Tuesday, May 23, 2006

At AK Steel, as at Enron, the Numbers Don’t Add Up

Following is the second part of an article on OSHA recordkeeping by guest bloggers "ERM." The first part appeared here on May 15.

AK Steel has a great safety record – at least on paper. Across the company the OSHA total recordable rate was 0.19 in 2005, the rate per 100 employees of the number of times per year an employee received treatment beyond basic first aid for an occupational injury or illness.

Things look even better at the company’s Middletown works: company spokesman Alan McCoy says the first quarter of 2006 was the best in history, according to an April 12 report in the Dayton Business Journal. The OSHA recordable rate of 0.13 was 42 times better than the steel industry average for all of 2005.

Maybe this safety record is just too good to be true, however. Two days later the Cincinnati Enquirer reported on an explosion that injured three workers at AK Steel’s plant. Worse, OSHA was having trouble investigating the blast because witnesses to the blast are working 12-hour shifts. Worse still, from a safety perspective, is that these are salaried employees and temporary replacement workers performing jobs that used to be done by the 2700 hourly workers who have been locked out since Feb. 28 when their contract expired.

Given these hazards, the explosion should have been a wake-up call.

Less than one month later, on May 4, Curtis Johnson died after being struck by a large coal cart, The Columbus Dispatch reported.

Twelve hour shifts plus temporary replacement workers equals one dead worker at AK Steel, but no effect on their injury/illness rates!

Of course, you might argue that the good rates are from the good times, before the lockout. Maybe. But first, remember how low those injury rates were and that for at least one month in the first quarter of this year the plant had brand new replacement workers.

Second, companies with superb safety records that are legitimate, have a deeply rooted safety culture. Organizations that value safety over production simply don’t ask replacement workers to work 12-hour shifts in a hazardous steel plant. This is just one example of how OSHA’s new math works, of how it is that every year injuries and illnesses go down, while fatalities and even fatality rates go up. Can both figures be correct? If not, which one do you believe?

If you are having trouble making up your mind, consider the AFL-CIO’s 15th annual “Death on the Job” report.

If you are still having trouble making up your mind, consider this recent study, as reported on April 12 by Occupational Hazards’ Josh Cable.
A NIOSH-funded study conducted by researchers from Michigan State University concludes that the current national system for calculating workplace injuries and illnesses "markedly underestimates the magnitude of these conditions."
The study, which is detailed in the April edition of the Journal of Occupational and Environmental Medicine, analyzes Michigan workplace injury and illness data in five different databases, including the Bureau of Labor Statistics, for the years 1999, 2000 and 2001. From that analysis, the researchers conclude that the number of work-related injuries and illnesses in Michigan is three times greater than BLS estimates.

The findings lead the researchers to believe that BLS significantly underestimates U.S. workplace injuries and illnesses. The researchers conclude that BLS needs to overhaul its current system "to address the undercount." (Emphasis added.)
Now, guess from where BLS gets its injury and illness data? OSHA! Remember, fatality figures can’t be gamed so easily and are collected separately – that is not only from OSHA.

If OSHA wants us to take its data seriously, it’s about time the agency started to take its own data seriously. That means enforcing the recordkeeping the standard.

OSHA will tell you it does audit injury and illness logs, but the problem is, declining injury and illness records are not just good for companies and their managers. Declining rates are good for OSHA too. It makes it look like the agency is doing a swell job. It just might be that OSHA is too close to the data to do a good job of auditing. Maybe OSHA needs to audited….After all, when was the last time you heard OSHA admit to a mistake?

Does this remind you of something? Enron cooked its books, of course. And those records were audited too, weren’t they? That’s right, but at least Arthur Anderson is now out of business.

Whereas every year, OSHA has the effrontery to congratulate itself on how injury and illnesses declined yet again. Last year, though, they were silent about one thing: fatality rates. That’s because for the first time in over a decade the fatality rate went up.