OSHA Injury & Illness Stats: Good News or No News?

I have come under a bit of criticism (shocking isn't it) on a safety listerve for not covering the Bureau of Labor Statistics' recent announcement that

nonfatal workplace injuries and illnesses in private industry declined in 2003, both in terms of the number of cases per 100 full-time workers, as well as the total number of injury and illness cases reported.

A contributor to the "Safety" listserve writes:

I find it interesting that in his little blog Jordan Barab does not mention that: "Approximately 4.4 million injuries and illnesses were reported in private industry workplaces during 2003, according to the latest report by the Labor Department's Bureau of Labor Statistics. The number translates to a rate of 5.0 cases per 100 full-time workers and a 7.1 percent decrease in the actual number of injuries and illnesses reported in 2002."

A 7.1 decrease in injuries and illnesses. Why all the focus on the negative? Things are improving out there. Perhaps he should highlight the positive as well.

And a private e-mail asks:

Could it be that, by concentrating on the news that you think could be interpreted as
negative to the current administration, you are using safety to promote your political agenda. Nah, people on the left would never do such a thing. Or would they?

For shame.

Yeah, I'm hanging my head. Wait... just a minute...let me get my tail...out from between my legs...OK, now where was I? "Little Blog?" Ooo, that hurts.

Seriously, I hadn't decided whether I was going to write about the BLS numbers or not. It's the holiday season and I'm not writing that much. And I'd rather write about more important issues. I'm not saying that the number of workers getting injured every year isn't important, I'm saying that the BLS data isn't important; that is, I don't think it can be believed, whether it's going up, down or sideways.

Why not? Let me count the ways:

• A recent study in the Journal of Occupational and Environmental Medicine shows the BLS estimates missed as many as 69 percent of all injuries.
  
  
• One reason is that OSHA's targeting system relies on employers' injury and illness reports, giving them an incentive to underreport in order to avoid inspections.
  
  
• Meanwhile, despite strong evidence that they discourage reporting, safety incentive program remain popular in American industry. As I wrote last year:

  Unions oppose safety incentive games as one of a variety of ways to discourage workers from reporting injuries or otherwise underestimate the rate of injuries and illnesses in the country. They are part of a management philosophy called behavioral safety, which assumes that workers' behavior is at the bottom of most health and safety problems (as opposed to hazardous working conditions) and that incentives (like money) or punishments will "correct" that behavior.

  Much more on behavioral safety at the Hazards Magazine website.
  
  
• Musculoskeletal disorders account for more than one-third of all injuries and illnesses involving days away from work and remain the biggest category of injury and illness, according to the BLS. OSHA has estimated that MSDs are understated by at least a factor of two and testimony delivered at the 2000 ergonomics standard hearings confirmed this estimate. OSHA made the problem worse in 2003 when it removed the column for reporting musculoskeletal injuries from its reporting form.
  
  
• I've written a number of times about large companies being caught not reporting injuries and illnesses. For example, a couple of months ago, OSHA fined General Motors Powertrain Corp $140,000 for not reporting almost 100 injuries and illnesses on the OSHA 300 Log.
  
  
• An article in the LA times last October described massive and deliberate underreporting at Southern California Edison. Why did they rig the numbers?:

  Because they had an incentive to do so. A 1997 Public Utilities Commission (PUC) program rewards or fines utilities for a number of measures, including employee safety. The decision to grant rate increases is also partially dependent on these numbers. Another measure that goes into the rate increase calculation is customer satisfaction. SCE was also found to have rigged those numbers as well, and agreed to return $14.4 million. Edison has now had to agree to return $20 million in safety awards already paid to Edison, plus $15 million pending for 2001 through 2003.

• And lets not forget the invisibility of occupational illnesses and deaths from occupational illness. As I wrote a couple of weeks ago:

  Revere at Effect Measure points out that by the most conservative estimate, 75 people die each week due to work-related cancer.

  If 75 people died in a train wreck or 75 people died in a building collapse or 75 schoolchildren in a school shooting it would make headlines in every newspaper in the country. But we won't read about it.
  
  Not only will 75 workers die today. But they did yesterday, too, and will tomorrow and so on, day after day after day. Bhopal in slow motion.

  And, of course, even those numbers don't include people who die of other work-related illnesses like silicosis, liver disease and black lung. Almost none of their names will ever be known.

  Nor do they show up in the BLS reports.
  
  
• I generally take workplace fatality statistics more seriously than injury and illness statistics on the assumption that it's harder (although not impossible) to hide a work-related dead body than to hide a work-related injury or illness. Workplace fatalities actually went up last year while injury and illness numbers were falling. In fact, if you look at the years 1992 - 2003, you find that the total number of injuries and illnesses fell 36% while fatalities fell only 12%. Seems like they should be running approximately parallel.

So what are the real numbers and which way are they going? Who knows? We haven't found an effective way to keep track, nor have we done what must be done to eliminate the incentives to underreport.

OSHA Director John Henshaw likes to shrug off criticism of his reign at OSHA by touting his "triple bottom line" -- bringing down workplace illnesses, injuries, and fatalities.

Well, I don't have an MBA and don't know much about bottom lines, but I do know that if a company based its bottom line projections on data as flimsy as the BLS data, they wouldn't be in business very long.

So, no, my lack of reporting has nothing to do with my "political agenda." In fact, I'm not using safety to further my political agenda; I'm using politics to further my safety agenda. There's a big difference.