A bill pending in the House of Representatives would allow businesses with union workers to reduce their company pension obligations by billions of dollars, because statistics show that most blue-collar workers do not live as long as other Americans.

The provision, which has gone largely unnoticed in a broad pension bill, is being supported by the United Auto Workers and manufacturing companies whose pension funds now have assets far short of what they are projected to need under previous assumptions about worker longevity.

The measure would allow companies to assume that their blue-collar workers will on average die sooner than pension plans now assume they will. So companies, not having to plan to pay future blue-collar pensions as long as they now do, would not be required to put aside as much pension money as government regulations now require them to do.

But the leader of a panel that developed the actuarial data on which the new provision is based said he had written to the Treasury Department, which regulates pension funds, to express concern that the data were being misapplied.

Edwin C. Hustead, chairman of the actuarial panel, said in an interview he was concerned that the data were being used in an improper way. White-collar workers are shown by statistics to live longer, he said, but the bill would not require companies to factor that into their pension calculations. If it were included, unionized companies with largely white-collar workers would have to set aside more to fulfill their promises to retirees in the future.

In addition, Mr. Hustead said workers' pay had been shown to be a more powerful predictor of life expectancy than whether a worker was blue collar or white collar, but the bill did not recognize that higher-paid workers live longer and therefore require longer pension payouts. Many auto workers and airline pilots are classified as blue collar in the bill, because they are covered by collective bargaining agreements, even though they are highly paid.

Aides to the bill's sponsors said they were unaware that the measure had overlooked these other significant mortality factors. A spokeswoman for Benjamin L. Cardin, Democrat of Maryland, said the congressman's goal was to help preserve the system of traditional pensions. A spokesman for Rob Portman, Republican of Ohio, said the bill was intended to make sure companies "aren't forced to overpay" into their pension funds.

"The pension system is a voluntary system, and if you force companies to artificially put more into their plan than is needed, that is a real disincentive," he said.

The United Auto Workers wrote a letter in support of the provision, according to people with ties to Congress, apparently in hopes that the money that companies saved from pensions could be used for higher wages. The Erisa Industry Committee, which represents large companies, helped with the bill but was primarily interested in supporting another measure on the interest rates used in pension calculations, said Janice Gregory, a vice president of the committee.

With most pension plans now underfunded, Mr. Hustead and some other actuaries fear that a reduction in contributions could increase the risk of defaults, at a time when companies are already defaulting on their pension plans at a greatly accelerated rate.

"I do not agree that the tables should be adjusted for differences between mortality for blue-collar and white-collar employees," Mr. Hustead wrote to Treasury, which regulates how companies set aside money to cover pension promises.

After spending five years collecting and analyzing data for the new mortality table, Mr. Hustead warned that his panel's findings seemed at risk of being used in a "curious" and "arbitrary" way. He said in an interview that he had not received any response from the Treasury.

The issue of worker life spans can be traced to the early 1990's, when pension plans were also strained, although far less so than today. Companies at that time were free to select any mortality rate when calculating their current pension obligations, and some were discovered to be using unusually high death rates, so as to shrink their pension debt.

In 1994, Congress required all companies with pension plans to use the same mortality table, or set of probability factors for death rates.

At the time, the table in common use was from 1983. So the Society of Actuaries, a professional group that engages in research and education, convened a committee to prepare a new table, based on current demographic trends. Mr. Hustead, who is also a senior vice president of the Hay Group, was the chairman.

Normally, actuarial proceedings are quiet affairs that make few waves, but Mr. Hustead said this one was different. "There were some very strong opinions, for an actuarial group," he recalled. "We've done mortality studies in the past, and you go out and do your study, and come back and say, `Here it is,' and just start using it."

The group decided to create a base mortality table, then test a number of factors to see what effect they might have on the death rates. One test was to examine whether the color of one's collar had any effect. The group defined blue collar as people who were represented by unions or who were paid by the hour ・a definition that turns such unlikely workers as unionized athletes, airline pilots, aerospace engineers and even some newspaper reporters into blue-collar workers.

Separately, the committee tested to see whether pensioners' incomes affected their mortality rates.

After analyzing 11 million "life years," the committee could see that the color of the collar and the income level affected life span significantly. Blue-collar workers and the poorly paid both tended to die young. White-collar workers and the well-paid tended to live longer.

But even after trying multiple regressions and hiring an outside panel of academics, the committee was unable to find a statistically valid way of handling those workers who did not fit the patterns ・those who were "blue collar," but who also happened to be well paid.

In such cases, Mr. Hustead said, the data showed that money has such power to lengthen human life that it more than offsets the blue-collar life-shortening effect.

But not everyone on the committee wanted this point included in the final report, Mr. Hustead said.

"There were those who wanted to use the higher mortality rates for `blue collar,' but who did not want to use any lower mortality for `high paid,' which would be like the auto industry," he said.

In the end, the society published the new table in a report that alluded to its members' difficulties in the treatment of these adjustments.

Its work completed, the society then passed the table to its sister organization, the American Academy of Actuaries, which has the responsibility of representing actuarial interests in Washington.

The academy concluded that what it called the "collar adjustment" was statistically sound. But it rejected the society's adjustment for income level, saying current trends make that adjustment suspect.

Mr. Hustead said he was concerned that the same forces that had caused dissent within his own committee had traveled to the academy. He recalled that one actuary on his committee, who had forcefully opposed any adjustment of death rates by income, had also participated in the academy's deliberations.

"At one time in the middle of this process, I asked him, `Are you working for one of the auto companies?' " Mr. Hustead recalled. The actuary, Vincent Amoroso, said he was not.

In an interview, Mr. Amoroso said he had, in fact, represented auto companies at the time when Congress was requiring that all companies use the same mortality table. But he said he ended those consulting relationships before the new mortality table was released in 2000.

"I had not had any connection to the auto companies, in a consulting capacity, since 1996," said Mr. Amoroso, who is now with the consulting firm of Towers Perrin.

Mr. Amoroso added that he had volunteered to serve on the Society of Actuaries' committee because he wanted "to make sure that everybody behaves fair."

"It turned out I was just aghast at the raw, political partisanship of it all," he said.

He added that he felt that the committee had an established membership and ingrained habits of mind, and treated him and his ideas as less than welcome because he was "the new kid on the block."

"Who would have thought there would be such a melodrama in a mortality table's construction?"