The Lewisization/Gladwellization of social science

Lewisization and Gladwellization are synonyms for abuse of social science in the service of readability, or what one might prefer to call narrative-arcification. Lewis’ Flash Boys is the latest and most damaging because, while the book hovers around a real problem in financial markets related to high-frequency trading, it doesn’t identify it clearly or explain the need to regulate–indeed, one could finish the book thinking that the market cured itself.

I wanted to write about this but I can’t improve on Felix Salmon’s sparkling review and blog posts. I made an argument similar to Salmon’s in this piece written with Glen Weyl, but you can’t do better than read Salmon.

Fragile by Design, by Calomiris & Haber

Fragile by Design, by Charles Calomiris and Stephen Haber, is a great book. The authors argue that the stability and efficiency of financial systems in different countries depend on political bargains that set the rules of the game. Authoritarian countries produce inefficient state-owned banking systems because governments cannot commit not to expropriate. But democracies produce a wide range of outcomes, depending on the configuration of constituencies and interest groups. The United States is cursed with a highly unstable banking system because local interests have been able to ensure a huge number of local unit banks that were insufficiently diversified. This system finally broke down thanks to the inflation of the 1970s, but our further misfortune was a political bargain between activist groups and large banks in the 1990s that resulted in a system where banks were encouraged to reduce underwriting standards so as to extend credit to low-income people. Meanwhile, countries like Canada were lucky enough that the initial political bargain at a national level led to a small number of stable and efficient banks that weathered the financial crisis of 2008.

There are a huge number of moving parts and epicycles (why was our unstable system so stable from 1936-1980?, and what explains the success of activists like ACORN?), but the book is nonetheless enormously illuminating, and contains the most powerful and concise account of the causes of the 2008 crisis that I have seen.

Goldman on Bitcoin

Business Insider exaggerated when it announced that Goldman completely obliterates Bitcoin in a new report. The report includes interviews with Bitcoin supporters. And while the Goldman analysts are skeptical that Bitcoin could serve as a currency—the view of nearly everyone nowadays—they do not rule out a role in the payments system. Currently, merchants pay 2-3 percent of purchase price to accept electronic payments. Bitcoin service providers charge 1 percent. But as the Goldman analyst notes, much of the cost of the current payment system is attributable to security and legal requirements that Bitcoin providers will eventually need to confront. Merchants who use bitcoin pay an additional 1 percent to exchanges in order to avoid exchange rate risk. Traditional payments system will also reduce costs in response to competition from Bitcoin. However this all works out, the long-term effect of Bitcoin will not be anarchist utopia but slightly lower prices—you may end up paying $100 rather than $101 for an item you buy over the web.

Other thoughts:

Bitcoin miners with 51 percent of the computer power over the Bitcoin network control the supply: they can decide to increase it. Question: don’t they have strong incentives to undersupply bitcoins—that is, to vote against increasing the supply to the social optimum while hording bitcoins—in order to maximize their profits, like De Beers?

Bitcoin’s legal problems are just beginning. An interview with a pair of lawyers reveals a potentially huge regulatory web that legitimate bitcoin institutions will need to navigate. Once bitcoin futures come into existence in sufficient volume, the CFTC will step in. We already know about money laundering laws, which require bitcoin services to keep tabs on customers and report suspicious transactions. The SEC has gotten into the act because of efforts to combine Bitcoin and securities. State regulatory agencies may require Bitcoin-related companies to obtain licenses akin to those that money transmitters like banks must obtain, which are costly. It also seems likely that Bitcoin services will, like existing money transmitters, be required to keep funds on hand to compensate customers if their bitcoins are lost—a further cost. Not discussed, but also worth considering, is the possibility that people will try to manipulate the bitcoin market—as I suggested above—necessitating another layer of regulatory scrutiny.

The bitcoin paradox

The implosion of Mt. Gox exposes a paradox about bitcoin, which I have been groping for in some writings. Assume that the bitcoin software works perfectly (though there is some question about this) or can be made to work perfectly (as advocates argue). Bitcoin still has a problem with the “joints”–the gap between the network itself and the ordinary (non-expert) users without which it could never be more than a marginal phenomenon. Ordinary people will need to rely on institutions–exchanges like Mt. Gox and other services–and they will not rely on them unless they can trust them. But, unlike bitcoin itself, these institutions are run by human beings who can make mistakes or engage in fraud. Hence the need for regulation. Thus, bitcoin will prosper only if it is integrated into the regulatory infrastructure, but that means that it cannot operate as a decentralized currency outside of government control. Yet it is that feature that makes bitcoin so attractive to its most ardent supporters. I expect that legitimate investors and merchants who may benefit from it will push the government to normalize bitcoin by regulating the intermediary bitcoin institutions, at which point it will no longer be an autonomous currency but just a useful piece of software.

N.B.: journalists reflexively describe bitcoin as a means of transferring value without using an intermediary, but for ordinary consumers that is the case only in the sense that it is true for currency as well. You could put a bunch of dollars into a wheelbarrow and wheel them to the store. Banks exist because this is impractical, and in the same way bitcoin intermediaries like Mt. Gox exist because it is impractical for most people to wheel around bitcoins on their own.

The Empty Call for Benefit-Cost Analysis in Financial Regulation by Jeffrey Gordon

Jeffrey Gordon has posted a paper arguing that applying cost-benefit analysis to financial regulation is a “serious category mistake.” He makes the arresting claim that CBA works best for the real economy, which is governed by the laws of chemistry and physics, but not for the financial economy, which is a “constructed system.” The bulk of the paper is devoted to showing the law of unintended consequences in action. Rules developed in the 1970s to permit money market mutual funds ended up harming S&Ls, which could not compete for funds, with the result that they were deregulated, whereupon they self-destructed. Etc.

Gordon is strongest in showing the sheer unpredictability of financial regulation. But is this problem worse than in other areas of regulation? Perhaps. Possibly connected to the idea that the financial economy is a “constructed system,” arbitrage seems to be a great deal easier in the financial world–limited only by imagination and computer power–than in the real world, where it can be hard to retool factories and move power plants. Still, I remain optimistic. Gordon’s is a nice companion piece to Coates’ paper, which I discussed here.

John Coates on cost-benefit analysis of financial regulation

John Coates recently posted a paper on SSRN entitled Cost-Benefit Analysis of Financial Regulation: Case Studies and Implications. This topic has been important ever since the D.C. Circuit struck down an SEC regulation for failing CBA in Business Roundtable in 2011. Glen Weyl and I held a conference on the topic last fall, and have written several papers arguing that, whatever one thinks of the reasoning in the (justly criticized) Business Roundtable case, CBA is the way to go.

The core of Coates’ paper is a description of efforts by agencies (and other institutions or persons) to perform CBAs of six major financial regulations. Coates persuasively argues that the existing CBAs fall flat and expresses skepticism that it is even possible, given the current state of knowledge, for meaningful CBAs of financial regulations (or certain financial regulations) to be conducted. What valuation should we assign to an avoided financial crisis? Hard to say. Coates concludes that while it makes sense for agencies to engage in cost-benefit balancing using rough guesstimates, their efforts should not be subject to judicial review.

Coates’ argument is sensible, but I am more optimistic than he is about the capacity of agencies to come up with numbers. This criticism–and many others that Coates makes–could be (and have been) made about other forms of CBA, for example, CBA of environmental regulation, which has improved greatly over the last 30 years but remains far from perfect. The fact is that when an agency proposes a capital adequacy rule, it is using an implicit valuation for an avoided crisis. The agency should be required to make that valuation explicit, and defend it. We can agree that a wide range of valuations is acceptable and also believe that public discussion of the range is useful.

I have mixed feelings about his criticisms of judicial review. On the one hand, it seems likely that if courts rigorously applied CBA to new financial regulations, we would not have any new financial regulations (a bad thing), at least not for some time. On the other hand, I’m not sure what will encourage agencies to perform CBAs properly–and this means paying money to consulting firms to generate valuations–if judicial review does not take place. OIRA has encouraged non-financial agencies to use CBA, but OIRA has more limited authority over the financial agencies–none at all in the case of the Fed.