The US
Financial Crisis Inquiry Commission has released its report. In a sign of how politicized economic analysis is in the USA, there are two Republican dissents from the main report -- it appears that in good Brendan Behan style, once the Republican commissioners had decided to split away from the main report, they then fell out among themselves.
Anyway, three of them -- Keith Hennessey, Douglas Holtz-Eakin, and Bill Thomas -- have published
their dissent. It proceeds with a variety of non-controversial findings and at times makes it appear as if the differences with the main report are mainly a matter of emphasis on different aspects of those findings. But at the end, it's clear that isn't simply a difference about the "causes" of the crisis. It's a difference about
what was the crisis.
To preview a longer story, the HEAT dissent (as it will hereafter be called, geddit?) essentially says the actual crisis was one month long -- September 2008 -- and everything else is either "foreshocks" or after-effects. This focus on September is part and parcel of their somewhat strange application of "contagion", a concept that any reasonable crisis analysis would need to have in there somewhere. So --
If financial firm X is a large counterparty to other firms, X’s sudden and disorderly bankruptcy might weaken the finances of those other firms and cause them to fail. We call this the risk of contagion, when, because of a direct financial link between firms, the failure of one causes the failure of another. Financial firm X is too big to fail if policymakers fear contagion so much that they are unwilling to allow it to go bankrupt in a sudden and disorderly fashion [bold font in original].
So far, so good. But now things get weird. After laying out some too big to fail examples, they say --
The risk of contagion was an essential cause of the crisis. In some cases the financial system was vulnerable because policymakers were afraid of a large firm’s sudden and disorderly failure triggering balance-sheet losses in its counterparties
Notice the subtle but radical substitution: contagion didn't cause the crisis. "Risk of contagion" leading to interventions was the cause of the crisis. The only non-intervention that appears on their list of contagion events is Lehman. And to that, they add --
The focus on Lehman’s failure is too narrow. The events of September 2008 were a chain of one firm failure after another
followed by a long list of policy interventions including the TARP, and concluding
The financial panic was triggered and then amplified by the close succession of these events, and not just by Lehman’s failure.
The problem is that "these events" were the just the manifestations of a fragility that had been building up for years and had metastasized in the summer of 2007, one of those "foreshocks" that receive almost no attention in the report. September 2008 was when the shakeout peaked, but the crisis had already been in full flow for a year by that point. There was indeed a strange summer 2008 reprieve (which, incidentally, appears to have suckered governments in
Ireland and Iceland into a false sense of security as well), but to pick out one month when the financial accident and emergency room was especially busy isn't going to tell you a lot about how the crisis started.
Cynical addendum: the dissenting report is fixed on quarterly GDP as a measure of crisis severity, an indicator that is negative (and therefore looks bad in some vague sense) for just 9 months, 6 of them being especially awkward for George W. Bush and John McCain. Is the message that except for one month of costly meddling, things would have been fine?
