Is Employment Growing or Shrinking? It Depends on Who You Ask, Says Robert McTeer
New York Sun
Oct. 2, 2002

When I started my central banking career thousands of years ago, my dad, who ran a truck stop in Ranger, Georgia, never quite knew what a central banker did for a living. The way he kept track of my progress was to ask me after each promotion how many "hands" I had—meaning how many employees reported to me. The quaintness of that term today is a reflection of how far the economy has come from the physical, or goods, economy to the service and information economy of today. While we no longer call them hands, the number of people working is still one of the main ways we keep track of how the economy is doing.

Few economic reports receive the close scrutiny routinely applied to the monthly employment figures issued on the first Friday of every month. The two headline numbers are the monthly employment changes, which economists focus on, and the unemployment rate, of more interest to the press and the public. You have to read the fine print to learn that these two numbers are derived from separate surveys. That’s why they often seem contradictory and confusing.

Friday’s announcement was a case in point. The Labor Department stated that employment declined by 43,000 in September, but the unemployment rate also declined, from 5.7 percent to 5.6 percent. How can that be? Was this good news or bad news? The markets are still trying to decide. I don’t think the report was favorable.

The net loss of 43,000 jobs was calculated from a survey of employment at business establishments. The unemployment rate drop to 5.6 percent came from a survey of households, which showed employment increasing by 711,000, not declining by 43,000. That’s a huge difference.

The survey of business establishments polls businesses and inquirers as to the number of workers on their rolls. The household survey, on the other hand, asks household members if they are currently employed, and if not, have they been actively seeking employment. If they are employed, they are in the labor force. If they are not employed, and are looking for work, they are considered unemployed. If they are not employed and they haven’t been looking recently, they are considered not in the labor force and thus are not considered unemployed. In this scheme, discouraged lookers who have given up looking for a job help hold the unemployment rate down, and persons not employed who claim falsely that they are looking for work push it up.

Usually, over time, the two surveys converge and tell a consistent story. But not lately. It’s almost as if employed household members haven’t bothered to inform their employers. Or, maybe they’re "consulting." Or, perhaps, they are not working and keeping it a secret as in the movie, "The Full Monty."

The National Bureau of Economic Research has not yet ruled on the end of the recession and beginning of recovery, but my guess is that the switch happened around November or December of last year. So, what has happened to employment during the recovery period so far? Since November, according to the household survey of employment, the economy has created some 932,000 net new jobs through September. By this measure, employment growth was strong enough to absorb all the entrants into the labor force during that period and hold the unemployment rate down. September’s 5.6 percent unemployment rate is the same as last November’s rate. The survey of business establishments, on the other hand, indicates that the economy has lost 233,000 jobs over the same period. That’s a difference of more than a million jobs. That’s not even close enough for horseshoes.

If the survey of business establishments is more accurate than the household survey—and most economists believe that it is—the low unemployment rate is overstating the health of the economy. That would mean that our recovery so far has been a "jobless" recovery reminiscent of the jobless recovery of 1991-92. During that period, job growth was so sluggish that the unemployment rate continued to rise for more than a year after the recovery officially began. Unemployment peaked at 7.8 percent coming out of the 1990-91 recession. The peak was 10.8 percent in the 1982 recession. So far, unemployment has not exceeded 6 percent during this cycle because of the methodology of the household survey.

It could be that the more accepted low job growth numbers will be revised upward and clear up the confusion. Last Friday’s report revised the August employment growth numbers upward to 107,000 from the originally reported 39,000. But it seems more likely that the outsized job growth numbers of the household survey will be revised downward, with the unemployment rate revised upward.

I hate to look a gift horse in the mouth, but the recession didn’t feel that mild and the recovery doesn’t feel that strong.

About the Author McTeer is president and CEO at the Federal Reserve Bank of Dallas.