Eighty-one-year-old Yale Goldman helped guide his family's machine factory through the Great Depression just fine. It's the recently heralded rebound of manufacturing that's killing it.
Though his machines for coiling wires have no direct competitors and the $14,000 price tag is less, after accounting for inflation, than it was a decade ago, sales have plummetted. The staff of the Baltimore-based Able Winding Machine Co., which peaked at 100 in the 1930s, has shrunk to nine.
So, when he hears government pronouncements that U.S. manufacturers are healthier than ever, Mr. Goldman says he is torn by conflicting emotions: " 'Thank God we are on the way,' and 'They are spitting in the wind.' "
Mr. Goldman's confusion about the health of manufacturing is being echoed in government offices and university economics departments.
In recent months, several U.S. government reports and studies have indicated that after the terrible retrenchment of the early 1980s, the nation's basic industries have bounced back and become more efficient than ever.
Department of Commerce statistics show, for example, that while the manufacturing work force fell from a high of 21 million in 1979 to 19.1 million last year, manufacturing's share of the nation's economic output rebounded from a recession-era-low of percent in 1982 to a healthy 23.3 percent last year. That's the same share of gross national product manufacturers held in the go-go era of the 1960s.
Pumping out more steel, clothes and cars with fewer workers means that manufacturing workers have become more productive.
In fact, the government statistics show that manufacturing workers' productivity has soared past that of other sectors, increasing at a rate of 3.5 percent a year in the 1980s. During the same period, the productivity of workers in non-manufacturing businesses improved at only 0.2 percent a year.
But some business people and economists argue that those statistics don't reflect what is really happening to the manufacturing sector. The strength of the nation's basic industries, which employ one in six working Americans, may be exaggerated by the way the government calculates output and productivity, they say.
Even the statistics' harshest critics say the bias isn't intentional. The problem: No one, including overworked and underfunded government economists, has figured out the perfect method to compare the output and productivity of manufacturers to service companies. It is, they say, like comparing workers at the Apple computer factory with waitresses under Howard Johnson's orange roofs.
"Most people agree there is a bias to the numbers. They show too large an increase for manufacturing," says Lawrence Mishel, research director of the Economic Policy Institute.
Mr. Mishel, a longtime critic of government manufacturing statistics, warns that inaccurate government statistics are dangerous for workers and businesses alike.
If the statistics paint manufacturing as healthier than it really is, there may be "an indifference to the plight of manufacturers . . . a lack of governmental programs" to aid them.
Though many economists both in and out of the government dispute Mr. Mishel's claims about the significance of the bias, there is little argument over the cause of what errors the statistics do contain:
Computers. Or rather, everything but computers.
The government gets many of its productivity and output numbers by sending hundreds of researchers out to stores to collect price and quality information on everything from shoes to modems.
Whenever the researchers notice a decline in price or an improvement in quality for a sample product, they tabulate the information as an improvement in productivity, said Jack Triplett, chief economist of the Department of Commerce's Bureau of Economic Analysis.
The federal government has developed a comprehensive system track the dramatic productivity improvements by domestic computer-makers, he said.
But it isn't so easy for researchers to figure out whether changes in shoe quality, for example, are improvements or not.
And the researchers have an even more difficult time figuring out whether the ongoing changes in some services, such as banking or medical care, are beneficial or detrimental.
"Sometimes quality goes up and it isn't detected. We don't know how to make the right adjustment," Mr. Triplett said. As a result, computers are given too much weight in productivity, some economists believe.
Some of this bias may be corrected later this year, when the Department of Commerce shifts its standard for "real," or inflation-adjusted, dollars from 1982 to 1987, some economists believe.
By moving the base year closer to the present, the statistics will more accurately describe the importance of computers, explained Robert Gordon, an economist at Northwestern University.
Though computing power today costs one-twentieth of its price in 1982, for example, computers are not 20 times more important, he said.