« "Social Status of Members of Congress Shifts Policy to Rich" | Main | Shovel Ready »

Friday, September 16, 2011

Why There Is No ‘V’ Rebound This Time

Before looking at some worrisome comparisons in the Economic Letter from Mark Wynne of the Dallas Fed indicating we are in for a protracted recovery, it's probably good to remember this from Charles Evans:

It bears keeping in mind that ... predictions of a slow recovery are based on historical averages of macroeconomic performances across many different countries... They highlight a challenge we face today, but from the standpoint of the underlying economic analysis, there is nothing pre-ordained about these outcomes. They are not theoretical predictions—rather, they are reduced form correlations. The economy can perform better than it did in these past episodes if policy responds better than it did in those situations.

Here's the Letter:

The Sluggish Recovery from the Great Recession: Why There Is No ‘V’ Rebound This Time, by Mark A. Wynne, Economic Letter, Vol. 6, No. 9, September 2011, Federal Reserve Bank of Dallas: The Great Recession of 2008–09 was by far the most severe United States economic downturn since the Great Depression of the 1930s. Real gross domestic product (GDP), the most comprehensive measure of U.S. economic activity, topped out in fourth quarter 2007 and has yet to approach that peak. Employment totaled just below 138 million jobs in January 2008 and, as of July 2011, was still nearly 5 percent below its precrisis level.
Conventional wisdom holds that severe recessions are usually followed by strong recoveries. This belief goes by many names. Milton Friedman termed it the “plucking theory” of business fluctuations, likening recessions to down plucks on a guitar string.[1] The essential insight is that the harder you pluck down, the faster the string snaps back to its original position. In the wake of the early 1980s recession, economist Alan Blinder wrote about the “Joe Palooka effect” following recessions, named after the comic book boxer of the same name who could take a punch.[2] In a series of papers I wrote with my colleague Nathan Balke after the 1990 recession, we documented this strong recovery phenomenon using time-series data on U.S. industrial production going back to the 19th century.[3]
Yet many have argued—and the facts seem to support—that we are not seeing the robust recovery from the most recent downturn that we might have expected given the recession’s severity. That is, the plucking theory, or bounceback effect, seems absent. Why is that? What is different this time that might cause a more protracted recovery?
Recessions and Recoveries
There are many ways to characterize the onset and ending of the ups and downs in economic activity known as the business cycle. The National Bureau of Economic Research (NBER), the official arbiter of when recessions begin and end, maintains a chronology of peaks and troughs in U.S. economic activity that extends back to the 19th century. The NBER chronology seeks to identify turning points in general economic activity rather than a transition in a particular economic measure. The selection of the turning points is based on the behavior of key indicators (such as employment, sales, production and income), with the weighting of each indicator reflecting the judgment of the committee that determines the dates. 
An alternative method of characterizing the business cycle is to examine the behavior of a single comprehensive measure of economic activity—such as GDP—and, furthermore, to define the cycle as movements in this indicator relative to some measure of its long-term trend. This is the approach favored by contemporary business-cycle theorists.
Chart 1 shows the behavior of real (inflation-adjusted) GDP relative to its long-run trend from 1947 to second quarter 2011.[4] When output is below trend, the deviations are negative. Note that these episodes tend to match up closely (albeit not perfectly) with NBER’s recession episodes (shown as shaded bars). For example, GDP was well below trend in every quarter of 1958, and in the last quarter of 1960 and the first three quarters of 1961. The NBER chronology notes recessions occurred between August 1957 and April 1958, and between April 1960 and February 1961. Fourth quarter 1982 is identified as a major downturn, matching the NBER’s recognition of November 1982 as a trough. And finally, the chart confirms the most recent episode as a major slump.

Chart 1: The Bounceback Effect in GDP

More importantly, the chart neatly illustrates the V-shaped recoveries that follow the most severe downturns. We see V-shaped recoveries after second quarter 1958, first quarter 1961, fourth quarter 1982 and even after first quarter 1975, which is not identified as a severe recession by this metric. We do not see as strong a recovery after the 1990 recession, although this was a milder downturn, as was the 2001 recession, which does not show up as a recession on this chart’s GDP plotting.
No V-shaped recovery is apparent after the most recent slowdown. The NBER reported the recession’s end as June 2009, and the chart shows that GDP was about 6 percent below trend in the second and third quarters of that year. Note that despite the severity of the downturn, real economic activity has not recovered at anything resembling the pace of previous recoveries.
This Time Is Different
Unlike all other post-World War II recessions, the 2008–09 episode was precipitated by a banking crisis. A number of researchers have shown that downturns associated with banking crises tend to be more severe, and furthermore, in their aftermath, output takes a lot longer to recover.[5] In some cases, the crisis seems to persistently affect the trend rate of growth, while in other cases, the growth path of activity seems to shift down.
Chart 2 is a summary of the average impact of financial crises on output. It shows average deviation of output from its trend path in a sample of countries that experienced banking crises from the early 1970s to 2002, along with a measure of the range of outcomes (shaded area).[6] During the first year of a banking crisis, output falls by about 2.5 percent on average and then slips further in subsequent years. The persistent decline in output relative to the precrisis trend is striking. The finding that banking crises tend to have persistent effects on output is robust to alternative definitions.

Chart 2: GDP Falls Below Trend After Banking Crises

How does recent U.S. experience compare? It depends to some extent on when we define the crisis start. Conventional wisdom holds that it began in August 2007, and the NBER dates the business-cycle peak in December 2007. Chart 3A overlays the recent behavior of U.S. real GDP relative to its precrisis trend along with the data plotted in Chart 2, taking 2007 as the first year of the crisis. Here we are just looking at annual GDP numbers. The numbers for 2011 and 2012 are based on projecting the 2010 number forward using the September 2011 Blue Chip Economic Indicators consensus. It is striking how closely the path of U.S. real GDP trend tracks the average path of output in countries that have experienced banking crises. In that sense, the pace of the recovery is more or less in line with what we might have expected based on the historical experience of other countries that have undergone similar banking calamities.

Chart 3: U.S. Postcrisis GDP Resembles Average Experience

However, reasonable people might argue that the crisis really started in 2008, when major U.S. financial institutions got into serious difficulties, ultimately prompting major policy initiatives from fiscal and monetary authorities to help stabilize the economy. Does this date change things?

Chart 3B shows how the comparison is affected if we take 2008 as the beginning. If anything, the fit to the historical patterns observed elsewhere is better. That is, the performance of real GDP in the U.S. is almost exactly in line with what we might have expected based on the average experience of other countries that have gone through banking crises.

Why does output tend to stay below its precrisis trend path in the aftermath of recessions associated with banking difficulties? There is little or no consensus on this issue. Banking crises tend to have persistent effects on productivity, the employment rate, investment and the capital-labor ratio.[7] Fortunately, there is little evidence of a persistent impact on growth rates: Most countries experiencing banking crises tend to return to their precrisis rates of growth over time.

Recessions and Banking Crises
While the U.S. economy has been in a recovery for almost two years, the pace has been unusually weak by the country’s historical standards. Rather than seeing the V-shaped recovery that might have been expected given the severity of the downturn, the nation is undergoing a more protracted process. However, when viewed in a broader international context, the pace of the recovery seems to be very much in line with what other countries have experienced. The persistence of the output losses associated with banking crises should serve as additional motivation—if any were needed—for preventing recurrences in the future.


The author thanks Adrienne Mack for her research assistance.

  1. See “The National Bureau Enters Its 45th Year,” by Milton Friedman, 44th Annual Report, National Bureau of Economic Research, 1969, pp. 7–25. (Reprinted as “The Monetary Studies of the National Bureau,” in The Optimum Quantity of Money, by Milton Friedman, New Brunswick, N.J.: Transaction Publishers, 2009, pp. 261–84.) See also “The ‘Plucking Model’ of Business Fluctuations Revisited,” by Milton Friedman, Economic Inquiry, vol. 31, no. 2, 1993, pp. 171–77.
  2. “The Joe Palooka Effect,” by Alan S. Blinder, Boston Globe, Nov. 13, 1984.
  3. See “Are Deep Recessions Followed by Strong Recoveries?” by Mark A. Wynne and Nathan S. Balke, Economics Letters, vol. 39, no. 2, 1992, pp. 183–89; “Recessions and Recoveries,” by Nathan S. Balke and Mark A. Wynne, Federal Reserve Bank of Dallas Economic Review, First Quarter 1993; and “Are Deep Recessions Followed by Strong Recoveries? Results for the G-7 Countries,” by Nathan S. Balke and Mark A. Wynne, Applied Economics, vol. 28, no. 7, 1996, pp. 889–97.
  4. The trend is measured as a simple least-squares regression of the log of real GDP on a constant, a deterministic trend and the square of the deterministic trend.
  5. See, for example, “Growth Dynamics: The Myth of Economic Recovery,” by Valerie Cerra and Sweta Chaman Saxena, American Economic Review, vol. 98, no. 1, 2008, pp. 439–57; and “The Effect of Financial Crises on Potential Output,” by Davide Furceri and Annabelle Mourougane, Organization for Economic Cooperation and Development, Economics Department Working Paper no. 699, May 2009. See also chapter 4 of World Economic Outlook, International Monetary Fund, October 2009.
  6. The data plotted here are from chapter 4 of World Economic Outlook, International Monetary Fund, October 2009, specifically the data plotted in Figure 4.15, alternative 1.
  7. Again, see chapter 4 of World Economic Outlook, International Monetary Fund, October 2009.

The views expressed are those of the authors and should not be attributed to the Federal Reserve Bank of Dallas or the Federal Reserve System.

    Posted by on Friday, September 16, 2011 at 03:33 AM in Economics | Permalink  Comments (25)


    Feed You can follow this conversation by subscribing to the comment feed for this post.