consistency; it concluded that FSOC had conflated two inquiries it had promised to perform into one. In particular, it concluded:

The [Council’s] Guidance divided six categories of analysis into two distinct groups. The first group (size, substitutability, and interconnectedness) was meant “to assess the potential impact of the nonbank financial company’s financial distress on the broader economy.” The second group (leverage, liquidity risk, and maturity mismatch) was meant “to assess the vulnerability of a nonbank financial company to financial distress.” The distinction was clear: FSOC intended the second group of analytical categories to assess a company before it became distressed and the first group to assess the impact of such distress on national financial stability. Yet in the Final Determination, FSOC posited that all six categories were meant only “to assess the potential effects of a company’s material financial distress.” That is undeniably inconsistent.[132]

But of course large, connected firms are unlikely to pose risk unless they are also highly leveraged or illiquid. And at any rate, the court never suggested that the Council had not considered the six factors, only that it had not done so in a way that could be expressed in two inquiries rather than one.

The Council was also faulted, a bit more plausibly, for failing to consider some relevant factors in making its decision about riskiness. The court expressed its view as follows:

[The Council] never projected what the losses would be, which financial institutions would have to actively manage their balance sheets, or how the market would destabilize as a result. . . . Predictive judgment must be based on reasoned predictions; a summary of exposures and assets is not a prediction.[133]

This sort of “you should have done more” second guessing is also easy to criticize. It is never clear how much more is enough, and suggesting that the Council’s procedures are arbitrary because it relied on data about exposure to conclude that MetLife was systemically risky—but that it failed to make an explicit prediction about systemic riskiness—holds the Council to an awfully high standard of connecting the dots.

As for the court’s argument that the Council’s “decision intentionally refused to consider the cost of regulation, a consideration that is essential to reasoned rulemaking,” the less said, the better.[134] Nothing about the APA requires a cost–benefit analysis, which means that the textual basis for requiring a cost–benefit analysis must be found in Dodd-Frank itself.[135] The portion of the statute that the court concluded required the agency to consider cost only invited the Council to use “any other risk-related factors that the Council deems appropriate” in making its designation decisions, an invitation that obviously includes no required factors, including cost.[136]

Although the court’s findings of procedural errors were unpersuasive as a matter of logic, the point is that the critique was based on the idea that the Council, regardless of the decision it made, failed to follow the processes that it and Congress had laid out. The court never suggested that MetLife did or did not present a risk to the financial system. It instead evaluated the designation according to the standards the Council had announced for itself, and found them wanting. While other judges might have been more procedurally lenient towards the Council, none of them got to evaluate MetLife’s case. The Council filed an appeal to the decision during the Obama Administration, only to drop the suit and acquiesce to de-designation during the Trump Administration.[137]

3. Deregulation by Congress.—Congress has not successfully passed any legislation that would remove any regulation of the insurance industry, but it has made noises about doing so. The House Financial Services Committee complained in a report that “FSOC’s nonbank designation process is arbitrary and inconsistent.”[138] The committee complained, among other things, that FSOC had failed to show that material financial distress at its insurance companies might have ripple effects, by causing “impairment of financial intermediation or of financial market functioning that would be sufficiently severe to inflict significant damage on the broader economy.”[139] But this oversight has not resulted in passed legislation that would undo the Council’s power to designate insurance companies as systemically risky. The House has tried; the Financial CHOICE Act, which it passed, would completely eliminate FSOC’s designation power.[140] But that statute, though blunt, has not gotten beyond the lower house. Instead, Congress passed a modest deregulatory bill that did nothing about insurance supervision (it modified, among other things, the Volcker Rule’s impact on other financial institutions, and reduced the number of banks subject to so-called “stress tests”).[141] The result is that Congress’s role in federal insurance deregulation has been all or nothing—and thus far, it has been nothing that has been the result.

Perhaps this did not have to be the case—the deregulatory Economic Growth, Regulatory Relief, and Consumer Protection Act of 2018 chipped away at the Dodd-Frank regulatory requirements, but made no attempt to fundamentally undo them (and did little about insurance regulation other than to encourage transparency in international insurance regulatory standard setting).[142] But Congress is limited on the substantive work it can do in deregulation by some constraints of the regulatory process. It must legislate prospectively, rather than retrospectively, and it must regulate classes, rather than individual firms, as the latter is left to the judiciary, and it is comprised of generalists, rather than experts—few members of Congress would claim expertise in matters of systemic risk.[143]

4. Conclusion.—What we see with deregulatory courts and a deregulatory Council is a traditional and important distinction between substance and procedure in administrative law. The Council has regulated the insurance industry by singling out for extra attention some institutions that posed risk to the final system. In deciding that these risks were no longer worth the extra attention, the Council considered how big the institutions were or had become, how seriously they had taken the effort to become less risky, and how interconnected the institution was once it was shrunken. It worried most about this interconnectedness because doing so is a textbook form of financial regulation, which is designed to forestall the type of contagion that really exists only in finance (manufacturing, for example, does not have this sort of risk; just because Chrysler goes bankrupt doesn’t mean that Ford and GM will as well).[144] In the last financial crisis, every investment bank in the United States looked set to collapse without a government intervention after the collapse of one of them—Lehman Brothers; by the end of the crisis, only two of the five largest in the country were left, and they had converted themselves into bank holding companies, exchanging the oversight of the SEC for that of the Fed, which has the power to make emergency loans to banks facing a run.[145]

The court’s reasoning, by contrast, had nothing to do with the question of whether insurance companies posed threats of systemic risk. Instead, it bespoke a traditional focus on procedure and a sort of second guessing about whether the procedures identified by the agency actually applied in the way the agency promised. While one form of deregulation focuses on the riskiness of financial institutions, the other focuses on a decision-making process that went into that riskiness determination.

This makes the Council’s form of deregulation substantive and much more satisfying than judicial deregulation, which said nothing about whether MetLife was actually risky or not but only critiqued the Council for making the decision it did.

This record is a retreat from Dodd-Frank’s clear desire to put the Council in an oversight position over nonbank financial companies. That statute’s creation of the FIO also suggests the desire for a federal role. Nonetheless, new regulatory schemes by definition create new burdens. While new executives and regulators cannot abandon oversight over industries that Congress has insisted be regulated, they can certainly turn down invitations to regulate from the Legislature.

Congress gave the Council the power to oversee systemically risky insurance companies, but it did not mandate a federal insurance regulation scheme. Perhaps unsurprisingly, then, an administration with a deregulatory bent has turned away from federal insurance oversight, preferring to leave such matters to the states.

III. International Regulation’s Role

The remaining federal role in insurance regulation will be provided by the part of the Treasury Department made responsible for negotiating international agreements on insurance. While the government is getting out of the business of the direct regulation of large insurers, regulatory cooperation orchestrated by the FIO has managed to keep the federal government involved in oversight. This indirect regulation has been based on international, rather than domestic policymaking. It is rooted in a deal between federal and European insurance regulators. The deal requires American regulators to remove some state requirements on insurance and reinsurance companies in exchange for similar EU concessions. It also encourages (and may ultimately require) state regulators to adopt a form of supervision used in the European Union for all U.S. insurance companies doing business in the EU, which includes all large U.S. insurers, making the regulatory impact of international agreements substantial.

The so-called covered agreement concluded with the European Union, which the Obama Administration negotiated and the Trump Administration signed, illustrates how the growing internationalization of insurance (or almost anything, really) can impel American regulators, even those of a deregulatory bent, to engage with their foreign counterparts.

It is perhaps enough to say that the covered agreement gives the federal government a regulatory role in the supervision of insurance companies through the operation of regulatory cooperation with foreign jurisdictions. It is a testament to the global nature of the insurance market and the way that all forms of finance cross borders. It has also meant that the FIO’s diplomatic role has been the statutory basis for the continuation of federal involvement in insurance oversight. It also is technical—the FIO is involved in insurance regulation not because of a sweeping deal on fundamental principles on insurance oversight, but rather on a give-and-take on reinsurance rules, and a concession to the EU on the appropriate way to supervise big insurance firms, a way that could revolutionize insurance supervision in the United States. But for aficionados, it is worth spelling out exactly how the covered agreement does so and how the FIO was charged with the power to negotiate it. Those less concerned with the intricacies of insurance supervision may find these details unnecessary for an understanding of the larger point.

The conclusion of the covered agreement raises some of the usual concerns about international financial regulation—is the cooperation by agencies sufficiently responsive to outside input, given that notice and comment implementing an international agreement can look like a fait accompli?[146] Time will tell, but it is clear that international financial regulation is hard to resist and tends to entangle—the covered agreement includes a provision for a Joint Committee of American and European regulators who will oversee its adoption and implementation over the next five years, meaning that international cooperation is more than a deal—it also creates cross-border institutions.[147]

A. Statutory Authority

Title V of the Dodd-Frank Wall Street Reform Act created the FIO within the Department of the Treasury.[148] That office has limited powers, especially domestically, where insurance supervision remains the province of the state insurance commissions.[149] FIO has nonetheless been charged with a particularly important outward-facing role. Congress instructed it “to coordinate Federal efforts and develop Federal policy on prudential aspects of international insurance matters, including representing the United States, as appropriate, in the International Association of Insurance Supervisors”; to “consult with the States (including State insurance regulators) regarding insurance matters of national importance and prudential insurance matters of international importance”; and to “advise the [Treasury] Secretary on . . . prudential international insurance policy issues.”[150]

It also has been given the power, in association with the United States Trade Representative, to conclude agreements on insurance regulation with foreign counterparties. These so-called covered agreements are defined in Dodd-Frank as follows:

a written bilateral or multilateral agreement regarding prudential measures with respect to the business of insurance or reinsurance that—

(A) is entered into between the United States and one or more foreign governments, authorities, or regulatory entities; and

(B) relates to the recognition of prudential measures with respect to the business of insurance or reinsurance that achieves a level of protection for insurance or reinsurance consumers that is substantially equivalent to the level of protection achieved under State insurance or reinsurance regulation.[151]

Covered agreements are meant to both strengthen insurance regulation and level the playing field between the United States and other countries. One of the problems for insurers who want to expand their operations abroad has been navigating the regulatory burdens posed by the government of every attractive market.[152] The efforts to liberalize this sort of trade—trade in services—has traditionally been thought to be less effective than the liberalization of trade in goods because of these regulatory barriers to entry.[153] Insurance companies seeking to do business abroad need to comply with local licensing requirements, rules about capital that must be retained to insure their solvency, and consumer-protection requirements relating to the kinds of insurance that can be sold and the basis for the denials of claims— all of these can be used to protect incumbent insurers, and the burden would be on foreign competitors.

Bilateral agreements are also meant to serve as a backstop for regulatory cooperation in cases where—as is the case with insurance regulation in particular—multilateral governance has not made progress. An analogy might be drawn to this country’s approach to progress on reducing barriers to trade. When multilateral trade negotiations like the Doha Round of the World Trade Organization have foundered, the United States has increasingly looked to pursue its trade interests through bilateral trade and investment deals.[154] In the case of post-crisis insurance supervision, the options offered in Dodd-Frank suggest that where multilateral efforts either to level the international playing field or to improve the supervision of systemically risky insurance companies have foundered, bilateral covered agreements might serve as a useful supplement.

B. The Covered Agreement

The covered agreement concluded by the FIO at the end of the Obama Administration gives the federal government a role in regulating insurance even as it has retreated from its SIFI designation role. The Trump Administration, after some hemming and hawing, announced on July 14, 2017, that the Treasury Secretary would sign the covered agreement; he duly did so September 22, 2017.[155]

In particular, the covered agreement creates a federal role for the capital rules for reinsurers and for the supervision of insurance companies, as a supervisor of what kind of rules the states can impose on insurers. In reinsurance, the agreement reduces regulatory barriers to foreign competition in the United States and EU. Its group-supervision principles, in contrast, harmonize the regulatory approaches of the supervision of large insurance companies’ operations in both jurisdictions. The agreement also includes an information-exchange component designed, among other things, to deepen regulatory ties between American and European insurance supervisors. The agreement thus sets regulatory parameters for the EU and U.S. insurance industries and requires the FIO’s monitoring and oversight of the implementation of the agreement in the United States. It means that while the federal regulators are marching away from the direct supervision of insuring through the Council and the Fed, the elimination of federal regulation of insurance has not been complete because of the role it will now have insuring that state regulators meet the terms of the covered agreement. Instead, through international agreement, the federal government will find that it must continue to supply regulatory oversight to the insurance industry.

As for reinsurance, the covered agreement is best understood as an effort to reduce regulatory barriers to foreign competition in the United States and EU. The agreement serves to remove posted collateral and local presence requirements for EU and U.S. reinsurers doing business across the Atlantic.[156] The reinsurance portion of the agreement thus reduces trade barriers in both the United States and the European Union in a way likely to benefit American consumers. It is something like a trade deal, contained within the more narrow confines of a limited agreement on international insurance regulation.[157] In particular, the requirement that foreign reinsurance firms post 100% collateral to do business in certain American jurisdictions makes little sense for well-supervised European reinsurers. This problem has been apparent for years, yet any reduction in the collateral requirements, which thereby would open up the U.S. reinsurance market and introduce new competitors, to the benefit of insurance companies and ultimately consumers, has been slow.

The agreement thus would prevent U.S. state insurance regulators from requiring EU reinsurers to post such high levels of collateral as a condition for U.S. firms to be credited for their contracts with EU reinsurers.

The United States also got something for American reinsurance companies. One of the covered agreement’s objectives, as announced in its Article I, is “the elimination, under specified conditions, of local presence requirements.”[158] Specifically, the agreement relieves U.S. reinsurers from the obligation to establish a local presence—i.e., a branch or subsidiary—in the EU. The local presence requirement in the EU was also a real burden on the ability of American reinsurers to access that market. The elimination of that burden should level the playing field for American and European reinsurance firms by making it easier for American reinsurers to access the European market without opening an office in every jurisdiction in which they do business.[159]

The agreement also contains provisions on group supervision. Under the EU’s “Solvency II” regime, European insurers are subject to group supervision, and foreign insurers seeking to do business in the EU are required to establish that they are supervised in a comparable way.[160] Most worryingly for American firms, the EU reserved for itself the right to impose additional capital and other regulatory requirements on firms based in countries that were not determined by the EU to have a supervisory system that is “equivalent” to the Solvency II supervisory system.[161]

The covered agreement provides that this requirement will not be imposed upon American insurers doing business in Europe, provided that they can establish that they are being adequately supervised as groups. The “consolidated” form of supervision assesses the solvency and soundness of insurance firms with reference to all of their subsidiaries; in the United States, solvency is traditionally assessed at the subsidiary, or operating entity, level, on a state-by-state basis, so that each state regulatory authority monitors the solvency of each insurance company subsidiary doing business in that state.[162] The agreement was in this way designed to “establish[] that the [American] supervisory authority, and not the [European] supervisory authority, will exercise worldwide prudential insurance group supervision,” as the agreement provides in Article I.[163] It means that U.S. insurance groups operating in the EU will be supervised at the worldwide group level by the relevant U.S. insurance supervisors rather than through a European process imposed on American insurers and based on Solvency II.

Group supervision is the appropriate way to supervise any large financial conglomerate; one of the lessons of the bailout of AIG concerned the downside of supervision on an entity-by-entity basis, for regulators did not realize that AIG’s credit default swaps business, to say nothing of its securities lending business, was in the province of local supervision. Banks are supervised at the holding-company level by the Federal Reserve, and the single-point-of-entry resolution scheme also looks to manage firms in crisis in a consolidated way.[164] Dodd-Frank, in the way it treated nonbank subsidiaries of broker-dealers and derivatives desks, also looked to the group rather than the operating subsidiary in assessing systemic risk.[165] The group-supervision component of the covered agreement brings this sort of focus to insurance conglomerates, and appropriately so.

Finally, the agreement provides for an information exchange that will amplify and improve contacts between regulators in the United States and EU.[166] Over four decades of “cooperation among central bankers and securities regulators has contributed to the capacity for the coordinated response that we have seen, to the degree that we have seen it,” in the response to the last crisis, by both.[167] In the midst of that crisis, “the SEC coordinated its shorting ban with its international counterparts at an IOSCO [International Organization of Securities Commissions] meeting, even though the coordination was done in the hallways rather than during the official session.”[168] By the same token, the coordination of the injections of capital through swap lines and other mechanisms by the world’s central bankers was facilitated by their already extant supervisory cooperation.[169] In other words, cooperation on matters of enforcement and understandings along those lines can create or further the relationships that can facilitate an international response to the next crisis.[170] That precedent suggests that the agreement on information exchange is a worthy and useful aspect of the agreement.


The deregulation of the federal oversight of insurance has been achieved but only mostly. In insurance at least, agencies have more carefully stuck to their substantive tasks in ordering deregulation than have the courts or Congress.

Deregulation is not the only story, however. Because the federal government committed itself to a deal on standards with the European Union on three particular aspects of insurance supervision, the federal role in insurance regulation is reduced but not entirely retired. In the increasingly globalized world, deregulation is likely to be complicated by the international commitments that are part of the modern regulatory toolkit. It all makes deregulation something easier said than done, but if you want to see how to do it, the way that regulators have gotten out of the business of the direct regulation of insurers offers insights on the comparative advantages and disadvantages of the country’s would-be deregulators.

  1. *Associate Professor, The Wharton School. Thanks to Peter Conti-Brown, Jeremy Kress, and Dan Schwarcz, and for comments received at presentations at Hastings Law School and Wharton.
  2. .See Exec. Order No. 13,772, 82 Fed. Reg. 9965 (Feb. 3, 2017) (directing Treasury Secretary to review financial regulations pursuant to principles, including to “make regulation efficient, effective, and appropriately tailored”); Exec. Order 13,771, 82 Fed. Reg. 9339 (Jan. 30, 2017) (directing agencies to repeal at least two existing regulations before issuing a new regulation and requiring “total incremental cost of all new regulations, including repealed regulations [for 2017] . . . shall be no greater than zero”).
  3. .Thomas O. McGarity, The Congressional Review Act: A Damage Assessment, Am. Prospect (Feb. 6, 2018), [] (“[T]he vast majority of bills enacted during his first six months in office stemmed from the Congressional Review Act of 1996 (CRA).”).
  4. .See, e.g., Exec. Order No. 13,798, 82 Fed. Reg. 21,675 (May 4, 2017) (directing the Treasury, Labor, and Health and Human Services Secretaries to roll back Obama Administration rules concerning contraception coverage under the Affordable Care Act); Press Release, Envtl. Prot. Agency, EPA Takes Another Step to Advance President Trump’s America First Strategy, Proposes Repeal of “Clean Power Plan” (Oct. 10, 2017), [] (proposing the repeal of the “Clean Power Plan”); U.S. Dep’t of Treasury, A Financial System that Creates Economic Opportunities: Capital Markets 132, 180, 205, 218 (2017), [] (proposing repeal or reconsideration of several Dodd-Frank era financial regulations).
  5. .Renae Merle, Trump Administration Strips Consumer Watchdog Office of Enforcement Powers in Lending Discrimination Cases, Wash. Post (Feb. 1, 2018), [] (“The Trump administration has stripped enforcement powers from a Consumer Financial Protection Bureau unit responsible for pursuing discrimination cases, part of a broader effort to reshape an agency it criticized as acting too aggressively.”); email from Mick Mulvaney, Acting Dir., Consumer Fin. Prot. Bureau, to _DL_CFPB_AllHands (Jan. 23, 2018, 12:59 CST), [] (“[T]he days of aggressively ‘pushing the envelope’ of the law in the name of the ‘mission’ are over.”). More generally, “[t]he steady pace of deregulation has shown how much can be done at the agency without legislative action, and more change is expected in the year ahead as agencies fill out their complement of Trump appointees.” Richard Satran, With Congress Gridlocked, U.S. Finance Regulators Roll Back Dodd-Frank, Thomson Reuters Reg. Intelligence (Jan. 24, 2018).
  6. .See, e.g., David J. Reiss, The Trump Administration and Residential Real Estate Finance, 23 No. 20 Westlaw J. Sec. Lit. & Reg. 02 (describing “a series of Treasury reports — titled ‘A Financial System That Creates Economic Opportunities’ — that . . . offer a glimpse into how this administration intends to regulate — or more properly, deregulate” financial markets).
  7. .The President, for example, sent the Secretary of the Treasury a memorandum ordering him to vote against any further federal efforts to supervise insurance companies. Memorandum on the Financial Stability Oversight Council, Daily Comp. Pres. Doc. 2 (Apr. 21, 2017), [] (instructing the Treasury Secretary in § 3, “consistent with law,” to vote against any federal effort to increase supervision over a particular insurance company).
  8. .There is no constitutional constraint against a federal role in insurance regulation, as the Supreme Court held in United States v. South-Eastern Underwriters Ass’n, 322 U.S. 533, 561–62 (1944) (holding that the Sherman Act, an antitrust statute, could be applied to insurance). But in the next year, Congress passed a statute precluding the government from regulating the business of insurance. McCarran-Ferguson Act, ch. 20, 59 Stat. 33 (1945) (codified as amended at 15 U.S.C. § 1012 (2012)). For a discussion of AIG’s problems, see Daniel Schwarcz & David Zaring, Regulation by Threat: Dodd-Frank and the Nonbank Problem, 84 U. Chi. L. Rev. 1813, 1824–30 (2017).
  9. .Dodd-Frank Wall Street Reform and Consumer Protection Act, Pub. L. No. 111-203, § 113, 124 Stat. 1376, 1398 (2010) (codified at 12 U.S.C. § 5323 (2012)) (providing this power, which covered not just insurance companies but all “nonbank financial companies”).
  10. .The FIO also has the power to monitor and collect data on the insurance industry in an effort to determine whether insurers were engaging in risky conduct that could jeopardize their solvency. This data collection activity gives the office a research function as well; in 2016, it issued a report on the effectiveness of terrorism-insurance research. See U.S. Dep’t of Treasury, Report on the Overall Effectiveness of the Terrorism Risk Insurance Program 6–7 (2016), [] (explaining the data collection processes under the 2015 Reauthorization Act).
  11. .Press Release, U.S. Dep’t of Treasury, Treasury, USTR Sign Covered Agreement on Prudential Insurance & Reinsurance Measures with the European Union (Sept. 22, 2017), [].
  12. .Oliver Milman, ‘Sloppy And Careless’: Courts Call Out Trump Blitzkrieg on Environmental Rules, Guardian (Feb. 20, 2018), [].
  13. .See Gillian E. Metzger, Foreword: 1930s Redux: The Administrative State Under Siege, 131 Harv. L. Rev. 1, 2 (2017) (discussing the way Congress’s use of the CRA meant that it has “eagerly repealed numerous regulations”).
  14. .See, e.g., Clean Air Council v. Pruitt, 862 F.3d 1, 4 (D.C. Cir. 2017) (concluding that “EPA lacked authority under the Clean Air Act to stay” a rule on greenhouse gas emissions and ordering the agency to implement the rule).
  15. .Chevron U.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 842–43 (1984), which held that courts should afford deference to any reasonable agency interpretation of its governing statute, was itself a deregulation case—in it, the Reagan Environmental Protection Agency reinterpreted a Carter EPA rule to reduce burdens on, among others, oil refiners.
  16. .Cass R. Sunstein, Interest Groups in American Public Law, 38 Stan. L. Rev. 29, 74 (1985).
  17. .See Perez v. Mortg. Bankers Ass’n, 135 S. Ct. 1199, 1206 (2015) (noting that the APA makes no distinction between initial and subsequent agency action).
  18. .See U.S. Dep’t of Agric. v. Moreno, 413 U.S. 528, 534–35 (1973) (explaining that when Congress regulates, it “must rationally further some legitimate governmental interest”). But see Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 59 (1983) (Rehnquist, J., concurring in part and dissenting in part) (arguing that different policy preferences by a new administration’s regulators could be a reasonable basis for a change in policy).
  19. .See FCC v. Fox Television Stations, Inc., 556 U.S. 502, 515 (2009) (“The statute makes no distinction, however, between initial agency action and subsequent agency action undoing or revising that action. . . . And of course the agency must show that there are good reasons for the new policy.”); Clean Air Council v. Pruitt, 862 F.3d 1, 9, 14 (D.C. Cir. 2017) (rejecting effort by EPA to stay a rule regulating methane emissions scheduled to go into effect because of the lack of process).
  20. .As the Court has said, “[A]n agency’s decision not to take enforcement action should be presumed immune from judicial review . . . .” Heckler v. Chaney, 470 U.S. 821, 832 (1985).
  21. .As Daniel Deacon has observed, “All else being equal, a deregulatory presidential administration would prefer to proceed by rulemaking or legislation because these forms of policymaking exhibit a ‘stickiness,’ or obduracy, that enforcement practices do not.” Daniel T. Deacon, Deregulation Through Nonenforcement, 85 N.Y.U. L. Rev. 795, 796–97 (2010).
  22. .Symposium, Exit and the Administrative State, 67 Duke L.J. 1615 (2018).
  23. .J.B. Ruhl & James Salzman, Regulatory Exit, 68 Vand. L. Rev. 1295, 1299 (2015). See Justin R. Pidot, Governance and Uncertainty, 37 Cardozo L. Rev. 113, 122–23 (2015) (“[Regulatory] exit is simply a form of radical legal evolution that occurs when lawmakers respond to new information or changing circumstances that demonstrate that a legal regime is beyond saving, or perhaps, that the targeted problem has been solved.”). But see Sarah E. Light, Regulatory Horcruxes, 67 Duke L.J. 1647, 1649 (2018) (noting that notwithstanding any efforts by regulators to build a formal exit strategy into a regulatory program ex ante, background legal rules like the Administrative Procedure Act permit successor administrations to exit regulatory programs).
  24. .Dodd-Frank Wall Street Reform and Consumer Protection Act, Pub. L. No. 111-203, § 113, 124 Stat. 1376, 1401 (2010) (codified at 12 U.S.C. § 5323 (2012)). For a discussion on the mechanisms by which § 113 of Dodd-Frank designates SIFIs, see Joshua S. Wan, Note, Systemically Important Asset Managers: Perspectives on Dodd-Frank’s Systemic Designation Mechanism, 116 Colum. L. Rev. 805, 813–16 (2016).
  25. .See Fin. Stability Oversight Council, Minutes of the Financial Stability Oversight Council 4 (2015),,%202015-Minutes.pdf [] (describing the Council’s annual review of designated firms); see also Fin. Stability Oversight Council, Notice and Explanation of the Basis for the Financial Stability Oversight Council’s Rescission of Its Determination Regarding American International Group, Inc. (AIG) 2 (2017),,_Inc._(Rescission).pdf [] (explaining the decision to de-designate AIG); Fin. Stability Oversight Council, Basis for the Financial Stability Oversight Council’s Rescission of Its Determination Regarding GE Capital Global Holdings, LLC 2 (2016), [

    -ADBT] (explaining the decision to de-designate GE Capital).

  26. .Hannah J. Wiseman, Regulatory Triage in a Volatile Political Era, 117 Colum. L. Rev. Online 240, 250 (2017); see also Sarah E. Light, Precautionary Federalism and the Sharing Economy, 66 Emory L.J. 333, 344 (2017) (“[S]cholars and policymakers should grapple more actively with the bases for a shift in the allocation of regulatory authority.”).
  27. .See Ryan Tracy, Washington’s $500 Million Financial-Storm Forecaster Is Foundering, Wall St. J. (Feb. 19, 2018), [] (describing the Trump Administration’s plans to reduce headcount and budget, noting that no official had been nominated to permanently head the office). The office is involved because its director participates in Council decisions as a nonvoting member. Fin. Stability Oversight Council, Who Is on the Council?, U.S. Dep’t of Treasury (Sept. 19, 2018), [].
  28. .See Michael P. Van Alstine, The Costs of Legal Change, 49 UCLA L. Rev. 789, 807 (2002) (describing statutory deregulation as the “most common” form of deregulation and the jurisprudential response to the “excesses of modern public law” responsible for lifting regulatory oversight over large industries).
  29. .Id.
  30. .The nondelegation doctrine, which polices congressional intrusions into the responsibilities of the other branches, has been successfully invoked in the Supreme Court only in 1935; it has had, as Cass Sunstein has said, “[O]ne good year, and 211 bad ones (and counting).” Cass Sunstein, Nondelegation Canons, 67 U. Chi. L. Rev. 315, 322 (2000). As for rational review, as the Ninth Circuit has put it, “the legislature’s decision to remove certain [] requirements that it no longer deems essential . . . is a rational and quintessentially legislative decision.” Merrifield v. Lockyer, 547 F.3d 978, 990 (9th Cir. 2008); see also Merrick B. Garland, Deregulation and Judicial Review, 98 Harv. L. Rev. 505, 532 (1985) (discussing judicial review of agency decision-making, which, as with congressional legislation, evaluates an agency decision under the rational basis test and accords it “great deference”).
  31. .H.R. 5, 115th Cong. (2017), [].
  32. .H.R. 4768, 114th Cong. (2016), [].
  33. .H.R. 10, 112th Cong. (2011), [].
  34. .5 U.S.C. §§ 801–08 (2012).
  35. .See Michael D. Shear, Trump Discards Obama Legacy, One Rule at a Time, N.Y. Times (May 1, 2017), [] (providing examples of administrative regulations that President Trump and Congress have reversed).
  36. .See 5 U.S.C. §§ 801(a)(3)(B), 802 (describing the procedural requirements for congressional disapproval).
  37. .Id. § 801(b)(2).
  38. .See id. § 801(d)(1) (providing a window of sixty session days for the Senate and sixty legislative days for the House); David Zaring, Guidance and the Congressional Review Act, Reg. Rev. (Feb. 15, 2018), [] (providing an example of congressional efforts to use the CRA to reverse Obama Administration regulations).
  39. .See CRA Resolutions, Rules at Risk, [] (listing the resolutions of disapproval and their dates).
  40. .See Pub. L. No. 115-5, 131 Stat. 10 (2017) (disapproving the stream protection rule).
  41. .See Pub. L. No. 115-4, 131 Stat. 9 (2017) (disapproving the resource extraction rule).
  42. .To see the full collection of rules, see CRA Resolutions, supra note 38.
  43. .For further discussion of the Bush Administration’s rescission of the Clinton Administration’s rule on ergonomics, see Nina A. Mendelson, Agency Burrowing: Entrenching Policies and Personnel Before a New President Arrives, 78 N.Y.U. L. Rev. 557, 638 (2003) and Note, The Mysteries of the Congressional Review Act, 122 Harv. L. Rev. 2162, 2172 (2009).
  44. .As Michael Gerhardt has explained, “[T]he Senate has adopted formal rules for its internal governance, including Senate Rule XXII, which expressly authorizes filibusters or protracted debate to delay or obstruct legislative action.” Michael J. Gerhardt, Dissent in the Senate, 127 Yale L.J. F. 728, 733 (2018).
  45. .Van Alstine, supra note 27, at 807–08. This can of course be exacerbated where there is a “vague judicial hostility to regulatory legislation.” Note, Intent, Clear Statements, and the Common Law: Statutory Interpretation in the Supreme Court, 95 Harv. L. Rev. 892, 911 (1982); see also Richard J. Pierce & Sidney A. Shapiro, Political and Judicial Review of Agency Action, 59 Texas L. Rev. 1175, 1206 (1981) (“The decision to enforce the delegation doctrine strictly and reject congressional decisions to adopt regulatory schemes without spelling out all details beforehand would effectuate a judicial deregulation of the economy.”).
  46. .George B. Shepherd, Fierce Compromise: The Administrative Procedure Act Emerges from New Deal Politics, 90 Nw. U. L. Rev. 1557, 1627 (1996) (stating that the “two groups that agencies harmed most, lawyers and regulated businesses,” pushed for judicial review of agency action through a codified APA).
  47. .David Zaring, Reasonable Agencies, 96 Va. L. Rev. 135, 137, 140–41 (2010) (reporting on a number of studies that show that courts reverse agencies approximately one-third of the time).
  48. .See id. at 170–76 (discussing how often agencies win their cases).
  49. .Patricia Wald, The Contribution of the D.C. Circuit to Administrative Law, Address Before the American Bar Association Section of Administrative Law (October 1987), in 40 Admin. L. Rev. 507, 528 (1988) (identifying an inadequate agency rationale as the reason for fifty-eight reversals or remands out of a total of 159 opinions issued by the D.C. Circuit for one year in the late 1980s). For further discussion of judicial review of agency rulemaking, see Richard J. Pierce, Jr., Seven Ways to Deossify Agency Rulemaking, 47 Admin. L. Rev. 59, 95 (1995).
  50. .3 Charles H. Koch, Jr., Administrative Law and Practice § 8:31[2](a) (3d ed. 2010) (observing the typicality of this remedy, “including remand for further proceedings, remand with instruction, and reversal and remand”).
  51. .Parrillo concluded that “even though contempt findings are practically devoid of sanctions, they nonetheless have a shaming effect that gives them substantial if imperfect deterrent power.” Nicholas R. Parrillo, The Endgame of Administrative Law: Governmental Disobedience and the Judicial Contempt Power, 131 Harv. L. Rev. 685, 697 (2018).
  52. .See, e.g., Motor Vehicles Mfrs. Ass’n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 55 (1983) (stating that “[t]he agency . . . failed to articulate a basis for not requiring nondetachable belts under Standard 208”).
  53. .As the Fed has explained in a proposed rule for the consolidated supervision of systemically important insurers, “the consolidated approach would categorize an entire insurance firm’s assets and insurance liabilities into risk segments, apply appropriate risk factors to each segment at the consolidated level, and then set a minimum ratio of required capital.” Press Release, Federal Reserve, Federal Reserve Board Approves Advance Notice of Proposed Rulemaking and Approves Proposed Rule (June 3, 2016), [].
  54. .The Treasury has suggested that the Council raise the stage 1 asset number to $250 billion. Jeff Bater, Mnuchin: Bank ‘Systemic Risk’ Label Should Start at $250 Billion, Bloomberg BNA (July 27, 2017), [].
  55. .For a general overview of the Act, see House Fin. Servs. Comm., The Financial CHOICE Act: Executive Summary, []. To view the Act’s full text and details of its progression through the House and Senate, see H.R. 10 – Financial CHOICE Act of 2017,, [].
  56. .The Council de-designated the last one, Prudential, by an order dated October 16, 2018. See infra section II(C )(1); see also Jesse Hamilton, Prudential Is Plotting Its Escape from Fed’s Tough Oversight, Bloomberg (Aug. 17, 2017), [] (describing the company’s de-designation strategy).
  57. .See Fin. Stability Oversight Council, Designations, U.S. Dep’t of Treasury

    (July 18, 2018), [] (listing designations and the basis for those designations).
  58. .MetLife, Inc. v. Fin. Stability Oversight Council, 177 F. Supp. 3d 219, 242 (D.D.C. 2016).
  59. .Douglas J. Elliott, Regulating Systemically Important Financial Institutions that Are Not Banks, Brookings Reports (May 9, 2013), [].
  60. .See infra section II(C )(1).
  61. .Id.
  62. .Office of Fin. Research, Asset Management and Financial Stability 19 (2013) (concluding that asset managers had suffered “material distress” during the financial crisis); Cary Martin Shelby, Closing the Hedge Fund Loophole: The SEC as the Primary Regulator of Systemic Risk, 58 B.C. L. Rev. 639, 674–75 (2017) (noting that the OFR study “was severely criticized by a large number of industry participants” and by the SEC commissioner).
  63. .Daniel Schwarcz, A Critical Take on Group Regulation of Insurers in the United States, 5 U.C. Irvine L. Rev. 537, 551 (2015). For a further discussion of how AIG collapsed, and what happened afterwards, see David Zaring, Litigating the Financial Crisis, 100 Va. L. Rev. 1405, 1427–30 (2014).
  64. .Schwarcz, supra note 62, at 551.
  65. .For further discussion of AIG securities lending, see id. at 551–55.
  66. .Schwarcz & Zaring, supra note 7, at 1827–28.
  67. .See id. at 1825, 1827–28 (describing the unforeseen risk in two of AIG’s pre-bailout activities).
  68. .Schwarcz & Zaring, supra note 7, at 1824–30, 1841.
  69. .See Int’l Monetary Fund, Global Financial Stability Report: Potent Policies for a Successful Normalization 101–02 (2016), [] (describing the reduction in risky assets by both large and small American insurers).
  70. .Schwarcz & Zaring, supra note 7, at 1817–18.
  71. .This pattern—the threat of designation followed by improved regulation—worked to improve the Securities and Exchange Commission’s regulation of money market funds. See generally Sec. & Exch. Comm’n, Money Market Fund Reform 22–23, 26–27, 29–31; Amendments to Form PF, 79 Fed. Reg. 47,736 (2014), amending 17 C.F.R. Parts 230, 239, 270, 274, 279; see Press Release, Sec. & Exch. Comm’n, SEC Adopts Money Market Fund Reform Rules (July 23, 2014), [] (describing amendments to the rules that govern money market funds).
  72. .McCarran-Ferguson Act, ch. 20, 59 Stat. 34 (1945) (codified as amended at 15 U.S.C. § 1012 (2012)).
  73. .See, e.g., Letter from Nat’l Conference of Ins. Legislators, to Barney Frank, Chairman, House Fin. Servs. Comm. (Oct. 23, 2009), [] (“We continue to disagree with the necessity for such an office and question its accountability and effectiveness.”). For a discussion of states’ reaction to and roles under Dodd-Frank, see Gillian E. Metzger, Federalism Under Obama, 53 Wm. & Mary L. Rev. 567, 581–87 (2011).
  74. .Indeed, the American Action Forum called it precisely this. Hotel California: FSOC Edition, Am. Action Forum (July 1, 2016), []. The reference is to a song by the Eagles, about a place from which “you can check out any time you like, but you can never leave!” The Eagles, Hotel California, on Hotel California (Asylum Records 1976).
  75. .For a discussion of the FSB generally, see David Zaring, Finding Legal Principle in Global Financial Regulation, 52 Va. J. Int’l L. 683, 698–99 (2012).
  76. .FSOC Accountability: Nonbank Designations: Hearing Before the Comm. on Banking, Hous. & Urban Affairs, 114th Cong. 1–2 (2015) (statement of Sen. Richard Shelby, Chairman, S. Comm. on Banking, Hous. & Urban Affairs).
  77. .Jeremy Kress, The Case Against Activity-Based Financial Regulation, CLS Blue Sky Blog (Nov. 16, 2017), [].
  78. .Schwarcz & Zaring, supra note 7, at 1869–81.
  79. .As Skinner has put it, “[A]lthough regulating nonbank financial institutions has, so far, proceeded in a black-and-white fashion—with all systemically important institutions regulated in a manner highly similar to banks—this approach is insufficiently attentive to the need for a more graduated, spectrum-like system.” Christina Parajon Skinner, Regulating Nonbanks: A Plan for SIFI Lite, 105 Geo. L.J. 1379, 1384 (2017).
  80. .See id. at 1396–97 (describing the “stark consequences of a binary situation” due to the “all-or-nothing” situation).
  81. .21 U.S.C. § 606(a) (2012).
  82. .Skinner, supra note 78, at 1385.
  83. .See id. at 1396 (describing FSOC declining to go forward with any new SIFI designations and instead commissioning a study on types of risks to be regulated).
  84. .See supra Part II.
  85. .See Metlife, Inc. v. Fin. Stability Oversight Council, 177 F. Supp. 3d 219, 239–42 (D.D.C. 2016) (analyzing the importance and necessity of using a cost–benefit analysis in the administrative process); see also Bus. Roundtable v. SEC, 647 F.3d 1144, 1149–51 (D.C. Cir. 2011) (determining that the agency “neglected its statutory obligation to assess the economic consequences of its rule”).
  86. .John C. Coates IV, Cost-Benefit Analysis of Financial Regulation: Case Studies and Implications, 124 Yale L.J. 882, 931 (2015). For example, financial institutions such as bank holding companies are subject to supervision by the Fed. 12 U.S.C. §§ 1841–52 (2012). They must comply with its rigorous capital standards, partly because of the risk the collapse of those firms can have on the larger economy. See Capital Adequacy of Bank Holding Companies, Savings and Loan Holding Companies, and State Member Banks (Regulation Q), 12 C.F.R. § 217.1 (2018) (setting out the minimum capital requirements and adequacy standards imposed on certain financial institutions). These requirements have been the centerpiece of financial regulation for decades, and the Fed changes the rules as time passes and new challenges are posed to the soundness of the financial system. The standards always impose costs on banks, but, when promulgating or amending the rules, the Fed has not tried to quantify the benefit in crises avoided, nor has it been asked to do so by the courts. See Henry T. C. Hu, Financial Innovation and Governance Mechanisms: The Evolution of Decoupling and Transparency, 70 Bus. Law 347, 404 (2015) (“[T]he Federal Reserve Board is generally not required to provide cost-benefit analysis with its rulemaking . . . .”).
  87. .Coates, supra note 85, at 999–1001.
  88. .Of these nine, eight are “real” regulators, and one is a voting member with insurance expertise. Fin. Stability Oversight Council, About FSOC, U.S. Dep’t of Treasury (Sept. 10, 2018), [] (providing a list of the nine regulators). The large number of federal financial regulators has, it is thought by many in the academy and elsewhere, made for too many financial overseers; the Treasury has proposed reducing the number of regulators in the past. The Department of the Treasury Blueprint for a Modernized Financial Regulatory Structure, U.S. Dep’t of Treasury 14 (Mar. 2008), [] (“In the optimal structure three distinct regulators would focus exclusively on financial institutions: a market stability regulator, a prudential financial regulator, and a business conduct regulator.”). For a discussion on alternative approaches to financial regulation, see Lawrence A. Cunningham & David Zaring, The Three or Four Approaches to Financial Regulation: A Cautionary Analysis Against Exuberance in Crisis Response, 78 Geo. Wash. L. Rev. 39, 45 (2009).
  89. .Press Release, N. Am. Sec. Adm’rs Ass’n, State Regulators Announce Representatives for the Financial Stability Oversight Council (Sept. 23, 2010), [].
  90. .Two federal bodies, the FIO and the Office of Financial Research, are also represented on the Council by their directors but do not vote on designations.
  91. .Scalia compared one balancing test to one that asked judges to divine “whether a particular line is longer than a particular rock is heavy.” Bendix Autolite Corp. v. Midwesco Enters., Inc., 486 U.S. 888, 897 (1988) (Scalia, J., concurring).
  92. .Dodd-Frank Wall Street Reform and Consumer Protection Act, Pub. L. No. 111-203, § 113(a)(2)(K), 124 Stat. 1376, 1398 (2010) (codified at 12 U.S.C. § 5323(a)(2)(K)).
  93. .Authority to Require Supervision and Regulation of Certain Nonbank Financial Companies, 12 C.F.R. § 1310 (2018).
  94. .Or so FSOC explained, “[A] nonbank financial company will be reviewed further if it has at least $50 billion in total consolidated assets . . . .” Fin. Stability Oversight Council, Nonbank Designations – FAQs, U.S. Dep’t Treasury (Nov. 21, 2016), [].
  95. .Id. In Stage 2:the FSOC notifies a company that comes under active review, and reviews existing public and regulatory information and information submitted by the company. If the Council decides to evaluate the company further, it notifies the company and begins Stage 3, a detailed, in-depth analysis that includes a review of confidential information provided by the company.Id.
  96. .MetLife, Inc. v. Fin. Stability Oversight Council, 177 F. Supp. 3d 219, 229 (D.D.C. 2016).
  97. .Id.
  98. .Fin. Stability Oversight Council, Basis for the Financial Stability Oversight Council’s Final Determination Regarding MetLife, Inc. 6 (2014), [].
  99. .Id. at 9.
  100. .Id. at 16.
  101. .See id. at 26 (describing the decentralized supervision of Metlife).
  102. .Fin. Stability Oversight Council, Basis for the Financial Stability Oversight Council’s Rescission of Its Determination Regarding GE Capital Global Holdings, LLC 2 (2016), [].
  103. .Id.
  104. .Id.
  105. .Id.
  106. .For a primer on how credit default swaps work, see William K. Sjostrom, Jr., The AIG Bailout, 66 Wash. & Lee L. Rev. 943, 947–52 (2009).
  107. .Fin. Stability Oversight Council, Notice and Explanation of the Basis for the Financial Stability Oversight Council’s Rescission of Its Determination Regarding American International Group, Inc. (AIG) 13 tbl.2 (2017),,_Inc._(Rescission).pdf [].
  108. .Id.
  109. .Id. at 5.
  110. .See id. at 13 tbl.2 (comparing the leverage ratio of AIG to the largest bank holding companies and insurers).
  111. .See id. at 17 (noting the decrease in the amount of outstanding single-name credit default swaps for which AIG is the reference entity). Of course, AIG’s financial crisis problems with CDS came from the CDS contracts it wrote rather than the ones referencing its own credit.
  112. .See id. at 13 tbl.2 (comparing the gross CDS outstanding for which bank holding companies and insurers are the reference entity).
  113. .Id. at 5.
  114. .Id.
  115. .Fin. Stability Oversight Council, Notice and Explanation of the Basis for the Financial Stability Oversight Council’s Rescission of Its Determination Regarding Prudential Financial, Inc. 2–9 (2018), [].
  116. .Id. at 11 tbl.1.
  117. .Id. at 61.
  118. .Id. at 14.
  119. .Id. at 32.
  120. .Id. at 5–6.
  121. .Id. at 7.
  122. .Id. at 11–12 tbls.1 & 2; id. at 62–63 apps. A & B.
  123. .Id. at 65 app. C.
  124. .See 5 U.S.C. §§ 701–06 (providing for judicial review of agency action).
  125. .Fin. Stability Oversight Council, Nonbank Designations – FAQs, U.S. Dep’t of Treasury (Nov. 21, 2016), [].
  126. .Id.
  127. .Alan S. Kaplinsky & Mark J. Levin, The Financial Stability Oversight Council: Another Potential Route for Overturning the CFPB’s Final Arbitration Rule, Ballard Spahr LLP (July 13, 2017), [].
  128. .See Eric A. Posner & Cass R. Sunstein, The Law of Other States, 59 Stan. L. Rev. 131, 140–43 (2006) (arguing that involving more voices in a decision improves the quality of the decision); see also Marquis de Condorcet, Essay on the Application of Mathematics to the Theory of Decision-Making (1785), reprinted in Condorcet: Selected Writings 33, 34–36 (Keith M. Baker ed., 1976).
  129. .MetLife, Inc. v. Fin. Stability Oversight Council, 177 F. Supp. 3d 219, 236–37 (D.D.C. 2016).
  130. .Id. at 230.
  131. .Id. at 239–40.
  132. .Id. at 234 (citations omitted).
  133. .Id. at 237.
  134. .Id. at 242.
  135. .The APA does not contain any language requiring a cost–benefit analysis, and the Supreme Court requires a clear statement from Congress. Am. Textile Mfrs. Inst., Inc. v. Donovan, 452 U.S. 490, 510 (1981) (“When Congress has intended that an agency engage in cost-benefit analysis, it has clearly indicated such intent on the face of the statute.”); Jennifer Nou, Regulating the Rulemakers: A Proposal for Deliberative Cost-Benefit Analysis, 26 Yale L. & Pol’y Rev. 601, 607 (2008) (describing the failed attempts to formally codify cost–benefit requirements in the APA). But see Cass R. Sunstein, Cost-Benefit Analysis and Arbitrariness Review, 41 Harv. Envtl. L. Rev. 1, 6 (2017) (“[A]ny decision not to quantify costs and benefits, or to show that the latter justify the former, does require some kind of explanation (at least if the governing statute does not rule cost-benefit analysis out of bounds).”).
  136. .12 U.S.C. § 5323(a)(2)(K) (2012); MetLife, 177 F. Supp. 3d at 242.
  137. .Pete Schroeder, MetLife, U.S. Regulators Agree to Set Aside Legal Fight, Thompson Reuters (Jan. 18, 2018), [].
  138. .Republican Staff of H.R. Fin. Servs. Comm., 115th Cong., The Arbitrary and Inconsistent FSOC Nonbank Designation Process 3 (Comm. Print 2017) [].
  139. .Id.
  140. .See House Fin. Servs. Comm., The Financial CHOICE Act: Executive Summary, [] (summarizing the proposed repeal of FSOC’s authority to designate firms as “systemically important”).
  141. .Economic Growth, Regulatory Relief, and Consumer Protection Act Pub. L. No. 115-174, §§ 203, 204, 401, 132 Stat. 1296, 1309–10, 1359 (2018).
  142. .Indeed, the most pro-deregulation members of Congress have indicated that they were unsatisfied with the 2018 statute and hoped for more in the future. Summary of the Economic Growth, Regulatory Relief and Consumer Protection Act (S. 2155), JDSupra (May 25, 2018), [].
  143. .For an analysis, see Michael Herz, The Legislative Veto in Times of Political Reversal: Chadha and the 104th Congress, 14 Const. Comment. 319, 336 (1997) (discussing cases where “congressional adjudication [would] so impermissibly intrude[] on judicial authority”).
  144. .After all, the Council’s mission is “to identify risks to the financial stability of the United States that could arise from the material financial distress or failure, or ongoing activities, of large, interconnected bank holding companies . . . .” 12 U.S.C. § 5322(a)(1)(A) (2012). For a discussion of the Council’s “sweeping” rulemaking authority, see Peter Conti-Brown, Elective Shareholder Liability, 64 Stan. L. Rev. 409, 434–35 (2012).
  145. .The history is reviewed in Steven M. Davidoff & David Zaring, Regulation by Deal: The Government’s Response to the Financial Crisis, 61 Admin. L. Rev. 463, 473–512 (2009).
  146. .For one look at the concerns raised along these lines, see David Zaring, Sovereignty Mismatch and the New Administrative Law, 91 Wash. U. L. Rev. 59, 82–84 (2013).
  147. .The Treasury Department has said that “the Joint Committee will serve as a forum for consultation and to exchange information on the administration and proper implementation of the Agreement.” U.S. Dep’t of Treasury, Statement of the United States on the Covered Agreement with the European Union 3 (Sept. 22, 2017), [].
  148. .Dodd-Frank Wall Street Reform and Consumer Protection Act, Pub. L. No. 111-203, §§ 501–02, 124 Stat. 1376, 1580 (2010) (codified at 31 U.S.C. § 313 (2012)).
  149. .So it has been since passage of the McCarran-Ferguson Act. That statute, as William Eskridge and John Ferejohn have explained, “exempt[s] insurance [from federal antitrust laws] and leav[es] that industry to state regulation.” William N. Eskridge, Jr. & John Ferejohn, Super-Statutes, 50 Duke L.J. 1215, 1233 n.72 (2001).
  150. .31 U.S.C. § 313(c) (2012). It has since directed the FIO to seek more transparency in these negotiations. Economic Growth, Regulatory Relief, and Consumer Protection Act Pub. L. No. 115-174, § 211, 132 Stat. 1296, 1317–19 (2018).
  151. .31 U.S.C. § 313(r)(2) (2012).
  152. .To take one example, a number of congressmen protested to the U.S. Trade Representative that insurance, among other industries, “face[s] serious market access barriers in Japan.” Letter from Max Baucus, Dave Camp, Orrin Hatch & Sander Levin to Ron Kirk, U.S. Trade Representative, Office of the U.S. Trade Representative (Nov. 8, 2011), [].
  153. .As Anupam Chander has observed, “International trade law has long recognized that internal regulations, not just border rules, might serve as barriers to trade in goods, but the even more extensive diffusion of regulatory authority over services heightens the challenge for discerning protectionist from other regulatory objects in services.” Anupam Chander, Trade 2.0, 34 Yale J. Int’l L. 281, 299 (2009); see Taunya L. McLarty, Liberalized Telecommunications Trade in the WTO: Implications for Universal Service Policy, 51 Fed. Comm. L.J. 1, 12 (1998) (“[T]rade liberalization in service markets was much further behind the liberalization of trade in goods . . . .”).
  154. .See Rafael Leal-Arcas, Proliferation of Regional Trade Agreements: Complementing or Supplanting Multilateralism?, 11 Chi. J. Int’l L. 597, 622 (2011) (“[W]orld leaders dropped a commitment to complete the troubled Doha Round in 2010 and vowed to push forward on bilateral and regional trade talks . . . .”).
  155. .For some background on adoption, see John S. Pruitt et al., Legal Alert: US-EU Covered Agreement—An Overview, Eversheds Sutherland (July 2, 2018), [].
  156. .U.S. Dep’t of Treasury, Bilateral Agreement Between the United States of America and the European Union on Prudential Measures Regarding Insurance and Reinsurance 4 (2017),   [] [hereinafter Bilateral Agreement].
  157. .And it was assessed as such. The American Insurance Association, an industry group, said that the “agreement on prudential matters will end the discriminatory actions against U.S. insurers and reinsurers, increase U.S. competitiveness, and boost the international standing of the U.S. state-based insurance regulatory system.” AIA Statement on the U.S.-EU Covered Agreement, American Insurance Association (July 14, 2017), [].
  158. .Bilateral Agreement, supra note 155, at 4.
  159. .Pruitt et al., supra note 154.
  160. .As Elizabeth Brown has said, “Solvency II will likely influence how insurance regulators outside of the EU regulate insurance, particularly those in the United States.” Elizabeth F. Brown, The Development of International Norms for Insurance Regulation, 34 Brook. J. Int’l L. 953, 972 (2009).
  161. .See id. (“Solvency II would require that non-EU insurance companies be regulated by a supervisory authority equivalent to the national authorities within the EU . . . .”).
  162. .See Schwarcz, supra note 62, at 543 (“[F]inancial data is generally not closely scrutinized by the states in which individual companies are licensed to conduct business but are not domiciled. Instead, these states defer to the financial analysis and regulation of the state of domicile . . . .”).
  163. .Bilateral Agreement, supra note 155, at 5.
  164. .For a criticism of the way these relate, see Arthur E. Wilmarth, Jr., The Financial Industry’s Plan for Resolving Failed Megabanks Will Ensure Future Bailouts for Wall Street, 50 Ga. L. Rev. 43, 48 (2015).
  165. .See Michael Greenberger, Overwhelming A Financial Regulatory Black Hole with Legislative Sunlight: Dodd-Frank’s Attack on Systemic Economic Destabilization Caused by an Unregulated Multi-Trillion Dollar Derivatives Market, 6 J. Bus. & Tech. L. 127, 162 (2011) (observing that “the Dodd-Frank Act includes both the ‘Volker Rule,’ which generally prohibits banks from engaging in proprietary trading or ownership of hedge or equity funds, and the ‘Lincoln’ or ‘Push-Out Rule,’ which requires bank holding companies to establish separate affiliated corporations for, inter alia, commodity swaps dealings . . . .”).
  166. .Bilateral Agreement, supra note 155, at 4–5.
  167. .David Zaring, International Institutional Performance in Crisis, 10 Chi. J. Int’l L. 475, 485 (2010) (describing the way this repeated conference had evolved).
  168. .Id.
  169. .Peter Conti-Brown & David Zaring, The Foreign Affairs of the Federal Reserve 33–37, J. Corp. L. (forthcoming).
  170. .For more on this, see Zaring, supra note 166, at 485.