Deep and Long: Private Debt and Financial Recessions

While he (uncharacteristically) doesn’t explain or deploy it terribly well, Paul Krugman points to some very excellent research on the relationship between private debt levels and the depth and duration of (especially financial-crisis-driven) recessions. The Schularick/Taylor Vox EU article is here. The Jorda/Schularick/Taylor paper is here (PDF).

I think the work is excellent in large part because JS&T address one of my pet peeves: sample size and selection. Unlike Reinhardt and Rogoff, who make wild claims about government debt/GDP levels above 90% based on a mere handful of sample points (examples that mostly aren’t representative of our current or recent situation), and unlike the kind of single-country and short/single-period analyses that you see all the time, and totally unlike shameless cherry pickers such as John Taylor (who dares to call 1981 — unequivocally a Fed- and interest-rate-driven recession — a financial crisis) JS&T look at a long-term, representative, multi-country data set culled from:

o Fourteen advanced economies that at least in that sense are representative of the economies that at least I am interested in understanding (i.e., ours). “The share of global GDP accounted for by these countries was around 50% in the year 2000.”

o Over 140 years

o Using a very clear definition of “financial recession” (below).

They end up with a sample of 208 recessions in advanced economies over that period, 35 of which were financial recessions.

And then they look at multiple lag times (something that is far too often absent from time-series studies): with conditions X in year zero, how do economies perform 1, 2, 3, 4, and 5 years later? (I rather took this issue to the limit in comparing US and EU growth rates, here.)

Results? I’ll cut to the comment that I just left on Krugman’s blog, with graphics added and some links and editing for your reading pleasure

The Schudarick/Taylor Vox article is excellent, as is the Jorda/Schularick/Taylor paper underlying it.

But I don’t think you describe them well.

First, ST&J don’t use “use pre-crisis credit growth … to identify financial-crisis slumps.” They use Laeven and Valencia’s (PDF) definition of a “systemic banking crisis.” [Emphasis and bracketed additions are mine.]

…a country’s corporate and financial sectors experience a large number of defaults and financial institutions and corporations face great difficulties repaying contracts on time. As a result, non-performing loans increase sharply and all or most of the aggregate banking system capital is exhausted. This situation may be accompanied by depressed asset prices (such as equity and real estate prices) on the heels of run-ups before the crisis, sharp increases in real interest rates, and a slowdown or reversal in capital flows. In some cases, the crisis is triggered by depositor runs on banks [yes we saw some -- Northern Rock -- but largely a cause rather than an effect], though in most cases it is a general realization that systemically important financial institutions are in distress. [No names, just initials: Lehman Brothers. AIG.]“

Yes: they find “that such [financial] slumps are indeed characterized by slow recovery.”

But their focus and key finding relates to private debt levels prior to such crises, and how those debt levels relate the depth and duration of the ensuing recession. They conclude that:

1. Higher private debt levels (and runups) prior to recessions correlate with much deeper and longer recessions (slower recoveries), and

2. That effect is especially pronounced with financial recessions, (those associated with systemic banking crises).

A typical financial recession in an advanced country that begins with high private debt levels takes two years to hit -5% GDP growth, and continues for three more years without regaining its previous level.

We saw U.S. private debt skyrocket off the charts in the thirty years leading up to this financial crisis (except in the early 90s…), utterly dwarfing the runup in government debt:

Based on ST&J’s work, that runup and magnitude in private debt, preceding (and causing) the financial crisis (combined with standard and straightforward Fisher/Minsky theory), is all the explanation anyone needs for the deep downturn and slow recovery.

A grateful nod to my fellow contributors on this subject: While I find Jazzbumpa and Arthur’s arguments about private debt and growth compelling, I do not find the single-country (U.S.) evidence that they provide conclusively convincing (for the reasons outlined above, and here [also see comments there]). I am pleased to find that a far more comprehensive and more systematically analyzed data set seems to support their conclusions. JS&T don’t do 20-year lags, so it’s hard to know from them what the long-term growth effects are of excessive private-debt runups. But they certainly display a mechanism whereby such runups could damage long-term growth: by savaging that growth in rare but devastating financial recessions.

And it’s worth remembering who that devastation is visited upon: the middle class and the poor. The rich do just fine.

Recessions are nature’s (and neoclassicals’) way…of keeping the little guy down.

Cross-posted at Angry Bear.

  1. Pedro Alvarez
    October 22nd, 2012 at 13:16 | #1

    Money is IOU. Debt is IOU. What is happening, then, in the private sector? A set of institutions (banks, pension funds, etc) hold IOU’s from ordinary folks, who bought homes, cars, televisions, etc.

    Money is both created and destroyed endogenously. Created when one creates IOU’s. Destroyed when one pays off the debt or when the debt is written off. Economy, as a whole, has to provide a way to pay off the debt; otherwise, debt holders should write off the debt. The interesting thing is that debt holders (like banks, mutual funds, pension funds, etc) are destroying the economy or the ability for ordinary folks to pay back the debt. This is just a micro-macro dilemma. At micro level, some action is rational; at macro level, same action becomes irrational.

    One may say that the rate at which endogenous money (private IOUs) is created should be reduced in such a way that the folks can service that debt. This reduces the aggregate demand, which destroys jobs.

    I think the whole paradigm of “get a job to pay bills, service debts” requires a rethinking.

  2. vimothy
    October 22nd, 2012 at 15:26 | #2

    Steve, Schularick and Taylor are good. Thanks for posting this. Another neat paper is: http://people.virginia.edu/~amt7u/papers/w15512.pdf

  3. Goldilocksisableachblonde
    October 22nd, 2012 at 19:17 | #3

    You might find this interesting :

    ” LEADING INDICATORS OF CRISIS INCIDENCE : EVIDENCE FROM DEVELOPED COUNTRIES ”

    http://www.ecb.europa.eu/pub/pdf/scpwps/ecbwp1486.pdf

    The focus is on advanced economies from 1970-2010.

    Bottom line : Watch out for private sector credit booms , the aftermath can be ugly. The good news is that if you pay attention , you can detect the boom 4 years prior to the crisis , and can thus nip it in the bud or , if not , you’ll have plenty of time to assume the fetal position.

  1. No trackbacks yet.