Workers' Bodies

Today the Standard had a guest post on ACC:

The investment losses have been a big part of it but there is also a rising accident rate stemming from our ageing population and climbing obesity rates, which has been foreseen by medical experts for some time. We cannot do much about an aging population really, but obesity is wholly avoidable with smart policy that has some guts behind it.

Why should we focus on obesity? Obese workers have a higher accident rate, take longer to recover, cost more treat and are out of work for a longer period of time. A 2007 Duke University study found that “obese workers filed twice the number of workers’ compensation claims, had seven times higher medical costs from those claims and lost 13 times more days of work from work injury or work illness than did nonobese workers”.

Although they don't provide a link I'm going to assume the guest poster is quoting from the press release about the study. Here's a link to the study itself for people who speak science article.

The numbers quoted are absolute numbers, they're not controlled for anything. In particular, they're not controlled for occupation.* I'm sorry to insult my readership by pointing this out, but the correlation between class and body size is pretty well established, as is the correlation between class and work-place accident rates.

Surprise! When the authors control for occupation (although not income, and managers appear to be treated as the same occupation as workers) the numbers look rather different. These numbers are expressed in risk ratios, whereby a control group is set at 1, and 2 means something is twice as likely when all the variables that are mentioned have been controlled for (full disclaimer, I could be lying, I don't understand statistics that well). The risk ratios for number of claims for people who have a BMI of over 25 range from 1.09 to 1.45. To understand how insignificant a risk ratio of that size is here are some of the risk ratios for occupational groups:

Laundry Staff: 7.35
Housekeeper: 6.44
Laboratory Animal Technician: 17.36
Inpatient Nurse: 4.01

The guest posts asks 'why is ACC costing so much?' And answers 'workers' bodies'. Even though its evidence is a study that demonstrates that the nature of work plays a far bigger role in the numberof workplace accidents than the nature of workers.

I'm a 'which side are you on' kind of a girl, and this post makes it very clear which side it's on. It blames workers and their bodies for workplace accidents. It chooses policing workers bodies, over fighting for workers bodies.

* I don't actually like debunking scientific research about fat, it seems to me to be conceding too much. Even if everything they said about the dangers of fat were true it wouldn't change my political analysis of fat at all.

** There are two other problems with those numbers. First that when it says 'obese' and 'non-obese' it appears to be comparing people with a BMI of between 18.5-24.9 and a BMI of 40+. In the article obese is defined as a BMI of 30+, so the terms used in the press release are not the same as those in the article, or the common medical use of those terms. I'm not going to dwell on that because I have less than no time for the BMI in the first place.

The other problem is that all the numbers apart from the numbers of claims made appear to be based on guesses at what the numbers might be rather than actual numbers:

Lost workday rates (days per 100 FTEs) were calculated by multiplying these stratum-specific claims rates by their corresponding mean number of lost workdays per claim. Similarly, multiplying the claims rate by the stratum-specific mean costs (including the amount already paid and the amount reserved) allowed calculation of cost rates (dollars per 100 FTEs) separately for medical and indemnity claims costs. Confidence intervals were calculated assuming that the number of events followed a Poisson distribution.

I'm not going to comment any more than that, because I don't speak science article, but will concentrate on the 'claims made' figure when explaining why this research doesn't prove what the standard thinks it proves.