San José State University
Department of Economics
Thayer Watkins
Silicon Valley
& Tornado Alley

Recession 2008:
A Matter of Terminology
and Timing

On the first of December 2008 the National Bureau of Economic Research (NBER) announced that according to their methodology the United States economy was in a recession and had been since December of 2007. The media immediately ballyhooed this announcement as "It's now official! We're in a recession!" The media did not reveal that the NBER's concept of a recession is different from the standard definition of a recession as a period in which the Gross Domestic Product has declined in real (inflation-adjusted and seasonally-adjusted) terms for two quarters in a row. Since at that time the real GDP had only declined for one quarter; i.e., from the second quarter of 2008 to the third quarter and then only by the statistically insignificant amount of 1/8 of 1% (according to the Preliminary Estimates from the Bureau of Economic Analysis) there was no basis for the declaration of a standard definition recession which refers to output.

The NBER is generally trying to identify a condition more general than a standard definition recession, something more in the nature of an economic malaise. This is a perfectly legitimate undertaking but the word recession should not be used for it. Recession is already taken to mean a receding of economic production. There is a desperate need for an expanded terminology for conditions of economic concern. In particular there needs to be a term to indicate a condition when unemployment is rising. If production is declining then definitely unemployment will be rising, but unemployment may rise even when production is not decreasing. In order to keep unemployment from rising the demand for labor must be increasing faster than the labor force. But the demand of labor will be increasing only if the level of growth of production (GDP) is increasing faster than the increase in labor productivity. Thus the GDP could be growing but unemployment increasing because the rate of growth of GDP is less than the sum of the growth rates of the labor force and labor productivity. Some people foolishly want to label that condition by the oxymoronic phrase growth recession. This is like talking about deflationary inflation.

The NBER has a specific general model of the economic that presumes there are internally generated cycles for the economy. It then seeks to identify peaks and troughs for the economy. The period from when the economy peaks to when it reaches a trough is the NBER notion of a recession. Likewise the period from a trough to a peak is a period of expansion. In practice the NBER cannot identify a peak when it occurs and typically waits until the economy is declining to say when the peak occurred and hence when a recession started according to their notion of a recession. There is much in this methodology that is patently absurd. The economy often reaches a plateau before going on to further growth. The economy could be on a plateau for some time and then experience an outside shock such as the passage of trade protection legislation in an important trading partner nation. The economy could experience a decline but that the period of the plateau may be completely unrelated to that decline but the NBER methodology would identify the beginning of the plateau period as the beginning of the recession. In many ways the NBER procedure is much like the procedures engaged in by the chartist investors in the stock market. There is no reason for the NBER dating of recessions and expansions to be considered official.

In the past the NBER's dating was just some harmless pontificating. In the present malaise it may be seriously harmful. In the past actual recession occurred because businesses decreased their investment in plant and equipment because they did not expect to need additional capacities. In the current situation the only major components of aggregate demand in the national income accounts that have been decreasing are residential housing construction and consumer purchases of motor vehicles. Business investment in plant and equipment had not been declining. Promoting the notion that there is an across-the-board decline in all economic sectors will decourage business investment and bring about a real decline in the economy.

Some people believe that because NBER looks at a number of different indicators besides real GDP that makes the NBER assessment of the economy superior to looking only at real GDP to define a recession. That makes no more sense than if the NBER made pronouncements about population growth by looking not only at population censuses by a variety of other statistics such as television sales, driver licenses and telephone lines. If one is concerned about population use the best population estimates. If there is information that could improve those estimates then they should be included. If NBER has a better index of economic production than real GDP then it should be published rather than mixing up notions of production and unemployment.

Update: March 2009

At this point there is an output recession occurring. However its onset can be dated to the financial crisis of September of 2008. Here is the record on the components of aggregate demand.

It is not until the third quarter that there is evidence of a downturn in production and sales for the economy. The component of demand that shows a long term downward trend is Investment. Here is the record of the detailed components of investment.

Residential housing construction has been following a declining trend since the fourth quarter of 2005. The crucial component is the one labeled Nonresidential. This includes the investment of businesses in plant and quipment. This did not show a downturn until the fourth quarter of 2008; only after months of a media frenzy of the declarations that the U.S. is in an output recession. With politicians declaring the outlook for the economy as dismal businesspeople would have a hard time convincing anyone that it would be wise to purchase a major piece of equipment at this time.

There has been an employment recession since December of 2007 but it may have its roots more in productivity increases rather than inadequate demand. However once the September financial crisis occurred then economic conditions changed significantly. Here is the record of total nonfarm employment for 2008 as compiled by the Bureau of Labor Statistics.

Employment is seen to be declining rather regularly from January to August. Then comes September and the declines accelerate significantly. This is displays more vividly in the form of the month-to-month proportional changes.

The NBER dating of a recession from December of 2007 is extremely misleading. The more significant downturn for employment came with the financial crisis of September. The output recession began with September. It had not existed before that time. For more on the nature of the financial crisis of September 2008 see September Crisis.

HOME PAGE OF applet-magic
HOME PAGE OF Thayer Watkins