Facebook Twitter Tumblr Close Skip to main content
A Project of The Annenberg Public Policy Center

Economy Producing Mostly Bad Jobs? Not so fast.

A new set of figures from the Bureau of Labor Statistics show HIGHER-paying jobs growing faster. A FactCheck.org exclusive. (But there's evidence on both sides).


Now that the economy is growing and creating new jobs, John Kerry has been saying that the quality of those jobs is “much lower” than the quality of jobs that have been lost. A recent ad by some Kerry allies even shows a middle-aged man reporting for his new job wearing a paper hat at a seedy-looking burger joint.

Well, hold on — there’s strong new evidence to the contrary.

A new set of numbers from the Bureau of Labor Statistics actually shows solid growth in employment in relatively higher-paying occupations including construction workers, health-care professionals, business managers, and teachers, and virtually no growth at all in relatively lower-paying occupations including office clerks and assembly-line workers. It’s the most detailed breakdown yet — looking at 154 different job and industry groupings. These statistics are a FactCheck.org exclusive — supplied to us by BLS at our request and not previously published.

Another statistic often overlooked by Bush critics is that average earnings of rank-and-file private-sector workers have increased since Bush took office, though modestly. Even after adjusting for inflation — including the rising price of gasoline –those earnings are up just over 1% since January 2001, despite the recession and the initially slow recovery.

These statistics come from a different BLS survey and cover a somewhat different time period than the figures cited by Kerry. They are going to be controversial and won’t settle the good jobs/bad jobs argument. There’s also plenty of evidence that large numbers of Americans are indeed worse off now than they were before 2001, including the fact that more than 1 million Americans have been out of work for a full year or more.

We can’t disprove Kerry’s claim that bad jobs are replacing good jobs over the past few months. But we do see good evidence that job quality has increased over the past year or more.


Let’s start with what’s new. Numbers supplied to us by the BLS highlight an important trend: more white-collar managers and professionals, relatively fewer low-wage jobs requiring less education. As we sort out the raw data, employment has recently increased by more than 1 million in categories that on average paid above the median earnings of $541 per week, while employment was virtually unchanged in categories paying below the median. That’s comparing the most recent 12 months with the same period a year earlier.  By that measure, the jobs gained are overwhelmingly good jobs — the very opposite of the claim made by Kerry and his allies.

MoveOn.Org Voter Fund
Ad: “Burgers”

Announcer: You put in 30 years at the company. You got good pay, health care. They send your job overseas. And under George Bush, the company gets a tax break for doing it. Now, Bush says we’re in a recovery.

And after a year, you finally land another job. And you wonder, is this what you worked your whole life for?

We’re not being led. We’re being misled.

Bogus Poster Boy

An extreme example of the Democratic line can be seen in the TV ad released June 22 by MoveOn.org Voter Fund, showing a dejected man who’s lost his job after “30 years at the company” reporting for his new job as a burger-flipper. It’s a caricature, of course. Somewhere in a nation of nearly 294 million people there must be some workers in that unhappy position. But the guy in the paper hat is by no means the poster boy for this economy.

(The ad blames Bush for sending the man’s job overseas. We’ll address the “offshoring” argument in a later article.)

Kerry has been making a similar but less excessive argument. After a West Coast appearance May 18 he was quoted by The Associated Press saying “The jobs that are being created in Oregon and elsewhere are paying significantly less than the jobs we’re losing.” A Kerry news release issued June 19 says “90 percent of the new jobs created since August of 2003 are in industries that pay an average hourly wage that is less than the national average.”

It’s true that looking at job changes by industry shows lower-paying industries have been growing faster than better-paying industries in recent months. A June 30 story in USA Today — headlined “Low-wage jobs rise at a faster pace” — also cites industry-by-industry statistics to document that idea. The newspaper even illustrated the story with a photograph of a food server, supposedly typical of the low-wage jobs said to predominate.

Earnings Have Risen

But we respectfully disagree: neither the statistics cited by Kerry nor the numbers cited by USA Today really prove the case, for the simple reason that broad industry averages tell nothing about the pay levels of the specific jobs that have been gained or lost within those industries. The USA Today figures tell us there’s been a gain in the relatively low-paid restaurant industry — but can’t tell us how many of those new jobs are dishwashers and burger flippers and how many are managers, chefs or wine stewards. Nor can it tell how many jobs lost in the relatively high-paid manufacturing industry are engineers, foremen, managers or other professionals, and how many are lower-paid assembly-line workers or janitors.

Besides, if the economy is adding mostly bad jobs, why are average earnings rising?

Take a look: Since Bush took office in January 2001, average weekly earnings of private-sector production workers have gone up nearly 8 percent — and that’s despite the eight-month economic recession that began in March 2001 and the long job slump that continued until August 2003.

Those “production workers” are roughly the lowest-paid 80% of those on private payrolls, including working-level supervisors. They don’t include managers or government workers, who tend to make more.

Even after discounting those rising earnings to account for inflation, the average production worker’s buying power — before any tax cuts are figured in — still gained just over 1%.

To be sure, as Bush critics point out, in the past few months inflation has slightly outstripped the rise in earnings. Since August, 2003 (the month the economy stopped losing jobs) weekly earnings have gained 1.6%. But after adjusting for inflation earnings are down — by less than two-tenths of one percent. Critics cite this dip in inflation-adjusted earnings as evidence that mostly poor-quality jobs have been added since the upturn began.

We disagree here as well. The very recent decline in inflation-adjusted earnings has more to do with rising oil prices than with any change in job quality. The Consumer Price Index has risen 2.3% since August, but with energy prices taken out it has risen only 1.6%. So, if energy prices hadn’t spiked, inflation-adjusted average weekly earnings wouldn’t have declined. In any event, the critics are mostly silent about the overall rise in inflation-adjusted earnings since Bush took office.

The New Figures: More Good Jobs

The figures we’ve obtained from BLS paint a brighter picture of what’s happening to job quality.

They focus on the types of jobs within broad industry categories. They also give a more detailed picture than earlier analyses. For example, the USA Today analysis was based on a breakdown of 78 different industries, while the new BLS figures give information for 154 different occupation groupings within industries.

What these new numbers show is strong growth in higher-paying employment categories over the past year — more than 1.1 million gained — and stagnation in lower-paying job categories (which recorded a loss that is within the statistical margin of error for this survey.) The pattern shows even more dramatically going back three years, over which lower-paying categories lost more than half a million while higher-paying categories gained 1.6 million.

(For the full set of figures for each of the 154 categories go here.)

Some of the big gainers were construction jobs such as carpenters and laborers, a category paying an average of $553 per week in 2003 — putting them just barely in the higher-paying category. Employment in this group increased by more than half a million jobs over the 12-month period we studied. The second-largest gain was in health and education professionals, a group including doctors, nurses, teachers and others, paying an average of $691 per week in 2003, and gaining 369,000 jobs.

The biggest loser among lower-paid groups was the category of production occupations in manufacturing, including assembly-line workers (but not higher-paid managers and professionals, as noted earlier.) These manufacturing production workers averaged $509 a week in 2003, well below the median of $541, and lost 179,000 jobs. Office and administrative workers in the information industry lost 90,000 jobs, and paid even less: $497 per week on average.

It’s not possible to reconcile these new figures with the very different figures cited by Kerry and others. Theirs come from the monthly BLS survey of payroll data from 400,000 US business establishments. Our new set of numbers comes from the BLS monthly survey of 60,000 US households. (For a technical discussion of the two surveys and how they differ, go here.)

The payroll survey is generally considered the more accurate of the two, and it’s the one the government and economists use to calculate changes in overall job levels. But it doesn’t provide demographic information on such things as race, gender, who’s unemployed — or occupation. Only the household survey distinguishes the chef from the dishwasher, the plant manager from the janitor.

Limitations in Data

There are limitations in our data, too. We can’t say what’s happened from month to month because the BLS (for technical reasons) has not yet adjusted these numbers to remove normal seasonal variations. So we can’t use these numbers to measure what’s happened since last August or since the precise month that Bush took office. We’ve only compared periods that are 12 months apart — which avoids any problem with seasonal changes. But even so, they can tell us something about what’s happened in the last year, or the last three years.

Note also that we’ve used 12-month “moving averages,” averaging together all the figures from one 12-month period and comparing them to the average for one year and three years earlier. That’s because the figures carve that 60,000 monthly sample into some very small categories that bounce around from month to month just due to random variations in who’s being surveyed.

Finally, both these numbers and the payroll survey numbers are taken from samples and thus subject to sampling error. BLS calculates that the actual median earnings for 2003 could be as high as $543 or as low as $539, for example.

Economists Disagree

Which of these conflicting statistical views of job quality is closer to the truth? Economists don’t agree on this.

David Wyss, chief economist for Standard & Poor’s, says it’s just not true that bad jobs are growing faster than good jobs. “They’re not,” Wyss tells FactCheck.org. “Good jobs are growing faster than bad, although there is some evidence that both bad and good jobs are growing faster than middle jobs.”

And Randy Ilg, a staff economist at the Bureau of Labor Statistics who has studied the question of job quality for nearly a decade, says the same thing. “Over the last couple of years, if you look at it from a perspective of occupations within industries, clearly there has been more growth in occupation categories that pay above the median.

But Mark Zandi, chief economist of Economy.com, who supplied USA Today’s industry-by-industry figures, focuses on the past few months and says, “We’re creating a lot more jobs but they are still largely lower-paying jobs . . . It means the jobs we are creating pack less of a punch for the economy.”

The best we can say is that no survey we know of even attempts to compare the specific jobs lost with the specific jobs added throughout the whole economy. Statistics solid enough to settle the question definitively just don’t exist. But — given the increase in average weekly earnings we mentioned — we’ll venture a guess that the trend in job quality has been up over the last year, or even three years, at least for the rank-and-file workers who make up 80% of the private-sector workforce.

Footnote: Kerry would be on more solid ground if he focused on the problem of long-term unemployment. BLS figures show that in June — and for 14 of the past 15 months — more than 1 million Americans have been out of work for a full year, or longer. The country hasn’t seen such levels of long-term joblessness since early 1994.



Watch MoveOn.Org Ad: “Burgers”


Supporting Documents

View Spreadsheet: “Job Growth by Occupation/Industry and Average Weekly Earnings.”


William McCall, “Kerry promises less outsourcing, more jobs; urges energy independence,” The Associated Press, 18 May 2004.

“Reality Check On The President’s Radio Address ,” news release, John Kerry for President, 19 June 2004.

Barbara Hagenbaugh and Barbara Hansen, ” Low-wage jobs rise at faster pace ” USA Today, 30 June 2004: B4.