Over at DealBook, I've got something on the financial regulatory reform that Europeans, in particular, love. Give it a look. And let me know if you agree with this bit:
Can a supervisory college work in lieu of a vibrant global resolution authority regime? The problem with these colleges is not that they are implausible, but that they have not really been tried in a crisis. The best-known supervisory college outside of the European Union was created in 1987 to monitor the Luxembourg-based, but international, Bank of Credit and Commerce International. Rumors of widespread fraud in the management of the bank were plentiful, but the collegiate approach did not mean that these problems were nipped in the bud. Although coordinated supervision led regulators to close many of bank’s branches at once after the bank’s accountant resigned and its insolvency became obvious, it is not clear whether Bank of Credit and Commerce International is a college success story or cautionary tale.
There are other reasons to worry about relying on colleges. The collegiate approach is meant to encourage communication more than action. Colleges operate as peers, convened by the home banking regulator, without the sort of hierarchy of decision-making and direction that leads to coordinated action.
I wanted to finish up my discussion of administrative enforcement by considering alternatives. We often take regulatory enforcement for granted. A securities regulator, for example, naturally will have the power to seek out violations of the securities laws and sanction violations. As is common in administrative law, scholars, courts, and Congress start with the assumption that expertise in the industry is the most important input into the enforcement calculus. If an agency is familiar with an industry, it will make good enforcement choices.
In my forthcoming article in the Minnesota Law Review, I argue that this question is actually much more complex than we usually assume. In particular, prosecutorial discretion has strong generalist aspects that largely do not depend on the regulatory subject matter. Giving enforcement authority to a specialist agency instead of a generalist enforcer (such as the Department of Justice) trades one type of expertise for another. Furthermore, specialists inherently see enforcement actions more narrowly. As a result, we shouldn’t see enforcement by regulatory agencies as inevitable or automatic.
Since it is still in draft form, I’d very much appreciate any and all comments. Thanks again to Erik for the chance to blog this week.
I blogged yesterday about administrative enforcement, an area that lies at the intersection of criminal and administrative law. Among other topics, my scholarship has considered the civil penalty process. In particular, what are the inputs and incentives that shape administrative agency penalties?
A standard model used to describe the penalty process emphasizes economic theories of deterrence. Financial penalties are a mechanism to raise the price for violations either to make misconduct completely unprofitable or, in the alternative, to force violators to internalize the costs of violations. I’ve pointed out one way that this theory may break down – administrative agencies might not focus on deterrence at all. Instead, their penalties may be crafted to achieve retributive ends.
In our recent Harvard Law Review article, For-Profit Public Enforcement, Margaret H. Lemos and I looked at penalties from another perspective: public enforcers may have self-interested reasons to maximize civil penalty recoveries. These incentives are widely recognized in private enforcement. Class action lawyers, for example, operating on a contingency fee basis have straightforward reasons to maximize recoveries.
Perhaps less obviously, public enforcement lawyers can have comparable incentives to impose large penalties. These incentives work most clearly in cases where enforcement agencies keep a portion of the civil penalties imposed. This structure is common in the asset forfeiture process used in connection with many criminal cases and also exists in other state and federal enforcement contexts. Even when penalties are turned over to the general treasury, enforcers may have reputational incentives to maximize penalties. Both agencies as a whole and individuals working in enforcement agencies may seek to build a reputation as an aggressive enforcer for reasons other than deterrence.
Assuming that these claims are right and that civil penalties can be driven by retributive or self-interested goals, is this a problem? Perhaps, perhaps not. Self-interested public enforcement may push enforcers to emphasize financial recoveries over other tools of regulatory control, such as injunctive relief. However, if our default assumption is that administrative agencies underenforce and usually do not impose adequate penalties, the pressure of self-interest may correct this trend to some degree.
The presence of retribution in civil penalties has similarly mixed effects. Of course, if penalties are supposed to be carefully calculated to deter, retributive ends will hamper this goal. On the other end, we now widely recognize the role of norms in shaping compliance behavior. Retributive punishment done well can shape and reinforce industry norms.
Usually thought of as unusually receptive, for a financial regulator, at least, to legislative pressure, the SEC, perhaps in a testament to its recent obsession with insider trading, has done the opposite and filed suit against Congress, subpeonaing a congressman and his aide to see whether the aide disclosed news to a lobby/law firm about health funding that caused a bunch of stock prices to spike ahead of the announcement of the new policy. DOJ is in on the game as well.
Congress is, it appears, displeased:
“What the SEC has done is embark on a remarkable fishing expedition for congressional records -- core legislative records,” [congressional lawyer] Kircher said in a court filing. “The SEC invites the federal judiciary to enforce those administrative subpoenas as against the Legislative Branch of the federal government. This court should decline that invitation.”
The so-called speech and debate clause in the Constitution protects members of Congress and staff from any outside inquiry into legislative business.
It is pretty juicy, and we'll outsource why to corp counsel. I'm just ballparking here, but a conversion between an aide and a lobbyist would seem to be deeply, deeply covered by the speech and debate clause, as unappetizing as it might seem. Here's a note on the clause, and here's the Heritage Foundation, which does these recaps pretty well.
And here's corp counsel:
the DOJ and SEC have sent subpoenas to Rep. David Camp, Chair of the House Ways & Means Committee, and Congressional Staffer Brian Sutter, regarding whether they tipped traders about a change in health care policy in the wake of a long-running investigation. And on Friday, as noted in this WSJ article, the SEC filed a lawsuit in the Southern District of New York seeking to compel the subpoenas. Possible grand jury to follow.
Here’s an excerpt from David Smyth’s blog about the case:
This is fascinating to me for so many reasons, among them: (1) the potential Constitutional cluster we’re about to witness; (2) the real test this poses for the recently passed STOCK Act’s effectiveness; and (3) another example of Mary Jo White’s severe distaste for those who defy Commission subpoenas.
And here’s an excerpt from the latest WSJ article:
“It’s not unheard of for an agency to serve a subpoena to Congress, but for an agency to sue is—if not unprecedented—at least very rare,” said Michael Stern, who was senior counsel to the U.S. House from 1996 to 2004. “It shows that there is a serious conflict; the SEC really wants the information and the House really wants it protected,” he said.
Erik, thank you for that introduction. It is a pleasure to join the Conglomerate for a week. My scholarly interests have recently focused on federal administrative enforcement – enforcement actions by agencies like the SEC, the CFTC, as well as a host of lower profile entities. This is a fascinating area of public law combining two scholarly literatures. Administrative enforcement actions share much in common with criminal cases. They are brought by public entities to vindicate public wrongs. However, the administrative context deeply shapes this type of enforcement. For example, unlike most prosecutor's offices, administrative enforcement bodies tend to be industry-specific.
As a result, administrative enforcement can go wrong in two different ways– the “criminal law” way or the “administrative law” way. Administrative agencies face the challenges of regulatory capture, inadequate or incorrect information, or simply the wrong incentives to engage in appropriate regulatory action. Criminal enforcement, though, often struggles with procedural fairness as well as the difficult task of assigning the correct level of punishment to different forms of misconduct.
Take, for example, this last issue: the fundamental question of penalty levels. Administrative agencies commonly use financial penalties to punish regulatory violations. How should these penalties be set? Which cases require the largest penalties and which only need more modest sanctions?
Criminal law scholars will recognize this question as an inquiry about theories of punishment. Speaking broadly, criminal law considers a couple of approaches. Utilitarian theories of punishment (e.g., deterrence, rehabilitation, incapacitation) seek to punish conduct to produce beneficial social outcomes. Retributive theories emphasize desert – punishment occurs because the violator deserves punishment, not because it produces a social benefit.
So what do federal agencies do? As I argue here, administrative agencies almost uniformly talk about deterrence, but usually engage in retribution. When setting penalty levels, agencies move penalties up or down in response to facts that justify retributive punishment but do not adjust penalties in the way deterrence requires. For instance, building on Gary Becker’s justifiably famous work, Crime and Punishment: An Economic Approach, most deterrence theories emphasize the role of the probability of detection in setting penalty levels. To deter appropriately, penalties need to increase when violations are harder to detect and punish. In practice, though, administrative agencies place little weight on this issue. Instead, agencies are deeply concerned with issues like mens rea, a topic far more central to retributive theories of punishment.
Is this retributive bent in administrative enforcement surprising? Perhaps not. A large literature suggests that most people are intuitively retributive when making punishment choices. In social science experiments, study participants set penalties based on retributive concerns, but do not adjust punishment levels in ways that would be required to deter appropriately. In this way, administrative agencies look like the rest of us. We mostly care about desert even when we talk about deterrence.
I've been on a bit of an international and administrative law kick these days, but as always, the case study is financial regulation. If you're interested in Sovereignty Mismatch And The New Administrative Law, available from the Washington University Law Review and here, you know what to do. Here's the abstract:
In the United States, making international policymaking work with domestic administrative law poses one of the thorniest of modern legal problems — the problem of sovereignty mismatch. Purely domestic regulation, which is a bureaucratic exercise of sovereignty, cannot solve the most challenging issues that regulators now face, and so agencies have started cooperating with their foreign counterparts, which is a negotiated form of sovereignty. But the way they cooperate threatens to undermine all of the values that domestic administrative law, especially its American variant, stands for. International and domestic regulation differ in almost every important way: procedural requirements, substantive remits, method of legitimation, and even in basic policy goals. Even worse, the delegation of power away from the United States is something that our constitutional, international, and administrative law traditions all look upon with great suspicion. The resulting effort to merge international and domestic regulatory styles has been uneven at best. As the globalization of policymaking is the likely future of environmental, business conduct, and consumer protection regulation — and the new paradigm-setting present of financial regulation — the sovereignty mismatch problem must be addressed; this Article shows how Congress can do so.
Comments and concerns welcome.
It is pretty fishy to argue that tenure deprives students of the right to an education (as opposed to being a reasonable call by the legislature that it is a way to vindicate that right), and one is taking one's chances when the first citation in an opinion is to Brown v. Board of Education, but that's what a California court just held and did. I'm guessing the deans at the state's various law schools will wait to see how this one plays out before sending out the pink slips.
After the SEC settled with Citigroup over misreprsentations made about a toxic security it sold during the financial crisis for a centimillion dollar fine among other things, Judge Rakoff rejected the settlement for failing to contain "cold, hard, solid facts established either by admissions or trials." I've been pretty critical of the decision, which was always headed for reversal. Not that Judge Rakoff cares: his familiarity with the agency (he once was in it), his generally respected status as a judge, and rumblings of discontent by other courts asked to approve other settlements once he fired his shot across the SEC's bow has led to a change in approach by the agency; I talked about the new policy here.
The problem with the decision was twofold, according to the Court of Appeals, at least as I interpret it.
Problem 1: Doctrinally, a settlement decision is an exercise of enforcement discretion, and enforcement discretion is basically unreviewable because the alternative - making it reviewable - would thrust the courts into the heart of what the executive branch does. Because the SEC wanted continuing court supervision of Citigroup as a consequence of the settlement, Rakoff did, indeed, have something to do. But if the SEC had simply dismissed its suit in exchange for the payment of a fine, which is less onerous than a fine plus continuing supervision by a court, Rakoff would have had, literally, no role to play in the resolution of the case. So requiring cold, hard facts to be established as a condition of signing off on a deal was a radical increase in the oversight of the SEC by a court.
No surprise, then, that the Second Circuit said that "there is no basis in the law for the district court to require an adminision of liability as a condition for approving a settlement between the parties. The decision to require an admission of liability before entering into a consent decree rests squarely with the SEC."
Problem 2: Settlements are not about right and wrong, while admissions of guilt are. Settlements are about moving on. We don't expect private parties to establish whether management caused the bankruptcy or someone else did, whether that product really was dangerous, or was misused by consumers, or whatever. And these can be matters of great public import. So it was never clear why the government, even though, yes, it is a state actor, should be treated very differently.
No shock, then, that the Second Circuit has said that "consent decrees are primarily about pragmatism" and "normally compromises in which the parties give up something they might have won in litigation and waive their rights to litigation."
According to the appellate court, the right way to review consent decrees is for procedural clarity and, as far as the public interest is concerned, with Chevron deference to reasonable decisions by the agency. It's not totally clear what that deference means - the court faulted Rakoff for figuring out whether the public interest in the truth was served by the deal when he should have been deciding "whether the public interest would be disserved by entry of the consent decree." But there you go.
Anyway, I think this stuff is interesting, because it's a tool in the regulatory arsenal, and indeed, my first baby law professor article was on just that.
Or so you could conclude if you look at my latest in DealBook:
The Business Roundtable has protested that it “is all too rare that agencies ask whether the original problem a regulation was issued to address has been solved or could be addressed more cost-effectively.”
Congress has heard these complaints. Bills with bipartisan sponsorship that would create an independent commission to recommend the repeal of antiquated rules are proceeding through both the House of Representatives and Senate.
But this is a case where the cure is dubious and the disease exaggerated.
Two bright spots yesterday from the Office of the Comptroller of the Currency. First, the OCC announced a new policy of rotating examiners among banks. Second, this recommendation stemmed from a peer review of the OCC by international financial regulators. With multiple eyes, all bugs are shallower.
In Governor Tarullo’s closely watched speech on bank regulation (I already blogged on the idea of a sliding scale of regulation based on bank size), he also argued for ending the IRB approach to capital requirements. How does IRB translate into Plain(er) English: DIY capital requirements for big banks. Tarullo’s argument is based in part on obsolescence: all the increases to capital in Dodd-Frank and Basel III dwarfed the IRB component. But it is also based on the fact that DIY capital requirements was a spectacularly bad idea (something I wrote about back when). Big banks have built in incentives (courtesy of government guarantees, explicit and implicit) to lower their capital and increase their leverage. Tarullo’s coming to bury not praise this part of Basel II calls for revisiting some of Joe Norton’s prescient work critiquing that accord as a political economy product of lobbying by behemoth banks. If we take New Governance and its experimentalism seriously, it is vital that we look at experiments that failed.
I want to draw your attention to this, now out in the Cornell Law Review. As regular readers know, I do a great deal of research on the international regulatory networks that increasingly set the standards for financial and securities regulation. This paper is an effort to connect the shaky legitimacy of that sensible impulse to a stronger body of doctrine, and so it touches more on foreign relations and administrative law than on pure financial regulation - and it is co-authored with a foreign relations and international law expert, Jean Galbraith. You should download it. The abstract:
The United States increasingly relies on “soft law” and, in particular, on cooperation with foreign regulators to make domestic policy. The implementation of soft law at home is typically understood to depend on administrative law, as it is American agencies that implement the deals they conclude with their foreign counterparts. But that understanding has led courts and scholars to raise questions about whether soft law made abroad can possibly meet the doctrinal requirements of the domestic discipline. This Article proposes a new doctrinal understanding of soft law implementation. It argues that, properly understood, soft law implementation lies at the intersection of foreign relations law and administrative law. In light of the strong powers accorded to the executive under foreign relations law, this new understanding will strengthen the legitimacy and legality of soft law implementation and make it less subject to judicial challenge. Understanding that soft law is foreign relations law will further the domestic implementation of informal international agreements in areas as different as conflict diamonds, international financial regulation, and climate change.
Please don't hesitate to send along your comments, concerns, etc.....
Oops, says BofA, we messed up our capital calculations. We don't have as much money on hand for shocks or emergencies as we thought. Since that's the principal thing that banking regulators care about, you might wonder what happens to banks who do this. Perhaps it would be interesting to consider some alternatives, might offer a sense of what bank supervision does and doesn't involve these days.
- The Fed could prosecute BofA executives for fraud. Call that the securities regulator/white collar approach. One problem, fraud must be intentional, so this would have to be not an error, but at the very least some sort of reckless accounting. It punishes individuals in management who contributed to the fraud.
- The Fed/FDIC could revoke their license or pull the inspectors. This is the USDA approach. The problem is that it is too nuclear - both of those things would shut down a bank that is far too big to fail.
- The Fed could fine them. This is the money laundering approach, and those fines are often imposed not just for tolerating the laundering of money, but for having inadequate controls in place to prevent it. We may see a fine here, BofA is pretty much saying that it had inadequate controls in place by acknowledging that it did the calculations wrong to the tune of billions of dollars.
- Or the Fed could do what the Fed is, for now, doing. It is suspending any dividend increases by the bank until it submits an accurate account of the state of its capital reserves, and has that account approved by the agency as sufficient. This is a somewhat new thing in high level banking oversight - punish the shareholders, thereby encouraging them to monitor management. Does it work? It is perhaps a little untested, although suspending capital distributions has been a tool used by the FDIC on the sorts of distressed small banks that were its old stock in trade. Seeing that tool applied to Citi and now BofA, however, is a dfferent thing altogether. It will be interesting to see if this trend continues.
- "That one million hours a year devoted to resolution planning is 500 full-time employees"
- "There are 8,000 employees 'dedicated solely to building and maintaining an industry-leading Anti-Money Laundering (AML) program.' JPMorgan employs more AML compliance officers than the Treasury and the Fed combined."
- Stress testing required 500+ FTEs
- Compliance with Basel's new securitization rules has required 35,000 hours of work (at 2000 hours per year, that's only 17.5 FTEs, so you can see why they moved to hours there).
That's a lot of compliance, and indeed, at these rates, way more people do compliance for JPMorgan than, probably, do actual investment banking. Of course, maybe we want all of this given that the firm is far too big to fail, and maybe we want to make banking burdensome and unprofitable. If so, we are on our way!
The DC Circuit rather shockingly threw out the SEC's conflict minerals rule ONLY because it compelled publicly traded companies to speak about the issue in their securities filings, which it concluded violated the First Amendment. EDIT: This means that the parts of the rule that require reporting but not a statement that goods are "not DRC conflict free," might still be okay. Bainbridge has takes here and here, Jonathan Adler here, Matt Levine here.
But, you are thinking, the SEC compels companies to do a million things in their securities filings! Does the very existence of a disclosure regime violate the First Amendment? The court's novel theory was that it is okay to mandate disclosures that aim to prevent consumer deception (so the books of publicly traded companies could be opened to investors), but any other goal must have more than a rational basis to be sustainable.
It is a crazy theory. Warning labels, origin labels, nutrition labels, mandatory agricultural marketing schemes, they don't involve consumer deception, and they're okay. And maybe this reveals a lack of adoration for the First Amendment, but if Congress could prohibit companies from using conflict minerals, which it surely can, then requiring them instead to disclose the use is both less burdensome and possibly more efficient. Why would we want a legal system that does not permit disclosure regimes, thereby requiring command and control?
Some other observations:
- One judge wanted the court to wait for a ruling in a related case going en banc before the now democratically controlled circuit, and the two majority judges declined to do so because now the SEC and the petitioners could participate in the en banc. Unless the new Obama judges on the court cannot hear the en banc, this seems like a request for a quick reversal.
- Also interesting, the court didn't bullet proof the opinion. The SEC survived the adlaw challenges, and the very controversial cost-benefit analysis requirement the DC Circuit has started imposing, though that is likely to change very soon, on the agency. There is only one ground for reversal here: disclosure is unconstitutional.
- There is a difference between speech and conduct in the First Amendment, but the other big thing the SEC does in foreign policy is corrupt practices prosecutions (bribes paid to foreign government officials, that is). Could that be affected by the holding of this opinion, were it to stand? It sure isn't consumer protection.
- One of my many pet theories about why people care about constitutional law, though they often overdo it, is the sense that stare decisis is only sort of a good way to think about the subject. Conservative judges clearly love commercial speech, and have been using it to reverse some settled doctrines that have been in place for decades. I doubt a single securities lawyer thought that this was a plausible holding by the Court. Some smarter on the subject than I were clearly surprised. Let's see if it lasts.