Skip navigation

Taking a closer look

Are declining workplace injury and illness rates too good to be true?

According to official statistics from the Department of Labor’s Bureau of Labor Statistics (BLS), the rate of workplace injuries and illnesses in private industry declined in 2005 for the third consecutive year. BLS reported that nonfatal injuries and illnesses declined to 4.6 cases per 100 workers in 2005, compared to 4.8 cases per 100 workers in 2004. The rate of injuries and illness requiring days away from work also declined 4 percent in 2005, according to BLS.

“The announcement that workplace injuries and illnesses in 2005 were at an all-time low is more good news for America’s workers and reflects the department’s effective worker health and safety strategy,” declared Secretary of Labor Elaine Chao. OSHA claims a 19 percent reduction in injury and illness rates since 2001.

Over the years, however, a growing number of voices have raised questions about whether these rosy announcements are “simply too good to be true.” The skeptics have pointed to exclusions in BLS-counted workplaces, continuing issuance of willful citations by OSHA against employers under-recording injuries, academic studies documenting massive undercounting by employers and a proliferation of corporate accounting scandals where “cooking the books” has occurred in many areas of business recordkeeping.

Accurate records of workplace injuries and illnesses are important because they are the essential starting point from which individual employers can identify high-hazard operations and develop appropriate controls. They also set government priorities for research, agency resources and enforcement actions.

Why Downplay?

There are powerful incentives for both government agencies and employers to downplay workplace injuries and illnesses. Declining injury rates can be highlighted as evidence of “success” of government efforts; at the same time, they can justify limited or no growth of government activity during a period of budget tightening.

Employers can benefit from lower reported injury and illness rates in reduced direct medical costs, lower workers’ compensation insurance premiums, use of a “clean safety record” in competitive bidding processes and avoidance of being targeted for high-hazard workplace inspections by state and federal OSHA. There are very real economic benefits to employers from underreporting.

According to its rules, BLS does not include in its annual survey workers who are self-employed; farms with fewer than 11 employees; federal, state and local governments (including fire and police departments); or private household workers. Given that nearly 15 percent of employed people are government workers and more than 7 precent of the work force is self-employed, these are significant exemptions.

Moreover, small firms, especially in construction with immigrant workers, may underreport injuries or fail to report them at all. In California, for example, 78 percent of the state’s employers have fewer than 10 employees. Small farms have the same high-hazard profile as agriculture in general; and police and firefighting work is known to generate significant rates of injury and illness.

Tip of the Iceberg

Over the last several years, even President George Bush’s employer-friendly OSHA has found it necessary to issue citations, often willful cites, for deliberate underrecording of injuries and illnesses. The companies cited are not simply isolated bad actors, but rather are major Fortune 500 firms that more likely represent just the tip of the iceberg.

The OSHA recordkeeping citations include:

June 2004: Federal OSHA issued a willful citation (later changed to “unclassified”) and $70,000 fine to Weyerhaeuser’s Trus Joint facility in Buckhannon, W.Va., for failing to record at least 38 injuries and illnesses on its OSHA 300 Log. The citations “paint a picture of an organization where underreporting of injuries and illnesses appeared to be a routine practice that was tolerated, and even rewarded, by company vice presidents.” (OCCUPATIONAL HAZARDS, September 2004)

October 2004: Southern California Edison underreported workplace injuries and illnesses for the previous 7 years and had to return $35 million in safety-related bonuses to the California Public Utilities Commission. “Edison found evidence that supervisors contacted outside medical personnel to influence treatment, change medical records or downgrade the seriousness of an injury. Other times, Edison said, its managers encouraged employees to dodge safety reporting requirements by undergoing physical therapy or using vacation days during recovery.” (Los Angeles Times, 10/22/04)

April 2005: Federal OSHA conducted 46 inspections of high-hazard industry facilities with very low workplace injury and illness rates, and found five plants (10.8 percent of those inspected) where recordable injuries and illness were not entered onto the facilities’ 300 logs. Citations and fines were issued in these cases.

November 2005: Federal OSHA issued three willful citations and $165,000 in fines to Fraser Paper’s Madawaska, Maine, paper mills for recordkeeping violations between 2003 and 2005. Federal OSHA found 59 instances of injuries and illnesses that were not recorded, 77 instances where recordable entries were not made within 7 days, and 2 years (2003 and 2005) for which incomplete annual injury and illness log summaries were certified as being complete.

June 2006: Cal/OSHA issued willful citations to the giant KFM construction consortium rebuilding the eastern span of the San Francisco Bay Bridge for failing to include 13 lost-time injuries in its OSHA 300 Log, including several resulting in major surgeries. Cal/OSHA found that KFM used an elaborate system of cash incentives, reprisals and careful management of medical treatment to reduce reported and recorded injuries.

November 2006: Federal OSHA issued 33 citations against the Volks Constructors company in Baton Rouge, La., with penalties of $47,600. Four of these citations related to failure to complete the OSHA Form 301 used for a first report of injury; failure to record 102 injuries on the company’s Log 300; company certification of Log 300s that were neither correct nor complete; and failure to provide the Log 300 and Form 301 upon request.

December 2006: The Pittsburgh Post-Gazette reports that deaths tabulated by MSHA excluded “natural causes,” even if heart attacks and hospitalizations were the direct result of accidents.

The newspaper’s review of MSHA records indicated 153 deaths have occurred at U.S. coal mines in the last 4 years, but another 72 deaths were labeled “nonchargeable” incidents, 55 of these attributed to heart attacks or other “natural causes.” In the same period in noncoal mines, 154 deaths were listed as “chargeable” and 100 classified as “nonchargeable” because the workers died of “natural causes.” (Pittsburgh Post-Gazette, 12/3/06)

There have been several recent academic studies done in the United States and Great Britain documenting very significant undercounting of injuries and illnesses, due to both employer underreporting and government data collection exclusions.

In the January 2004 issue of the Journal of Occupational and Environmental Medicine, University of California at Davis researchers reported that BLS may miss as much as 69 percent of all workplace injuries and illness. Instead of 5.7 million nonfatal injuries in the United States each year, there are more likely 11 million to 12 million cases, according to the report.

In the same journal’s April 2006 issue, researchers at Michigan State University analyzed data collected by BLS and found that the statistics failed to count roughly two-thirds of nonfatal workplace injuries and illnesses in Michigan over a 3-year period. The researchers estimated that 869,034 work-related injuries and illnesses occurred on average each year in Michigan from 1999 to 2001, compared with the BLS estimate of 281,567 per year. The report estimates that 75 percent of the injuries and illnesses missed by BLS resulted from employer underreporting.

The problem of underreporting is not exclusive to the United States. In February, a University of Liverpool study prepared for the United Kingdom’s Health and Safety Executive concluded that 70 percent of workplace accidents in the United Kingdom – including major injuries – are not being reported. The study found that out of 581 injured workers interviewed at the Royal Liverpool University Hospital, only 30 percent of the injuries were reported under Britain’s RIDDOR regulations (the equivalent of the OSHA 300 Log rules).

While it is very difficult to “prove” a negative, that is, provide exact instances of what was not accurately reported, recent high-profile investigations of corporate safety culture indicate that “creative accounting” has been done in the safety realm as well as in the financial records scandals of Enron, WorldCom and Arthur Andersen. If so many major corporations have been willing to mislead investors and the Securities Exchange Commission about sales, profits and executives’ pay, it is not surprising that “cooked injury and illness books” resulting in significant economic gains to employers also have occurred.

Clearly it is time for a serious look at what workplace injury and illness rates include and how they are calculated. Without an accurate picture of injuries and illnesses on the job, setting effective priorities for employer and government resources is not possible.

In 1987, the Keystone Center in Colorado brought together 46 people from corporations, labor unions, various health professions, government agencies and academia to discuss how to fix underreporting and other problems in government job injury statistics. Their recommendations died on the vine and were never enacted. Twenty years later, it’s time to try again.

Garrett Brown, MPH, CIH, is a compliance officer for the California Division of Occupational Safety and Health (Cal/OSHA) and the volunteer coordinator of the Maquiladora Health and Safety Support Network.

TAGS: Archive
Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.