The Employment Picture

I’ve been calling for a weak cyclical recovery in the US economy to begin either late in the third quarter or early in the fourth.

The most common question and push-back I receive concerning this projection concerns the employment picture. Many investors wonder, quite rightfully, how robust an economic recovery can be while many Americans continue lose their jobs and the unemployment rate continues to rise toward the psychologically key 10 percent level.

The answer to this seeming paradox is that historically some employment indicators tend to lead economic recoveries, others are coincident, and some are lagging. As I’ve noted in recent issues, the US employment picture is actually improving, but this is only true if you’re looking at the leading aspects of employment data.

Let’s take a closer look at the data to illustrate how it all fits together.

The most important employment report of all is the Bureau of Labor Statistics (BLS) monthly Employment Situation report that’s released on the first Friday of each month. This month, the report was released a day early because on Friday, July 3, the US observed Independence Day.

There are actually two major parts of the BLS report: Table A, based on the Household Survey, and Table B, based on the Establishment Survey. Data for the former is derived from a survey of roughly 60,000 households conducted each month. Meanwhile the eponymous Establishment survey collects data on 160,000 different businesses in both manufacturing and service industries.

One of the most often-quoted statistics from the monthly employment report is the unemployment rate, a statistic derived from the Household Survey within the BLS report. Here’s a chart of the unemployment rate going back to 1960.

Source: Bloomberg

The current seasonally adjusted US unemployment rate stands at 9.7 percent, the highest level since the late 1970s and early ’80s. I suspect this indicator will ultimately top out somewhere between 10 and 11 percent, if not higher.

However, take a closer look at the chart above. The last spike in unemployment came after the 2001 recession; the recession ended in November 2001, but the unemployment rate didn’t top out until June 2003. Similarly, the early ’90s recession ended in March 1991, but the unemployment rate didn’t top out until June 1992.

This lagging effect isn’t a new phenomenon. Look back at the vicious economic downturn of July 1973 through March 1975; the US unemployment rate hit its peak of 9 percent in May 1975.

Based on the historical record, however, it does seem fair to say that the lag between the end of a recession and the top in unemployment has lengthened somewhat in recent cycles.

If this trend persists, it would make the unemployment rate even more useless as a predictor of economic turns. Investors who wait for the unemployment rate to fall before buying stocks will find they’ve missed out on the sweetest part of the rally.

The unemployment rate is defined as the total number of unemployed divided by the labor force. This measure depends both on the measure of the total number of unemployed and on the measure of the size of the labor force. Many people that one might consider unemployed aren’t actually counted as unemployed by the BLS survey.

The reason is that to be unemployed one has to actually be looking for work. For example, a “discouraged worker” isn’t working, nor has he or she recently looked for work. But he or she does indicate that they want a job and have looked for work in the recent past.

Discouraged workers include potential workers that just don’t feel they can find full-time employment. Discouraged workers aren’t considered part of the labor force and, therefore, don’t effect the unemployment statistic.

Here’s a chart of the total number of discouraged workers as a percent of the total labor force.

Source: Bloomberg

Because data on discouraged workers only goes back to 1994, it’s difficult to make a call on the value of the indicator. One thing we can say is that in the cycle earlier this decade, the number of discouraged workers topped out in June 2003, long after the recession ended.

That being said, the index seemed to plateau beginning in late 2001, around the time the recession ended. This may be an interesting statistic to watch in coming months but it appears to be at best a coincident index, meaning that it turns at around the same time as the business cycle as a whole.

Although the popular press tends to focus on the unemployment rate, investors tend to watch the monthly change in non-farm payrolls more closely.

Source: Bloomberg

This non-farm payrolls data can look noisy if you chart too long a time frame, so I’ve only displayed data going back to 1989. As you can see, the month-over-month change in non-farm payrolls tends to bottom out around or just before the trough of major recessions. This is a leading indicator.

For example, the monthly change in non-farm payrolls bottomed out in October 2001 at around -390,000. This was about a month before the official November 2001 end to the last recession.

Of course, even after the end of the recession, the non-farm payrolls numbers saw a few prominent spikes to the downside. Such month-to-month noise is common.

Going back to the early ’90s the non-farm numbers bottomed out in February, a month before the end of the recession. Again, there were a few notable spikes in job losses even after that prominent low.

If the pattern holds, it would bode extraordinarily well for the economy this year. Note, in particular, how the monthly change in non-farm payrolls bottomed out in January of this year with between 700,000 and 800,000 jobs lost, a record figure.

Although much was made of the worse-than-expected June jobs report, the downside blip in that report doesn’t look at all unusual when compared to the 2001 and 1991 cycles.

Finally, a more timely indicator to watch is the weekly jobless claims data released every Thursday. This data comes from individual states that record the number of workers filing for first-time jobless benefits. It’s considered a good indicator of trends in the monthly data. I look at the four-week moving average of claims rather than week-to-week data as it offers a smoother pattern.

Source: Bloomberg

Once again, initial claims are a good leading indicator for the economy. The last spike in initial claims topped out in September of 2001, before the end of the 2001 recession. Back in 1991 claims data topped out about a month prior to the recession’s end.

You can clearly see that initial jobless claims data topped out in March and has been plummeting since in characteristic fashion. This indicator is also pointing at a recovery for the US economy at some point this quarter.

Don’t believe everything you read or see on TV. Although the jobs picture isn’t pretty right now, the data is actually pointing to a cyclical recovery, not continued recession. Short-term market pullbacks aside, this bodes well for an end-of-year rally for the S&P 500.

Speaking Engagements

Eight score and one year ago, with the onset of the California Gold Rush, San Francisco earned a reputation as a prospector’s town. It’s time again to seek paths to prosperity–and to enjoy one of the most beautiful natural settings in the US, if not the world.

Venture west for the San Francisco Money Show Aug. 21-23, 2009, at the The San Francisco Marriott and discover how Elliott Gue, Roger Conrad and GS Early can help you profit in these adventurous times.

Elliott will detail the new direction for Personal Finance and provide his forecast on energy markets for 2009. Roger will discuss utilities, Canadian income and royalty trusts as well as his new service focused on exploiting the greatest spending boom in history, New World 3.0. GS, a constant at PF for two decades, will be there to speak on emerging tech, nanotech and defense tech.

Click here or call 800-970-4355 and refer to priority code 014315 to register as a guest of Personal Finance.