Summary: U.S. payroll employment or industrial production growth for one quarter will suggest the recession is ending.

(June 5, 2009) The most frequent question we hear is: ‘When will the recession end?’1 Last May 2 we announced the U.S. economy entered recession in January 2008. Our forecast missed by one month. The National Bureau of Economic Research declared Dec. 1, 2008, that economic activity peaked in December 2007. 3

Looking in the Wrong Places

Many observers look to GDP and the stock market for evidence of recession, and expansion. Observers fixated on GDP missed the December 2007 peak because the indicator had not contracted for two consecutive quarters, a common recession definition that is inaccurate. The NBER’s Business Cycle Dating Committee states:

A recession is a significant decline in economic activity spread across the economy, lasting more than a few months, normally visible in production, employment, real income, and other indicators.

One FAQ posted on the NBER site notes the “financial press often states the definition of a recession as two consecutive quarters of decline in real GDP.” How does that relate to the NBER’s recession dating procedure?” The Committee’s answer:

Most of the recessions identified by our procedures do consist of two or more quarters of declining real GDP, but not all of them. As an example, the last recession, in 2001, did not include two consecutive quarters of decline. As of the date of the committee’s meeting, the economy had not yet experienced two consecutive quarters of decline.

Observers fixated on the stock market as an indicator make the same mistake. The Dow traded between 13,059 and 13,850 in the month the recession started (December 2007), not far from its peak of 14,279 (Oct. 11, 2007). The Dow still traded above 13,000 on the day in May 2008 when the Foundation issued its recession call, and did not close below 10,000 until 2008-4Q.

Relying exclusively on GDP and the Dow is akin to a hitter who expects only fastballs and curves, and discounts sinkers, changeups or sliders.

Key Indicator: Payroll Employment (First Friday)

The Bureau of Labor Statistics releases its monthly “employment situation summary” on the following first Friday (7:30 a.m. CST). Payroll employment is considered a coincident indicator, i.e. it tends to peak or trough with cyclical turning points in the economy in the postwar era. Employment peaked in December 2007, and has contracted for 17 consecutive months. Exceptions were expansions following the last two recessions (1990-91, 2001). Slow job growth and late employment troughs (May 1991, August 2003) led some observers to use the term “jobless recovery.”

Employment peaked in December 2007, and has contracted for 17 months. Today’s report showed a decline of “345,000 in May, about half the average monthly decline for the prior 6 months.”

Key Indicator: Industrial Production (Mid-Month)

The Federal Reserve releases its monthly industrial production index mid-month. Production, another coincident indicator, measures the physical output of the nation’s factories, mines and utilities. Our March 2002 memo noted: “Postwar, in 90 percent of recessions, industrial production stopped contracting prior to or in the month later declared by the NBER as the end of the recession.” Industrial production peaked in December 2007, and has contracted for 15 of the last 16 months.


Employment and industrial production are more reliable than GDP and the Dow in determining cyclical turning points in the economy.

— Greg Kaza

1 Today parallels 2002-1Q when a Foundation memo noted the most frequent question asked of (us) is ‘When will the recession end?’ See “Today’s Industrial Production Data” (February 2002) and “Industrial Production Expands” (March 2002). The Foundation, in March 2002, announced the recession was over, and the NBER’s July 17, 2003, declaration set a November 2001 trough.

2 “Economist Claims Nation in Recession.” Fayetteville Morning News. May 17, 2008.

3 The NBER announcement is posted at: