Rethinking the 'Strong Jobs Recovery' Scenario

Job creation is crucial to any economic expansion. It directly affects consumer spending, and it's one of two key factors determining the health of real estate (the other being interest rates). One cannot overstate the importance of job creation to the economy.

Lately, the White House and Treasury Secretary John Snow have been trumpeting the fact that the economy has created 4.4 million new jobs since May 2003.

Inquiring minds want to know: How legit is that number? How was it derived? How does this job-creation data compare to prior cyclical recoveries?

Let's zoom in on the actual employment numbers and see what's there:

First question: How did the White House come up with that 4.4 million new-jobs number? Is it accurate?

The answer is simple math: Measured trough to peak, there were actually almost 4.5 million new jobs created. In May 2003, there were 129,827,000 people employed, according to the Bureau of Labor Statistics. As of November 2005, there were 134,289,000. That reflects 4,462,000 new jobs. So the "over 4.4 million jobs created" statement is numerically accurate.

  • So if that number is mathematically accurate, what's the problem?
  • As those of us who work on Wall Street know, you typically don't get to pick your time periods when measuring performance. You especially don't get to base it on trough-to-peak numbers. In most any series, there are more natural time periods, e.g., year to date, one, three and five years.

    As opposed to cherry-picking the most favorable-looking time periods, job creation historically has been measured from the end of the recession, which the National Bureau of Economic Research puts at March 2001. Another commonly used period is from the start of the president's term (Jan. 20, 2001).

    When we plug those time frames into the BLS data, we derive a significantly less rosy picture: From the beginning of the recession to last month, about 1.8 million jobs were created. Measured from the end of the recession, we see 3.4 million new jobs. None of these measures take into account the 2.6 million jobs lost from 2001 to 2003.

    Over the course of four years, those numbers fail to keep up with population growth. The U.S., with about 275 million people, needs more than 1 million new jobs per year -- between 125,000-150,000 per month -- just to maintain the same percentage of employed relative to the labor force.

    As with any data series, you can make the numbers better or worse depending upon when you mark the beginning of your time period, as the chart below of nonfarm payroll data since January 2000 shows.

    So the answer to our second question is that the 4.4 million number significantly overstates the true jobs picture since the end of the recession. Indeed, if this were a mutual fund, the Securities and Exchange Commission would not allow such an advantageously selective timeline to be used in the advertising.

  • Third question: We know the BLS model is a bit quirky and has some warts on it. How "real" are these numbers, and how much is theoretical conjecture?
  • That's a complex question, but let's take a stab at it. In 2001, the BLS started a new numerical projection called the birth/death adjustment. This replaced a prior adjustment known as the "bias factor." This new adjustment has been gradually phased in since 2001, and became fully implemented in 2003. That's convenient for our analysis, as it was fully integrated at about the same time that Treasury and the White House have used to reach their 4.4 million new-jobs number.

    The birth/death adjustment was created, according to the BLS, to capture job creation of new firms that is missed by bureau methodology "due to an unavoidable lag between an establishment opening for business, and its non-sampling methods must be used to estimate this growth."

    What the BLS does is estimate the number of new businesses coming into existence. It then projects how many new jobs these new firms create.

    Click here for larger image.

    There is some debate on how accurate the birth/death adjustment is. Morgan Stanley's analysts have found it to be (mostly) reliable; John Williams of Shadow Government Statistics thinks it significantly overstates job creation.

    I fall somewhere between the two. Given all the effort that goes into actually counting establishment jobs vs. merely estimating them, one would expect that this projection would be a relatively small number of the total new-job count. A modest estimate of new jobs created would be acceptable as a reasonable adjustment to the true data collection.

    But that's not what we see when we take a look at the data closely: Of the 4.4 million new jobs from the March 2003 low until present, the birth/death estimate accounts for 1,639,000. That is an extremely significant 36.7% of new jobs.

    By any measure, that's a hefty estimated adjustment to an actual data-based number. It is not particularly credible to me to have a statistical projection be more than a third of a measured data series. To be blunt, it is a game-changing "adjustment."

    To get an idea of how big this is, imagine how your performance might be enhanced by goosing it more than a third: This would raise the output from so-so to spectacular. A mediocre baseball hitter (.250) becomes an MVP (.392); an average bowler (200) rolls a perfect game; meanwhile, a weekend duffer who shoots an 88 becomes club champ with a 56.

    Nice statistical work, if you can get it.

    Our last measure is not quantitative, as the prior three issues have been. It is qualitative:

    As I've noted previously, the jobs recovery compares rather poorly with prior post-World War II recessions and their aftermaths. We've seen that the jobs created are unusually dependent upon the real estate complex. We also know that the private sector jobs created have, on average, paid less and have had weaker benefits than the jobs they've replaced. And we also see there has been an unusually large number of government jobs created. Overall, the jobs quality during this recovery is mediocre.

    Not Up to Snuff
    Job growth in the current recovery is weak by post-World War II standards
    Source: Barry Ritholtz

    After all this data-crunching, one query remains: How does this jobs-recovery era compare to prior ones?

    The answer, it turns out, is not particularly well. In fact, this is the eighth-worst jobs recovery of the prior 10 recessions, according to The New York Times.

    What made this cycle somewhat unique was that the job count fell for another year and a half after the recession ended.

    The reason for this is quite simple: Most economists have been looking at the post-recession period incorrectly. Instead of viewing this as a post-bubble economy, with the 2000 crash a rare event, they are looking at it as if it's just another postwar recession-recovery cycle. That misses the bigger issues.

    By nearly any honest measure, this has been a lackluster jobs recovery. That is not widely believed among the investing population -- though consumer sentiment shows plenty of hesitancy.

    Base of the Bear

    This is another in our occasional series of data analyses, looking beneath the headlines at the actual numbers to discern what's truly going on in the economy (see prior commentary on home sales data; Black Friday, and inflation).

    This variant perception -- that the macro environment is far worse than most people believe -- forms the basis of my bearish expectations for 2006, which I'll explore in more detail in a forthcoming column.

    Barry Ritholtz is chief market strategist for Maxim Group, where his research and market analysis are used by the firm's portfolio managers and clients in the U.S., Europe and Japan. He also publishes The Big Picture, his macro perspectives on the economy and geopolitics, entertainment and technology industries, and is a member of the board of directors of, a streaming media software company. At the time of publication, Ritholtz had no position in any securities mentioned in this column, although holdings can change at any time. Under no circumstances does the information in this column represent a recommendation to buy or sell stocks. Ritholtz appreciates your feedback; click here to send him an email.

    More from Markets

    Week Ahead: Trade Fears and Stress Tests Signal More Volatility To Come

    Week Ahead: Trade Fears and Stress Tests Signal More Volatility To Come

    Trump Takes Aim at Auto Imports; Markets End Mixed -- ICYMI

    Trump Takes Aim at Auto Imports; Markets End Mixed -- ICYMI

    Video: What Oprah's Content Partnership With Apple Means for the Rest of Tech

    Video: What Oprah's Content Partnership With Apple Means for the Rest of Tech

    REPLAY: Jim Cramer on the Markets, Oil, Starbucks, Tesla, Okta and Red Hat

    REPLAY: Jim Cramer on the Markets, Oil, Starbucks, Tesla, Okta and Red Hat

    Flashback Friday: The Market Movers

    Flashback Friday: The Market Movers