Employment Data Fails the Test as an Indicator

Friday, the market-watchers continued a pattern of obsessing over the employment numbers with all the accompanying hand-wringing over how the market might respond. Funny thing is, this pattern of analysis is backward. Generally (though not always), stock market improvement leads the economy, which in turn leads most employment metrics.

There is one employment data point, the peak of initial unemployment claims, that does serve as a modestly useful leading economic indicator. However, both the trough in total employment and the unemployment rate are badly lagging economic indicators of both the economy and the stock market.

Looking at the last six recessions (1969-2001), we analyzed when the economy bottomed (as measured by the "official" National Bureau of Economic Research recession trough) and compared that trough to when various metrics hit their worst levels. We constructed the table below to show the results:

Leads and Lags of Recession Bottom
Recession Bottom	Initial Unemploy Claims Peaked	Employment Trough Occurred	Unemployment  Rate Peaked		ISM New Orders Bottomed
November 1970	6 months before	Same time	9 months after		Same time
March 1975	1 month before	1 month after	2 months after		3 months before
July 1980	1 month before	Same time	Same time		1 month before
November 1982	1 month before	3 months after	One month after		12 months before
March 1991	Same time	4 months affer	15 months after		2 months before
November 2001	1 month before	21 months after	19 months after		1 month before
Average	1.7 months before	4.8 months after	7.7 months after		3.2 months before

Our conclusion is that at the bottom of an economic cycle, the initial jobless claim peak tended to lead the economy by a month or two. The employment trough, however, lagged the economic bottom by almost five months, and the unemployment rate pulled up the rear, lagging the economy by almost eight months.

This pattern makes a lot of sense if you think about it. As the economy starts to stabilize, the first thing employers do is nothing. Corporations initially stop the layoffs but don't hire more people, either. As the economy further improves, businesses begin to add jobs, improving that metric. However, as potential employees see things improve, previously discouraged workers dust off their resumes and re-enter the employment pool, causing unemployment to go up. (Remember, unless someone is actively looking for work, they cannot be counted as unemployed.)

As logical as this sounds, others might argue that if people don't have jobs, they can't spend money, and the economy won't improve. Certainly there is some feedback in terms of lost jobs and spending, but this analysis misses one key point. In most of recessions, more than 90% of people who want a job have a job. How those employed people are compensated and what they believe is the prognosis for their own job will determine their spending. The actions of the employed majority will be far more important than a percent change in the overall unemployment rate. The table above tends to validate the argument that employment is not the key driver of the economy in the short run.

While all three of the major economic statistics are showing a decreasing rate of decline, none, not even the initial claims, has bottomed. We will continue to monitor initial unemployment claims closely, every Thursday morning when they are released. Remember, though, the lead time on this metric is very short.

There are a number of indicators that are more helpful than the employment measures. One of our favorites has been the New Orders Index from the Institute of Supply Management (ISM). As shown in the table above, this metric has led the trough in the economy by about three months and has always been a leading indicator in each recession. This metric bottomed in December 2008 at 23.1, improved in January, was flat in February, and was over 41 for the March reading. Some other indicators are also showing improvement, but this is one of the more dramatic ones and is one of the reasons we believe the economy could easily hit its bottom by June.

DIV style="MARGIN-TOP: 0px; FLOAT: right; MARGIN-LEFT: 7px; MARGIN-RIGHT: 0px" block? >