Note to Simon Johnson: I do not think antitrust means what you think it means

Geoffrey Manne —  21 January 2010

Simon Johnson is at it again, advocating the use of antitrust to break up the banks because they are, you know, big, and antitrust is about busting up big companies, right?

As Josh suggested back in July, the idea is gaining momentum, it seems.  The Financial Times is also pushing the idea.  What’s remarkable about both the FT’s and Simon Johnson’s “analysis” is that it is actually largely devoid of modern antitrust analysis.  It reflects the outdated, non-economic (dare I say anti-economic?) logic of the structure-conduct-performance paradigm of the 1960s.  Here, for example, is the FT:

The issue is not just market share in deposits, which attracts public attention. Many corners of the financial services market with a lower profile are highly concentrated and highly profitable.

Concentrated markets can still be competitive and high profits are not in themselves proof of anti-competitive behaviour. Market concentration is high in many industries – including new areas of the economy such as web search.

Anti-trust investigation would focus only on areas where critics allege anti-competitive behaviour, for example, the puzzle of why investment banks charge standard rates to raise equity capital for businesses. But even where there is no suspicion of collusion, regulators could examine market structures and practices that create barriers to entry.

Or, the administration could simply float the idea of a wide-ranging push on competition in banking and use it to gain leverage over big banks on issues such as regulatory reform and whether or not they pass on the cost of the financial levy to their customers.

So, yes, there is a nod (“concentrated markets can still be competitive . . .”) toward recognition that Demsetz does, indeed, exist and he did write Industry Structure, Market Rivalry, and Public Policy in 1973.  But the thrust is pure SCP.

Here’s Johnson:

There are definite elements of oligopoly in wholesale markets, underwriting new issues, and mergers and acquisitions both in the United States and around the world. This is part of the explanation for very high profits in banks — particularly big banks — over the past decade.

The question becomes: Is there evidence that our leading banks have used their pricing power or other aspects of their market muscle to keep out competition or otherwise distort behavior in very profitable arenas, like over-the-counter derivatives?

I don’t even know what that means:  “pricing power or other aspects of their market muscle to keep out competition?” Huh? It’s just hand waving.  And my favorite part:

We may also need new theories of antitrust.

Sure. Why not? Here “new” seems to mean “old,” but by all means we should launch antitrust investigations with the intention of developing new theories of antitrust that can justify the a priori policy conclusion that banks are violating the antitrust laws.  That’s some sound policy advice from the IMF’s former chief economist.

His “analysis” has not changed, as far as I can tell, since the last time he advocated bank busting, so I can hardly do better than repeat what Josh said back in July:

FWIW, I’m skeptical about the utility of introducing “too big to fail” as an antitrust concept.  Antitrust has come a long way since its economically unprincipled approach several decades ago to its current state.  It has done so largely by staying relatively hinged to microeconomics.  This approach has done antitrust well as evidenced by the evolution of the doctrine over the past 30 or so years.  We now have a substantial body of economic theory and empirical evidence that tells us quite a bit about sensible approaches to at least cartel and merger enforcement that are likely to help rather than harm consumers on net.  Injecting “too big to fail” as an antitrust concept whether under the Clayton Act or otherwise is not a minor tweak to the system.  Too big to fail is not an antitrust concept and attempts to operationalize it within the antitrust framework are likely to cause more harm than good by undermining the progress that has been made by sticking to a disciplined economic approach.  As I commented for a related story in The Deal, and consistent with some of my own research on economic education and complexity in antitrust cases,  “Consumer welfare is complicated enough” for judges and enforcement agencies as is.  But the threat is not just increasing the risk of errors associated with introducing this factor into the antitrust calculus, but also allowing it to substitute for and gradually subsume the economic approach which has served us well.

This kind of hubris–that would throw out restrained and rigorous policy analysis precisely because it is restrained and rigorous and therefore not necessarily supportive of one’s preferred outcomes–is pernicious.  And, unfortunately, endemic.

Geoffrey Manne

Posts

President & Founder, International Center for Law & Economics

One response to Note to Simon Johnson: I do not think antitrust means what you think it means

  1. 

    It sounds much like the Hart Deconcentration Act, which Sen. Hart introduced back in the late `60s. Essentially, it would have broken up the largest firms in any “industry” (defined, I guess, by Census SIC codes) where certain concentration threshholds were exceeded (you can read this Act in all its glory in the back of the famous Goldschmidt “New Learning” book). No theory behind it, really, beyond a simplistic S-C-P syllogism. But this actually is more than what Simon Johnson has to offer — as Geoff notes, what he really is searching for is an ex post rationalization for a predetermined policy. Pretty shameful for an economist.