Dubious data clouds job safety gains
Executives at Smurfit-Stone Container Corp. were jubilant. The big packaging company, a self-proclaimed leader in workplace safety, had smashed its own record for lowest injury rates in its industry.
It was another milestone in Smurfit’s “incredible tradition of safety achievement,” said a February 2008 press release.
Yet the month before, Monterey County authorities filed criminal charges against two officials of a local Smurfit plant and a medical provider, accusing them of conspiring over several years to cover up injuries and discourage workers from filing workers compensation claims. The men pleaded not guilty, and the case is pending.
Official statistics show remarkable declines in injury rates for workers, not only at Smurfit, but at U.S. companies generally. According to government estimates, since 1992 injuries and illnesses in the private sector have tumbled by more than half, from 8.9 cases per 100 full-time workers to 3.9 cases in 2008. But the estimates rely exclusively on injury logs kept by employers. Officials cite mounting evidence of gross undercounting, even for injuries as serious as amputations. Rep. Lynn Woolsey, D-Calif., who chairs the House subcommittee on workforce protection, says true injury totals could be “two or three times greater” than official counts.
Employers have financial motives for fudging injury records, observers say, including a desire to cut workers compensation costs and avoid safety inspections. Workers sometimes fail to report injuries out of fear of getting fired. And safety incentive programs at some companies encourage injured employees to keep quiet to win prizes.
Moreover, official injury estimates by design exclude large numbers of workers and job sites — including the self-employed and farms with fewer than 11 employees.
Experts said bogus injury counts reduce the chance of hazards being identified and hinder safety regulators in setting priorities for inspections. Unreliable data can drive other decisions too. In July 2006, Gov. Arnold Schwarzenegger vetoed funds to hire 15 more inspectors for the state’s job safety agency, Cal/OSHA, on grounds that injury rates in the state were below the national average.
Official injury rates are based on Bureau of Labor Statistics surveys of employer logs. Academic research has exposed wide gaps between these estimates and other data, such as hospital and workers compensation records.
Studies published in 2006 and 2008 found the official estimates missed 68 percent of injuries and illnesses in Michigan, and at least 24 percent of those in six other states.
Official estimates also missed 77 percent of work-related amputations in Michigan in 2007, according to researchers at Michigan State University. In contrast to the official estimate of 160 amputations, the study identified 708, mostly involving the loss of fingers. “Amputations should be a pretty obvious type of event,” Dr. Kenneth Rosenman, an epidemiologist and one of the authors, said in an interview.
It would not be the first time government safety data were significantly in error. In 1987, a study by the National Academy of Sciences found that more than half of work-related deaths were being missed by the official estimates. The Bureau of Labor Statistics switched from estimating fatalities to counting them, and in the first year the number spiked — from 2,800 deaths in 1991 to 6,217 in 1992.
Reflecting growing concern, the U.S. Occupational Safety and Health Administration last October announced a “National Emphasis Program” to crack down on inaccurate reporting through stepped up audits of injury logs. Cal/OSHA will be “fully participating” in the initiative, a spokeswoman said.
Jordan Barab, deputy assistant secretary of labor for OSHA, acknowledging that injury data suggest the agency is doing a great job, said it is “tempting for any agency to put out stats that make it look good.” But he said in an interview that federal OSHA has no interest in “looking good based on statistics we really don’t have any faith in.”
Others said the concern is overblown. “There are probably some employers who knowingly may be underreporting,” said Baruch Fellner, a Washington, D.C., lawyer for the U.S. Chamber of Commerce.
But “the majority … are doing a good job,” he said, adding that undercounting may stem from difficulty in following complicated rules, not deliberate cheating.
In recent years, a number of companies have been cited for cheating on injury records. In July 2004, a Weyerhaeuser subsidiary in West Virginia agreed to pay $77,000 in penalties for failing to log at least 38 injuries and illnesses. In October 2004, OSHA sought $160,000 in fines from General Motors Powertrain Corp. of Massena, N.Y., alleging failure to log 98 injuries. After an appeal, a hearing judge cut the penalties to $39,000. But in January 2007, the company was cited for repeating the violations and fined $28,700.
OSHA sought $170,000 in penalties from Fraser Paper of Maine, including for 59 instances of failing to log injuries and inaccurately certifying injury totals as complete. The company appealed and paid a $107,000 settlement in May 2006.
In 2008, utility regulators ordered Southern California Edison Co. to surrender $35 million in safety bonuses for manipulating records, including by characterizing job-related injuries as occurring at home.
The same year, a medical group complained about members coming under pressure to help employers maintain spotless injury logs. In a letter to OSHA, Dr. Robert K. McLellan, then president of the American College of Occupational and Environmental Medicine, described “the dilemma faced by occupational physicians when they are pressured by employers to minimize treatment or fudge” records.
The Government Accountability Office followed up with a survey of more than 500 occupational doctors, nurses and physician assistants. The GAO reported last October that more than half said they had been pressured by workers or employers to downplay injuries; 44 percent said this affected how injuries were recorded.
Under OSHA rules, conditions requiring medical care must be recorded, but not those involving simple first aid — a distinction that can lead to inappropriate treatment to avoid an injury report.
Incentive programs designed to encourage safe work practices can also exert pressure on supervisors and workers to hide injuries. In many cases, the only way to win prizes is to keep a clean injury log. And when awards are tied to the performance of entire crews, peer pressure kicks in. “If I’m injured, not only do I lose out, but my fellow workers lose out,” said Les Boden, a professor of environmental health at Boston University.
Last summer, the AFL-CIO surveyed 868 leaders of union locals from various industries, more than half of whom had incentive programs at work. Of these, 13 percent said the programs encouraged injury reporting, while 58 percent discouraged it.
A string of unreported injuries on a major bridge project offers a glimpse of what happens when incentive programs go too far.
In 2006, Cal/OSHA accused a consortium of contractors, known as KFM, of failing to log 13 injuries of workers rebuilding part of the San Francisco-Oakland Bay Bridge, and of recording late or failing to investigate four others. A separate review by the California Bureau of State Audits found that workers were offered incentive payments of $200 to $600 apiece — and foremen twice the amount — if certain phases were completed without a recorded injury.
“During my experience at KFM, I witnessed a pattern of deliberate underreporting of injuries,” Winston Peart, a former field safety manager, said in a written statement to state audit officials. Arne Paulson, a pile driver foreman, told Cal/OSHA that after suffering a knee injury, crew members sometimes had to carry him onto boats heading to construction barges to avoid logging a lost-time injury. “It was known by everyone not to report any injuries because that would mean no BBQ, no tool prizes, no tool box prizes,” Paulson told the agency. “Everyone would want to know who ‘lost’ the prizes for the crew.”
Cal/OSHA issued two citations and sought $5,790 in penalties. Without admitting guilt, KFM agreed last April to pay the fines after Cal/OSHA downgraded the most serious citation from “willful” to “regulatory.”
Dole Fresh Vegetables, a Dole Food Co. unit in Soledad, Calif., also faces Cal/OSHA charges of hiding at least 11 employee injuries in 2006 and 2007. The company’s appeal is scheduled for a hearing in April. A Dole spokesman declined comment.
A leaked letter from a Dole lawyer to a company official acknowledged that some supervisors may have failed to “accurately report claimed industrial accidents.”
The 2006 letter followed an internal review by attorney Peter R. Nelson of tactics allegedly used to keep ailments of lettuce workers off the books. Some employees reported being directed to seek treatment for injuries “in Mexico on a personal basis with physicians and/or unlicensed ‘sobadors’ [healers],” the letter said. Others were asked to get treatment through group health plans, rather than workers compensation, or to let supervisors “personally pay for this treatment.” The letter was first reported in January by KCET television’s “SoCal Connected.”
Marty Ordman, vice president and spokesman for Dole, said that following the review Dole fired one supervisor and suspended a foreman for violating company policies.
In the Smurfit case in Salinas, former plant officials Douglas Tateoka and David L. Polk, along with physician assistant Eugene G. Guzman, Jr., are charged with 109 insurance code violations involving injury claims of more than two dozen workers. The Monterey County District Attorney claims that workers were pressured not to file workers compensation claims, and that injured employees were sometimes ordered back to work without adequate recovery time.
Among them was Francisco Pulido, a maintenance mechanic who lost the tip of his little finger when it caught in a roller in 2007. Pulido allegedly was sent back to work the same day as the accident, and Smurfit later suspended him without pay for three days, claiming he broke safety rules by incorrectly using equipment.
Smurfit, based in Chicago and Creve Coeur, Mo., was not charged and a spokesman declined comment. The defendants have pleaded not guilty, and their lawyers said their actions were completely aboveboard.
Authors: Joanna Lin, Bridget Huber, Matthew Richmond, Jill Replogle and Myron Levin contributed to this report.