Why This U.S. Recovery Is Weaker
There is a vigorous debate over whether the U.S. economy’s recovery from the recent deep recession and financial crisis is weaker than similar events in economic history.
In work with Joseph Haubrich, an economist at the Federal Reserve Bank of Cleveland, I have argued that this recovery is unusually weak compared with previous episodes. The Harvard University economists Carmen Reinhart and Kenneth Rogoff say the pace of the current recovery is consistent with the aftermaths of serious U.S. financial crises of the past.
This is more than simply an academic debate. If, as Haubrich and I contend, this recovery is unusually slow, then the policies of President Barack Obama’s administration must bear some of the responsibility. (I would like to make clear that I am not affiliated with Republican Mitt Romney’s presidential campaign, though I was among 670 prominent economists who signed a letter of support for the candidate.)
Reinhart and Rogoff criticized my work with Haubrich in a Bloomberg View op-ed article last week and in a subsequent article. It is important to set the record straight on our research and its differences with Reinhart and Rogoff’s 2009 book, “This Time Is Different: Eight Centuries of Financial Folly,” and other publications.
We wanted to answer this question: Are recoveries from recessions associated with financial crises weaker than recoveries from recessions that were not accompanied by financial crises? To do this, we examined U.S. business cycles from 1880 to the present.
We were driven by two ideas: to determine if the current recovery is weak because it followed a financial crisis in 2008, as Reinhart and Rogoff and others assert, and to see whether Milton Friedman’s plucking model of business cycles -- which posited that the strength of a recovery depended on the depth of the preceding recession -- holds up against the evidence of the past 130 years.
What we found was that Friedman’s theory held up and that historically the U.S. economy has bounced back faster from deep recessions accompanied by financial crises than from those without financial crises. In addition, we found that the weakness of the current recovery is a major departure from the historical pattern of recessions with financial crises.
The methodology we used came from a long line of research on business cycles going back to Friedman’s work in the 1960s and Victor Zarnowitz’s 1992 book, “Business Cycles: Theory, History, Indicators, and Forecasting.”
We looked at the recovery after the trough of the business cycle and focused on the subsequent pace of economic growth for a period that matches the duration of the recession. With the 1907 crisis, for example, the economy contracted for four quarters, so we measured real gross domestic product growth for four quarters of recovery; the Great Contraction lasted 14 quarters starting in July 1929, and we measured real growth over 14 quarters of recovery.
So how does that compare to the present? The recent recession lasted six quarters until its trough in the summer of 2009, as designated by the National Bureau of Economic Research, so we looked at GDP growth for six quarters.
By contrast, Reinhart and Rogoff focus on the behavior of real per capita GDP from the peak preceding the financial crisis to the point in the succeeding recovery at which the earlier peak of real per capita GDP is reached again.
Our conclusion -- that recoveries after deep recessions accompanied by financial crises are faster, with the notable exception of the recent experience -- is even clearer when we exclusively focus on what Reinhart and Rogoff call systemic crises, such as 1893 and 1907. In addition, our findings were the same even when we used per capita real GDP as they do and when we looked at the global financial crises demarcated in their book.
The crux of the difference between us is that Reinhart and Rogoff combine the downturn with the recovery and argue that our method is based on arbitrary measures of business-cycle dating. In the U.S., however, the dating of recessions and recoveries is done by the NBER’s Business Cycle Dating Committee, which most economists would agree has a pretty good track record. Moreover, several well-known statistical algorithms that aren’t based on human decision-making seem to track the NBER business cycles quite well.
Nonetheless, there are legitimate concerns over whether the NBER’s dating of business-cycle peaks and troughs is the right measure for the analysis of the recovery from a deep recession associated with a systemic financial crisis. That is a contentious issue and should be the subject of future research.
For now, however, comparing their findings with ours is like comparing apples with oranges.
Our approach focuses directly on the question of whether recessions with financial crises are associated with slower or faster recoveries. Their approach answers the very different question of whether systemic financial crises, like the recent one, lead to deeper and longer lasting declines in economic welfare (measured by the level of per capita real GDP) than occur in normal business cycles.
(Michael Bordo is professor of economics at Rutgers University and a fellow at Stanford University’s Hoover Institution. The opinions expressed are his own.)
To contact the writer of this article: Michael Bordo at email@example.com.
To contact the editor responsible for this article: Max Berley at firstname.lastname@example.org.