In yesterday’s FT Carmen Reinhart and Kenneth Rogoff argue for the creation of an international financial regulator. Their analysis essentially internationalizes the argument in favor of creating a super-regulator in the US. Unfortunately, I think that this conventional diagnosis and prescription is completely misguided. An understanding of the transactions costs of regulation and politics implies that this solution will make things worse, rather than better, because it aggravates the underlying factors that caused the regulatory failure in the first place. I know many people want to improve financial market regulation in the worst way–and this would be it.
Reinhart and Rogoff argue that an International Super-regulator (sorry, but that reminds me of the title of Green Day’s greatest hits CD, International Superhits) would be beter able to overcome the influences of domestic financial services interests than national regulators:
Finding ways to insulate financial regulation from political meddling is critical to creating a more robust global financial system in the future.
Indeed, the need for greater regulatory independence is a compelling reason for establishing an international financial regulator, another topic the G20 conspicuously avoided. We recognise that international financial institutions are far from perfect. Nevertheless, a well-endowed, professionally staffed international financial regulator – operating without layers of political hacks – would offer a badly needed counterweight to the powerful domestic financial service sector lobbies. The independence argument is in addition to the well-recognised need for better mechanisms for co-ordinating regulation to reduce regulatory arbitrage in a world of global capital markets.
. . . .
Important regulatory appointments are, of course, largely political, and governments determine the extent and power of the regulatory agencies. During the boom, many regulators throughout much of the world were put under pressure by leading politicians to lighten up. The lack of transparency, which the G20 leaders complain of so vehemently, also served as a convenient shield to keep politicians’ interference out of public view.
With asset prices soaring and rating agencies giving their usual overly optimistic boom-time assessments, regulators had few available weapons. This was no accident. Look at the list of leading contributors to the presidential and congressional candidates in the US election (including the primaries). Financial companies dominate. Thus it is no surprise that, during the boom, all the supposed market watchdogs were neutered. This is an international problem, not just a US one.
I have no doubt that interest group influence can neuter regulators. I am just far from convinced that creating an international regulator will solve the problem of “neutered” regulators. Indeed, there is some economic theory that predicts the exact opposite.
Specifically, in his “The Making of Economic Policy: A Transaction-Cost Politics Perspective,” Avinash Dixit draws on agency theory to show that the more diverse a regulator’s constituency and the broader its responsibilties, (a) the weaker its incentives will be, and (b) the less discretionary authority it will possess.
Agency theory shows that in the presence of measurement costs it is usually optimal to provide multi-task agents with weaker incentives and less discretion than single task agents in order to mitigate moral hazard problems. This, by itself, implies that regulatory agencies (note the word “agency”–they are agents of the legislative principal) with broad and diverse responsibilities will typically face weaker incentives and have less discretionary power; it is in the interest of the principal–the legislature–to structure the agency in that way.
Dixit extends the analysis based on an observation by James Q. Wilson that regulatory agencies usually have multiple principals. For instance, a regulatory agency deals with the executive, the legislature, and disparate interest groups. Dixit shows that the greater the number of principals-constituencies, the weaker should be the incentives and the more circumscribed the discretion that the agency faces.
Applying that to the idea of an international financial regulator suggests that is a very bad idea indeed. An international super-regulator as envisioned by Reinhart and Rogoff and others (e.g., Sarkozy) would have numerous responsibilities, over banking, equity markets, derivatives markets, in multiple countries. Thus, an international regulator would be more of a multi-tasker than traditional national regulatory agencies with authority over a single slice of the market in a single country. Moreover, it would have more principals than any national regulator. Creation of a super-regulator would not make domestic financial services interests go away; it would just give them an interest in influencing a new regulator. Moreover, whereas a regulator such as the SEC might have a single national legislature and a single national executive as principals, a super-regulator would have myriad national legislatures and myriad national executives as principals.
Thus, a super-regulator–either within a country, but especially an international one–has many tasks and many principals. The Dixit analysis implies that rather than being an aggressive regulator that would overcome petty national interests, a super-regulator would in fact be extremely weak. To make a play on one of Reinhart-Rogoff’s words criticizing incumbent national regulators, the super-regulator would be super-neutered. Put differently, expanding the scope of the regulator both in terms of its responsibilities and the nations which it would regulate would actually result in weaker regulatory oversight.
Indeed, Reinhart-Rogoff mention an example that illustrates my point: “The Basel agreements on bank regulation have succeeded to some extent in achieving this, but they take too long to negotiate. The Basel II agreement is already dead on arrival.” I would also argue that the European response to the crisis, notably the halting response of the European Central Bank, also should sound a note of caution to those who believe that a super-regulator with authority over institutions in multiple states would actually be more aggressive than a narrower regulator in an individual country.
Of course, issues of coordination must be considered as well, and perhaps a single regulator could more effectively coordinate a response than multiple, autonomous ones. But in my view, the “agency” problems would overwhelm any benefits from improved coordination. Reinhart-Rogoff and many others bewail the inability of national regulators with multiple regulatory responsibilities answering to multiple constituencies to address the financial crisis in an aggressive way. The Dixit argument implies that this is an inevitable feature that is purposely baked into the design of these agencies–to address agency problems their designers provide weak incentives and limit discretion. This is not a mistake–it’s supposed to be that way! That’s exactly what Congress and national parliaments designed them to act.
These agency problems arise from multi-tasking and mutiple principals. From this perspective, adding tasks and adding principals will make things worse rather than better.
Overall, the Dixit book (which I highly recommend–it’s a quick and pleasant read packed with insights) should make anyone highly doubtful about the prospects of regulation to fix serious problems like the financial crisis, or to prevent them from recurring in the future. Dixit makes a persuasive case that agency problems and transactions costs inherently limit the capability of political and regulatory institutions. Political, incentive, and information constraints sharply circumscribe what regulators can do–and what they should be allowed to do.
And the most persuasive part of the book is that which uses standard economic tools to demonstrate that the usual prescription for regulatory failure–make the regulators bigger with more responsibilities–is exactly the wrong thing to do. It aggravates, rather than ameliorates, the sources of regulatory failure. It is, in short, the political equivalent of bleeding to cure anemia.