The Business Cycle Dating Committee of the National Bureau of Economic Research (NBER) sets the official dates for the beginning and end of recessions. One might imagine that these are not exciting meetings, but the next few may be livelier. That's because the major indicators they follow to determine whether the economy is in expansion or is contracting are sending off conflicting messages. One of the major indicators that the group follows is consistent with an economic recovery. One is unimpressive, but not strongly at odds with a recovery. The two remaining indicators imply that the economy may still be in recession.

This divergence has led committee members to express different views of where the economy is in the business cycle. Following the November employment report (but before last week's disappointing job loss) Robert Hall, who currently heads the Business Cycle Dating Committee, said that the report "makes it seem that the trough in employment will be around December. The trough in output was probably sometime in the summer. The committee will need to balance the mid-year date for output against the end-of-year date for employment."

Earlier this month Martin Feldstein, the former president of the NBER, focusing on the job market, said that "the current downturn is likely to last much longer than previous downturns ... We will be lucky to see the recession end in 2009."

If we take a look at the primary components of the economy that the NBER tracks to determine periods of recession, this uncertainty about the end of the recession becomes understandable. The NBER defines a recession as 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." The four primary indicators that the group tracks are: Industrial Production, Real Manufacturing and Trade Sales, Real Personal Income Less Transfer Payments, and Nonfarm Payrolls. The graphs below show how each of these measures perform around the end of a recession, along with recent performance. The blue line is an average of the performance in each measure, beginning with the 1953 recession (where the data is available). The red line tracks the performance of each measure during this business cycle. The black line marks the end of each recession. For this analysis, we'll assume the recession ended last June, which seems to be the consensus of Wall Street analysts.

Industrial Production is the first of two indicators that measure the economy's output. Industrial Production rose .8 percent last month from the previous month and now has risen or remained unchanged for 5 consecutive months. The graph shows that while the contraction in production was more severe during the current recession (as all the indicators will show), the recent rise in output is in line with prior recoveries, though recent momentum has been uneven.

Manufacturing and Trade Sales send a similar message. This metric bottomed in June and has advanced modestly since then, although at a slower rate than during prior recoveries. The implication of measures that track manufacturing and output is that economic activity may have bottomed last summer, but with uneven progress compared with previous recoveries. Measures that track employment and household income are even less clear.

While the Real Personal Income Less Transfer Payments Index has stopped declining, it has been mostly moving sideways since April. Flat is better than declining, of course, but the lack of clear upward progress may make determining the end of the recession trickier. In its note announcing the trough of the 2001 recession, the committee remarked, "the NBER's practice has been that if economic activity is roughly flat at the end of a recession or expansion, the turning point is placed at the end of the flat period."

The weakest measure of the four primary indicators that the NBER considers is employment. The NBER's preference for tracking the health of the jobs market is to use Non-Farm Payrolls. The graph below shows how non-farm payrolls typically respond around the end of a recession, compared with the data's recent performance. As the graph shows, it is not typical for employment to contract this persistently during a new recovery. While there is sometimes a lag between the end of a recession and the beginning of an expansion, it tends to be quite small. Looking at the recessions prior to 2001, payrolls typically began expanding just one month after the trough of the economy.

Prior announcements by the business cycle dating committee may shed some light on their considerations of income and employment in determining the end of the current recession. In 2003 the group wrote, "Identifying the date of the trough involved weighing the evidence provided by the behavior of various indicators of economic activity. The estimates of real GDP ... are only available quarterly. Further these macroeconomic indicators are subject to substantial revisions and measurement error. For these reasons, the committee refers to a variety of monthly indicators to choose the exact months of peaks and troughs. It places particular emphasis on real personal income excluding transfers and on employment, since both measures reflect activity across the entire economy."

If the group still places "particular emphasis" on personal income and employment, it's difficult to confidently conclude that the recession is over. And even if it is, stock investors can be sensitive to continued weakness in the job market. The graph below shows the changes in non-farm payrolls around the 2001 recession, a period that also witnessed persistent flat or declining employment after the recession was officially declared over.

The current job market is closely tracking the employment situation in 2001. Here it's important to keep in mind that although the recession was declared to have ended in November of 2001, the bear market in stocks resumed after a brief period of strength, and ultimately continued for another 15 months. The bear market finally ended about 5 months prior to a sustained improvement in the job market.

Even if the recession is eventually determined to have ended last year, the NBER may take their time in dating it. In each of the last three recessions, because the primary metrics they follow were mixed, the group waited until GDP increased to a new high - or was forecasted to reach a new high in the current quarter - to declare that the recession had ended. With the signals clearly mixed here - manufacturing and output indicators rising and household income and job prospects moving sideways or contracting - the group may also take a wait-and-see approach. With real GDP still down more than 3 percent from its peak, it would go against precedent for the group to declare the end of a recession with the mixed signals that are currently in place.