Recession Dating: When in Doubt, Wait for a Crisis

REAL TIME ECONOMICS Real Time Economics

Economic insight and analysis from The Wall Street Journal.

December 1, 2008, 7:04 pm
Recession Dating: When in Doubt, Wait for a Crisis


Travel back to the first half of the year, after an asset bubble burst, and you’ll find a group of economists struggling with the question of whether the nation is in a recession. Payrolls hit a clear peak and started declining, the unemployment rate was rising and Americans were talking increasingly about a troubled economy. The common recession definition — two straight quarters of falling economic output — was far from being met, since official economic data hadn’t even shown a single quarter of contracting GDP yet.


Then a major international crisis hits, economic data start looking uglier and the recession question is effectively settled.Those were the circumstances of the U.S. recession call in 2008 — and in 2001. The worsening of the financial crisis in September proved to be the clarifying event that the 9/11 terrorist attacks became seven years earlier for the National Bureau of Economic Research’s Business Cycle Dating Committee. In both cases, GDP releases (or revisions of earlier figures) that weren’t available in the first half of the year later showed that a contraction clearly was in the works. But making the recession call in real time proved difficult based only on output. So the committee used nonfarm payrolls to identify the recession’s starting point in both cases.


The NBER committee, the nation’s generally accepted recession arbiter, restarted its activity this year after the labor market started contracting. But a few months of mild payroll declines weren’t enough to call a recession without a single quarter of declining GDP, given the chance that the economy could turn around quickly. (The fourth quarter 2007 contraction didn’t show up until the government’s GDP revisions came out in late July.) Then September 2008 hit, promising to worsen the housing and credit crises.


“Employment declined less than is normal in a recession until about September,” said Stanford University economist Robert Hall, chair of the NBER’s recession-dating committee. Until that point the committee didn’t expect to take action until next year, Mr. Hall said in an interview today. “Then so many negative numbers came through that made it completely clear this was a recession.”


Once that was established, the key task became identifying the precise starting point for the NBER’s monthly chronology. “The sudden worsening of the financial crisis since the end of the summer obviously sent the economy into a nosedive,” said Harvard University economist Jeffrey Frankel, a committee member. “There’s no doubt for a few months we’ve been in recession. The only question is whether you want to count the shallower period.”


The committee’s explanation of its decision today ran through the various measures it generally considers (including real personal income, wholesale-retail sales and industrial production), noting flaws in just about every one of them except employment. GDP fell in this year’s third quarter and the fourth quarter of 2007, but not in the two quarters in between. Gross domestic income peaked in the third quarter of 2007, fell in the two quarters that followed, then rose for a quarter and fell again in this year’s third quarter. The different figures — whose discrepancies (debated all year) eventually may be revised away next July — still suggest a rough plateau in economic activity from the third quarter of 2007 to the second quarter of 2008, Mr. Frankel said. “Any way of doing this is imperfect given how the two main output measures give such different answers,” he said.


With ambiguous GDP data, the payroll figures proved critical to calling the start of the recession just as they did in 2001. But if the committee follows the 2001 recession playbook again, employment won’t be key to marking the end of the recession. After a lengthy debate, the NBER committee dated the end of the recession to November 2001 even though payrolls continued declining into 2003. It explained its call in 2003: “From October to November, industrial production and sales fell sharply, employment fell moderately, personal income rose very slightly, and monthly real GDP rose moderately. Based on this information, the committee concluded that the economy reached a trough in November.” (The highly subjective nature of the call may be one reason the government never got involved in recession dating or even selected a group — the NBER or otherwise — for the task. But government economic statistics use NBER recession dates, making the committee the semi-official keeper of the nation’s business cycle chronology.)


GDP contractions continuing through June 2009 — as many economists forecast — would put the current recession at a year and a half, or two months longer than the worst two downturns since the Great Depression. With the economic outlook still so uncertain, we’ll probably have to wait until late 2009 or even 2010 before getting that call from the committee.


— Sudeep Reddy




Earlier on RTE:
Waiting For The Recession Call
A Recession Announcement Coming? Probably, But Not Soon
The GDP Debate: Did a Recession Start in 2007?
If It Is a Recession, the Longest in a Quarter Century?
Economists Weigh Possibility of a Recession Amid Economic Growth


Permalink Trackback URL: http://blogs.wsj.com/economics/2008/12/01/recession-dating-when-in-doubt-wait-for-a-crisis/trackback/



http://blogs.wsj.com/economics/2008/12/01/recession-dating-when-in-doubt-wait-for-a-crisis/?mod=rss_WSJBlog