Not out of the Woods Yet with Employment

Year-end employment data from the Bureau of Labor Statistics (BLS) make interesting reading. If you are an optimist, you can rejoice in the gains posted during 2004 in both payrolls and the total number of people employed. If you are more of a pessimist, you can fret over the number of people on payrolls at the end of 2004 still being slightly below what it had been at the end of 2000.

Four years of demographic-driven growth in the potential workforce--approaching 2 million per year in good economic times--coupled with no net growth in payrolls is not a healthy situation.

Another way to look at this disconnect is that while the population of 16- to 65-year-olds was increasing by 9.0 million and of 25- to 65-year-olds by 7.6 million between December 2000 and December 2004, payrolls failed to grow, and growth in total employment averaged a meager 600,000 per year.

Preliminary BLS data for January are not encouraging. Although they indicate that payrolls, at 132.6 million, finally crept past the 132.5 million record high set in February 2001, the gains for the month in total employment and payrolls were both well below the levels needed to keep up with population growth.

The good news from the year-end BLS data is that total employment increased by 1.7 million during 2004, bringing it to about 2.5 million higher than it had been at the end of 2000. And although the growth rate for payrolls slowed later in the year, they posted a respectable 2.2 million gain for the year.

The big question is whether 2005 will be a second year of payroll gains and so confirm and consolidate, if somewhat belatedly, continuation of the post-World War II pattern of bust and boom employment cycles of relatively short periods of payroll decline followed by much longer periods of payroll gains.

The payroll decline in the early 1980s was very sharp--2.8 million jobs--and the downtrend lasted for 18 months. The subsequent upturn brought payrolls back to the previous peak within another 11 months and ended up lasting for seven-and-a-half years and adding 21 million jobs.

The downturn of the early 1990s lasted only 12 months, with the loss of 1.6 million jobs. It took another 21 months to retrieve this loss. The recovery went on to last almost 10 years and boost payrolls by 24 million.

The payroll decline of the early 2000s, which started in April 2001, was not exceptionally large, with a drop in payrolls of 2.7 million. But it lasted for an unprecedented 27 months, and it has taken another 20 months of recovery for a total of 47 months.

There have been a total of nine periods of employment decline since World War II. For the first six, the declines lasted an average of 12 months with another 10 months to regain the previous peak for an average total of 22 months. For the early '80s and early '90s, the decline-and-recovery periods lasted a longer 29 and 33 months, respectively.

Is the further extension of the decline-and-recovery period to 47 months for the latest downturn just in the natural order of things in an ever more complex economy? Or is it possibly reflecting systemic changes in the job market and in employer-employee relationships in the U.S.?

Part of the problem in answering that question is the plethora of data on employment--data that do not all point in the same direction.

Payroll data, which are generally regarded as the most meaningful employment measure, are developed monthly by BLS from information from a sampling of more than 400,000 employers. They are checked and adjusted periodically against hard data on unemployment insurance.

BLS also collects less comprehensive monthly data on total employment from a smaller sample of about 40,000 households.

Historically, payroll and household data have moved more or less in parallel, with the household employment level being somewhat higher because it includes workers not in the payroll count. These include farm workers, consultants, and those in day and more casual labor.

The continued growth in total employment over the past four years, even if modest, while payrolls have stagnated raises the possibility that payrolls--the most stable employment relationship--are starting to lose out to less binding, less secure, and likely less well paying alternatives.

Be that as it may, there are still a number of job-related issues that chemists, most of whom still work for organizations that make things, should keep an eye on:

◾ What BLS defines as total private employment is dragging. In January 2005, it was still more than 800,000 lower than its peak, set in December 2000.This follows a 21 million gain over the previous eight years.
◾ Most of the 2.5 million manufacturing jobs lost over the past four years will not be retrieved anytime soon. Such jobs were down a further 145,000 in January.
◾ In a steady and continuing decline, jobs in the nonpharmaceutical component of the chemical industry have dipped by 23%, or more than 180,000, over the past decade.
◾ The impact on domestic employment of globalization and job outsourcing overseas has yet to be definitely quantified. But there is little doubt that it is real. And the supply of well-qualified scientists in China, India, and other nations is substantial and growing very fast.

Views expressed on this page are those of the author and not necessarily those of ACS.