Economics One

Paul Krugman disagrees with my recent post that the recovery is weak compared to recoveries from past serious U.S. recessions including those associated with financial crises. I’ve been writing about the reasons for weak recovery for two years, but the issue has heated up because of its relevance to the elections this fall.

In making his critique, Krugman appeals to a recent oped in Bloomberg View by Carmen Reinhart and Ken Rogoff who criticize the research of economic historian Michael Bordo and his coauthor Joseph Haubrich of the Cleveland Fed, which I have referred to. Bordo and Haubrich demonstrate that the recovery from the recent recession and financial crisis has been unusually weak compared to recoveries from past recessions with financial crises in the United States. In separate research, Jerry Dwyer and Jim Lothian report the same finding. Neither Reinhart-Rogoff nor Krugman disprove this finding.

To see this, consider the points made by Reinhart and Rogoff, and also by Krugman.

First, they argue for a narrower definition of a financial crisis. Reinhart and Rogoff say that one should “distinguish systemic financial crises from more minor ones and from regular business cycles.” Thus they exclude some cases studied by Bordo and Haubrich. But narrowing the focus to systemic crises in this way does not change the Bordo-Haubrich findings because the recovery from the recent recession is weaker than the average of past recessions cum financial crises even with these exclusions.

Second, Reinhart and Rogoff argue that one should look at recessions together with recoveries when looking at severity. In fact, in their work they explicitly “don’t delineate between the ‘recession’ period and the ‘recovery’ period.” But there is no disagreement that recessions associated with financial crises have tended to be deeper than those without financial crises. I certainly don’t deny that there was a serious financial crisis. In fact I wrote one of the first books on the crisis and found that government policy prior to 2009 was to blame, just as government policy was to blame for the even more serious Great Depression.

The issue that I and others have focused on is whether the recovery is unusually weak. By mixing recessions with recoveries Reinhart and Rogoff blur the classic distinction, which has long been at the heart of macroeconomic analysis. Because they do not examine recoveries per se, their empirical analysis does not disprove the fact that the current recovery is very weak as Bordo and Haubrich and others have shown.

Third, there is the complaint that in a simple chart I used to show how weak the recovery has been, I only looked at the first four quarters of recovery. But I also mentioned that Bordo and Haubrich use a different measure, which goes well beyond 4 quarters, and come to the same conclusion, and also that the current recovery has weakened further since the first four quarters. In any case, the focus on four quarters has nothing to do with it. Here is a chart that looks at growth during the first eight quarters. The story is the same.

Since most of the Reinhart, Rogoff and Krugman criticism is implicitly aimed at the historical work of Bordo and Haubrich, it is appropriate to conclude with what Bordo wrote in response to a press inquiry which he shared with me. Bordo puts it this way:

Aside from the unnecessary political rhetoric and ad hominems, the basic difference between my research with Joseph Haubrich on U.S. recoveries and that of Carmen Reinhart and Ken Rogoff is over the methodology of defining a recovery. Reinhart and Rogoff focus on the behavior of the level of real per capita GDP from the peak preceding the financial crisis to the point in the succeeding recovery at which the earlier peak level of real per capita GDP is reached. We look at what is called the bounce back, the pace of recovery from the trough of the business cycle.

We find that deep recessions accompanied by financial crises bounce back faster than recessions which do not have financial crises. These results are even stronger when we focus on what Reinhart and Rogoff call systemic crises, like 1893 and 1907. The recent recession and financial crisis is a major exception to this pattern. The recovery remains tepid after three years.

Reinhart and Rogoff s methodology combines the downturn with the recovery. Using our data but following their approach one would get the same results as they do, that recessions with financial crises have slow recoveries as I show in my Wall Street Journal op ed. Their use of real per capita GDP rather than just real GDP would not make any difference to our results. Thus comparing their methodology with ours is like comparing apples with oranges. Our approach focuses directly on the question– are recoveries after recessions with financial crises associated with slower or faster than average recoveries. Their approach answers a different question than we ask.

In sum, the weak recovery deniers have not made their case.

The weak recovery continues to be a major topic. Over the weekend, Russ Roberts issued the second episode of his three part “chartcast” series on the topic, which is based on interviews with me and builds on his highly-regarded podcast series, but with helpful charts and illustrations. (Here is the first episode). Among other things we discuss the work of Mike Bordo and Joe Haubrich on the unusual nature of this recovery, and their views of the work by Carmen Reinhart and Ken Rogoff.Also, as Jon Hilsenrath and Ezra Klein report, over the weekend Carmen Reinhart and Ken Rogoff released a short rebuttal to Bordo and Haubrich as well as to several opeds. Reinhart and Rogoff argue in favor of a narrower definition of a financial crisis, and they thus focus on a subset of the eight Bordo-Haubrich recessions with financial crises (for example, they exclude 1913 and 1982). This alone does not change the Bordo-Haubrich results as the figure in my post of last week makes clear. But Reinhart and Rogoff argue that one should look at the downturn as well as the recovery when looking at severity. There is no disagreement that recessions associated with financial crises have tended to be deeper than those without financial crises. The disagreement is over the recoveries. By mixing downturns with recoveries Reinhart and Rogoff get different results from Bordo and Haubrich.

But the question for policy now is whether the recovery has been unusually slow compared to earlier recoveries from recessions with financial crises, and the evidence is still clear that it has been. Papers in the book Government Policies and the Delayed Economic Recovery edited by Lee Ohanian, Ian Wright, and me show that policy is the reason.

Twenty years ago this month my colleague Bob Hall and I wrote an op-ed for the New York Times about how “in recent months press reporting about the economy has become so pessimistic that it has completely lost touch with reality.” (October 16, 1992). The Times editors headlined our article “Economic Scare Stories,” which captured our point perfectly and fit the Halloween season. In October 1992, economic growth was improving following the 1990-91 recession, but most reporting looked beyond good economic news and said that the economy was doing poorly. Amazingly, the frequently-reported view that the economy in October 1992 was like the Great Depression went unchallenged. So we challenged it, hoping that our article would in some small way result in improved reporting.

Today press reporting seems to have switched to the other side of reality. Compared to October 1992, economic growth is now slower, unemployment is higher, and tragically the long-term unemployment rate is twice has high. And reported economic growth has been declining rather than improving as it was in 1992. Yet, in recent months much reporting about the economy has turned so upbeat that it has again lost touch with reality. Many look beyond the tragic growth or employment news and say that the economy is improving, or that things could have been worse, emphasizing that it is fortunately nothing like the Great Depression.

When asked what caused the switch, I answer, facetiously, that people must have read our article, remembered it, tried to make a correction, but unintentionally overcorrected. That answer, of course, is out of touch with the reality that both October 1992 and October 2012 constitute the final days of a presidential election where the main issue is the economy, and, as Bob Hall and I wrote, “people’s perceptions about the economy affect elections.”