Since the beginning of what has come to be termed “The Great Recession” in December 2007, we’ve been looking for analogies to understand the hole into which our economy has fallen. “The Great Recession” itself is a reference to the Great Depression of the 1930s. Others have compared the most recent recession to the twin recessions of the early 1980s, frequently with unfavorable views of federal government response to the recent downturn. Others have looked farther afield to the persistent doldrums that struck the Japanese economy in 1990 and continues to the present.
The comparison of the recent downturn to the Great Depression of the 1930s is particularly inapt. The Great Depression saw a decline in GDP of 25% or more in the United States, was worldwide, and saw sharp declines in production, wholesale prices, and foreign trade. Contrariwise the Great Recession has none of those characteristics.
Increases in unemployment between the two economic downturns are, however, quite comparable. [Update: this is an error. I am grateful to Steve Verdon for pointing out that at no time has present unemployment neared Depression Era levels, however measured, cf. this comment. I had incorrectly compared Depression Era U3 unemployment with present-day U6.]
Comparison with the Fed-induced recession of the 1980s is even farther off the mark. Unemployment increased sharply but recovered even more sharply. It was sharp, harsh, and over relatively quickly, a classic V-shaped recession in which growth rebounded almost as fast as the NBER could declare that a recession was in progress.
Over the period of the last several years the term “balance sheet recession” has come increasingly into vogue until not only is it the prevailing wisdom that the recession of 2007-2009 was in fact a “balance sheet recession” but any attempt to call that into doubt is met with considerable derision. “Balance sheet recession” derives from economic Richard Koo’s 2003 book, Balance Sheet Recession: Japan’s Struggle with Uncharted Economics and its Global Implications. The timeliness of Dr. Koo’s work has given him near-oracular status on this subject. Under the circumstances and, particularly, considering the increasing likelihood of further declines or, at the very least, persistent unacceptably slow growth for the foreseeable future, I think it’s reasonable to consider the evidence that we are, in fact, experiencing the after-effects of a balance sheet recession, examine the relationship of our own economic woes to Japan’s, consider the implications of a balance sheet recession, and assess its implications for policy.
The graphic above is from the St. Louis Reserve’s FRED system and illustrates the growth in household debt from 1950 to the present. The graphic below, from the San Francisco Fed, juxtaposes U. S. household debt to income and Japanese nonfinancial corporate sector debt to GDP over 10-year periods before and after the leverage-ratio peaks. I have been unable to locate the data on actual post-war Japanese nonfinancial corporate sector debt. I think it would be very interesting.
The first graph is astonishing to me for how perfect an example of exponential growth it presents. Growth proceeds virtually at a fixed rate up to the point at which it becomes asymptotic, then declines sharply. Anything that cannot be sustained will not, indeed. In the second graph the Japanese experience presents nearly as many contrasts with our own as similarities. Yes, the graph goes up, then down. There the similarities end. Unlike the smooth exponential growth expressed in the U. S. instance, the growth in Japanese debt to GDP is more irregular and exhibits a very different pattern. Whatever is happening here differs from Japan rather dramatically.
Even if the analogy holds, I don’t think we should take much solace from it. Twenty years later Japan continues to be burdened with slow growth. The Japanese have tried Keynesian stimulus, massive infrastructure projects, and quantitative easing much as we have. None of these have reawakened growth in Japan.
Consider, too, the downturn in household debt here. Nearly all of that decline is due to foreclosures. Real retail sales have recovered by nearly three quarters while; real disposable personal income has continued unimpeded throughout the downturn. If there is balance sheet rebuilding here, I’m not seeing it.
Additionally, how far would household balance sheets need to recover? I have seen a return to 1990s levels bandied about: we haven’t even begun such a process. And why is that the right level? Why not the 1970s? The straight line curve-fitting of the 1970s is even better than that of the 1980s and 1990s and to my eye both look like partial segments of what is obviously an exponential growth curve.
Thirty five years ago 54.4% of our economy was based on personal consumption. Now 77.3% is. To return to 1990s levels let alone 1970s levels over anything but the very long term would constitute economic collapse. I conjecture that a considerable portion of that excess consumer spending is healthcare spending.
If we have, indeed, experienced a balance sheet recession what is the appropriate policy response? The Japanese experience isn’t helpful in that regard. Debt forgiveness? (something I’ve mentioned here before) More and faster foreclosures? We are a very different society than the extremely homogeneous and consensus-driven Japan. I don’t think we can stand 20 years of slow or no growth household balance sheet rebuilding peaceably. Can stand ten?
I think there’s another interpretation that I think should be considered: we have reached the end of a 70 year period during which a debt-financed model of economic activity has been pursued. It did not prove sustainable and it is not being sustained.