Wednesday, May 25, 2011

Worker health in a boom-bust economy

This morning I'm giving a short presentation on worker health in a boom and bust economy to a class at the U of A. Tomorrow night, Yessy Byl and Kevin Flaherty will be picking up this theme in the Parkland Speakers Series.

Worker health in a boom-bust economy
CSL 350/360 & 550/560: Oil + Community: Health Equity in a Petro-Environment, University of Alberta, May 25, 2011.

Thanks for inviting me to speak today about the effect of a boom-bust economy on worker health. I’d like to start by talking about worker injury in Alberta and specifically the oil patch. Then I’m going to segue into a broader discussion of injury in a boom-bust economy and conclude with some observations about Alberta’s approach to workplace injury. I’m happy to take questions as we go and there is also 15 minutes for questions at the end.

Injury in Alberta
When we talk about workplace injuries in Alberta in public, there are basically three kinds of conversations. The first is the “my cousin Kelly” conversation where you hear the details of Kelly’s injury. These stories are compelling and often tragic. But they don’t give you a very good sense of what happens in the workplace overall.

The second kind of conversation—which is often a response to “my cousin Kelly”—is the “dumb worker” conversation. Basically workers are blamed for their injuries because they are reckless or careless. This ignores that a moment’s inattention can only result in an injury if the employer organized work in a manner that was inherently unsafe. If there is no latent hazard, inattention won’t result in an injury.

Almost no one wants to explore the question of why an employer might introduce a hazard into the workplace. That leads to all manner of awkward discussion about power and profit. It is easier to blame the workers.
The third kind of conversation is basically mindless community boosterism. “Workplaces have never been safer!” This is the kind of conversation you’ll have with a politician or business person. This is an interesting conversation because the other person will usually cite the latest government stats about injury.

These stats tell us important things. They show us that injuries are socially constructed—that is, what we call an injury can vary, often wildly. They show us that Alberta’s occupational health and safety system simply doesn’t work. And they show us whose side the government is really on in the question of workplace injury.

Counting Injuries
So let’s talk about how we count injuries in Alberta. The government uses two main measures: the lost-time claim rate and the disabling injury rate.

Lost-time claims are injuries reported to the workers’ compensation board (WCB) where the worker could not go to work the next day. The LTC rate is the number of LTCs per 100 person years worked. Expressing this number as a rate controls for changes in the size of the workforce so numbers become comparable across time.

Disabling injuries is a broader category. It includes injuries reported to the WCB where the worker either could not go to work the next day or could go to work but could not do their job in some way. The DI rate is also the number of disabling injuries per 100 person years worked.

There are a number of short-comings with these measures. For example, they obscure the actual number of injuries. A disabling injury rate of 3.09 per 100 person years worked sounds less bad than “this year we injured 53,0000 people so badly they could not do their job the next day”.

It is also possible for employers to “convert” more serious injuries to seemingly less serious injuries by how the employer handle the claim. So, you might need time at home but the employer tells you to come in and do light duties. This converts a lost-time claim to a modified work claim, which reduces the lost-time claim rate. Employers do this because it saves them money on their workers’ compensation premiums. Or your employer might simply tell you not to report an injury and, poof, your injury becomes statistically invisible.

These injury indicators also excluded many injuries. For example, if you get injured badly enough you need to go to the doctor or have rehab, but you can do your job the next day, you are not counted as injured. If you just need first aid or the rest of the shift off, you are not counted as injured. If you are outside of the ambit of workers’ compensation or don’t report your injury you are not counted as injured. And occupational disease is almost entirely absent from these stats as well.

A quick example is often useful. In 2009, Alberta’s workplace injuries were as follows:

Lost time claims: 28,688
Modified work claims: 24,625
Total Disabling injury claims: 53,313

These are all of the injuries Alberta officially “counts” when it talks about injuries. But this is a minority of overall injuries. For example, Alberta collects data about injuries requiring medical treatment but no modified work:

Medical aid claims: 95, 854

Adding in these numbers, suddenly we can see that actual number of injuries is 149,167—three times what the official injury numbers are. But, again, that’s not the whole story.

Not covered? 13% 171, 456
Not reported? 40% 285,760

Minor injuries? Occupational disease? Unknown but it likely doubles the count to about 500,000 injuries a year. This is important for three reasons:

1. Injuries are socially constructed. Official statistics report about 1 in 10 workplace injuries, skewed towards the most serious injuries.
2. OHS does not work. No reasonable person could conclude Alberta’s OHS system works when half a million injuries occur every year.
3. The government does not protect workers. That the government hides this massive number of injuries and runs a system using techniques that were a failure when they were first used in Ontario in 1886 tells us that they are playing for the employer’s team.

In any event, the LTC and DI rates are the indicators the government uses. So how does the oil and gas industry fare? Surprisingly well.

Oil patch injuries
According to government statistics, injury rates in upstream oil and gas declined each year from 2005 to 2009. The LTC rate went from 1.29 LTCs per 100 person years worked in 2005 to 0.70 LTCs per 100 person years worked in 2009. The disabling injury rate also trended down, going from 4.14 DIs per 100 person years worked in 2005 to 2.22 DIs per 100 person years worked in 2009 (Alberta, 2010).

These injury rates are lower than the average rates across the province (1.69 LTCs and 3.09 LTCs). Oil-and-gas injuries accounted for 3% of provincial LTCs and 5.25% of provincial DIs. Small employers tended to have higher claims rates. This might suggest they operate more dangerous workplaces (or operate workplaces more dangerously). Or, alternately, it may suggest they are less able to manage claims in a way that drives down the reported numbers.

Overall, though, this would appear to be a good news story. Before we get too excited, it is useful to have the raw numbers. There were 788 LTCs and 2497 DIs in the oil and gas industry in 2009, the vast majority (95%) being traumatic injuries (sprains, breaks, cuts, burns, etc.). That is to say, there were 2500 pretty serious injuries reported to the WCB in a “good” year (Alberta, 2010). Keeping in mind the 1 in 10 reporting ratio we observed province-wide, this could mean upwards of 25,000 actual injuries.

Interestingly, there were also a lot of fatalities in upstream oil and gas. In 2009, there were 14 workers killed, almost 13% of the provincial total. While the number of fatalities was small as an absolute number, fatalities are important because they are one of the hardest types of injury for employers to hide.

The fatality rate in upstream oil and gas is almost double the provincial average (Alberta, 2010). And 9 of these 14 fatalities were the result of a workplace incident. The provincial average would be around 5 deaths in 14 from workplace incidents and the rest from MVAs and disease. This suggests upstream oil and gas worksites tend to have more serious hazards than normal worksites.

The research literature on safety in Alberta’s oil and gas sector is fairly thin. I found two recent studies. Rothe (2008) interviewed 45 northern Alberta oilpatch workers about seatbelt use and found they “… believe that taking safety risks is an essential characteristic of who they are and where they work. Employers demand consecutive number of hours on the job and offer attractive incentives for working overtime that encourages risk-taking (p.226).” Houser (2010) presents an ethnography of rig hands, suggesting that the uncontrollable nature of the risks involved in derrick work (due to technology and time pressure) trigger a selective fatalism among workers.

What I’m struck by is how thin the literature is when both the injury data and popular culture suggest that oil-and-gas is very dangerous. I don’t know why that is the case but how society treats occupational disease gives us a hint. And occupational disease is also germane to the oil-and-gas industry.

Occupational disease
An occupational disease is a chronic ailment that results from work. For example, we often think of various respiratory illnesses such as black lung, asbestosis, and silicois. These illnesses are the result of exposure to hazardous substances in the workplace.

Occupational diseases are subject to significant under-reporting. Let’s consider occupational cancer because it is pretty scary. A 2011 estimate by Alberta Health Services (Curley, 2011) suggest that there are roughly 761 cancers diagnosed in Alberta each year with a strong or suspected link to occupational exposures (313.6-1283). These stats are conservative and don’t include non-cancerous occupational diseases.

So let’s return to the question of how accurate government statistics are. In 2008, only 31 claims for occupationally-related cancer were accepted by the Alberta WCB despite the government’s own stats which suggest that there were 760-odd new occupationally related cancers that year (Auditor General, 2010). This suggests 96% of occupational cancers are never reported in workers’ compensation statistics. This reflects a combination of murky causality, long latency periods and ignorance.

These factors, in turn, reflect that occupational contributions to disease is under-researched and “under promoted”. One reason is that companies have mis-represented the toxicity of their products. While cigarettes are perhaps the best-known example, corporate deceit is widespread and spans decades (Michaels, 2008). I bring this up because it suggests an answer regarding our earlier question: perhaps there is the limited research into oil-and-gas safety because it is bad for business.

The skeptical among you might ask, do oil companies really care about research on workers’ health and occupational hazards? That is a valid question, but research on worker health is not just about worker health. The boundaries between occupational hazards and environmental hazards are quite blurry and permeable.

In both cases there is a substance that causes adverse health effects. Workers are often the first and the most intensively exposed to hazardous substances. In this way, corporations actually use workers as guinea pigs. But then they often ignore the warning signs that show up.

Lead is a good example. Petrochemical workers got sick when lead was introduced into gasoline in the 1920s. No one paid any attention. General Motors, DuPoint and Standard Oil simply blamed the injuries on workers not following safety precautions. But worker carelessness wasn’t root cause of worker injuries and death—exposing workers to a toxin at work was.

As a result of blaming the workers, not only were workers injured, but a hazardous product became widely used. The United States now has up to 11 billion pounds of lead deposited in soil from car emissions. This constitutes a significant hazard to children playing outdoors. Viewed in these terms, suddenly research on worker health and occupational hazards becomes a much riskier proposition than it would first appear.

Injury rates in boom and bust
The broader theme of this course is worker health in a boom-bust economy. So what happens during the boom and the bust? The generally accepted wisdom is that with an economic boom comes an increase in the absolute number of injuries. This seems sensible: more activity means more workers which means more injuries.

The literature seems to bear this out (Boone and van Outs, 2006). Alberta’s injury stats are less clear on this point. I’d suggest that has to do more with aggressive claims management by employers during the last boom than a real change in this relationship.

The more important question is whether the rate of injury (i.e., injuries per X workers) goes up when there is a boom. This is a very complicated question. The arguments in favour of an increased injury rate during a boom swirl around more inexperienced workers on the job, new worksites, less maintenance, longer hours and/or a greater intensity of work. Conceivably, these conditions could cause the rate of injury to rise

There is some support for the assertion that injury rates rise during a boom and decrease during a bust. Brooker (1997) examined both acute and back pain claims in Ontario and found this pattern. US researchers note a similar results, although there are sometimes variations by industry—more dangerous occupations tend to see greater swings (Asfaw et al., 2010; Robinson and Shor, 1989; Kossoris, 1938).

Yet it is not clear that changes in claim rates mean an actual change in injury rates. That is to say, we might see an increase in injury claims rates but no real change in the rate of actual injury. This reflects that workers have some discretion in whether or not they will file a claim. Sometimes workers may be more willing to put up with pain than at other times. An interesting finding from the UK is that the rate of minor injuries in the UK is pro-cyclical (i.e., rising during a boom) while the rate of major injuries is not affected by economic change (Davies et al., 2009). On the face of it this appears a bit strange.

Analysis of injury data from 16 OECD countries shed some light on this. It found that fatalities rates tended not to change much during a boom (Boone and van Ours, 2006). Fatalities are hard to hide and stable fatality rates suggests that injury rates remains stable regardless of the boom or bust. But the study found that the rate of reporting claims went up during a boom and down during a bust.

The authors’ conclude that workers felt more comfortable making injury claims during a boom because they were less likely to be terminated for doing so. Basically a tight labour market means workers are less vulnerable if they report injuries. Conversely, during a bust, workers were less likely to make a claim because of fear of termination. In effect, what we’re seeing is that workers with injuries are more likely to tough it out during a bust thus claim data under-reports injuries during a bust.

So do booms increase injury rates? I’d say, assuming nothing else changes, the answer is likely no. But, during a boom in Alberta, we experience significant growth in high-injury occupations. So, while the rate of injury within occupations may stay constant, the proportion of the population engaged in high-risk employment may increase. This would create a boom-driven increase in the overall provincial injury rate as well as an increase in the absolute number of injuries.

I think considering workplace injury—in a petro-state or elsewhere—reveals that workers and employers see workplace injuries differently. This is because workers shoulder most of the consequences of injury while employers and their investors reap most of the rewards. For employers, injury risk is mostly an economic issue (Hilgartner, 1985). And the risk of workplace injury is cast as minimal, unavoidable and acceptable. This economic perspective dominates popular discussion and public policy (Iverson and Barling, 2005).

One implication of this economic approach is that, since perfect safety is unattainable, safety initiatives should be assessed on a cost-benefit basis. Put bluntly, safety should only be improved when it costs less to prevent the injury than the injury itself costs. Employers assert that they ought to make these decisions, because government regulation is said to cause rising prices, job losses, and a declining standard of living.

On the surface, this economic perspective appears quite sensible. Every activity does entail some risk. And risk reduction can be very expensive. Nevertheless, workers—those most often injured and killed—tend to see things differently.

Workers note that workplace injury is not a natural phenomenon that no one can control. Rather, the risks workers face reflect decisions employers make—decisions about what, when, where and how goods and services are produced. Employers make these decisions with the goal of maximizing profitability. In this way, injury is a cost imposed on workers by employers. And allowing employers to do this is a political choice by the state.

Workers also know that the most important consequence of health and safety risks is not economic. It is the injury and death of workers. Reducing injury, disease and death—not maximizing cost-effectiveness—is the pre-eminent goal of occupational health and safety activities.

That is not to say that workplace injuries don’t have economic consequences. Clearly they do. Injured workers cannot earn a living and lose then their houses. Society must pay for medical treatment. Employers profit from dangerous work. But these economic outcomes are secondary effects—byproducts of workers being exposed to the risk of injury and death by their employers.

The political perspective of workers and the economic perspective of employers start with contradictory views about the nature of risk in the workplace. Employers see risk as natural; workers see it as imposed. Consequently, their prescriptions for reducing risk differ. How governments choose to regulate workplace injury reflects the respective abilities of workers and employers to influence public policy. In Alberta, the government chooses to regulate ineffectively and tries to hide that with deceptive statistics. That tells you a lot about the relative power of employers and workers in this petro-state.

Alberta. (2010). Occupational injuries and diseases in Alberta: Lost-time claims, disabling injury claims and claim rates: Upstream oil and gas industries, 2005-2009. Edmonton: Employment and Immigration.

Alberta. (2011). Occupational disease fatalities accepted by the Workers’ Compensation Board. Edmonton; Employment and Immigration.

Asfaw, A., Pana-Cryan, R. and Ros, R. (2010). The business cycle and the incidence of workplace injuries: Evidence from the U.S.A. Journal of Safety Research. 42(1):1-8.

Auditor General. (2010) Report of the auditor general, April. Edmonton: Author.

Boone, J. and van Ours, J. (2006). Are recessions good for workplace safety? Journal of health economics. 25(6): 1069-1093.
Brooker, A. (1997) Back pain claim rates and the business cycle. Social science & medicine. 45(3): 429-439

Curley, P. (2011). The business case for cancer prevention: The economic burden of occupational cancer in Alberta. Paper presented at the Occupational Cancer Research Centre Research Day, March 23. Edmonton, Canada.

Davies, R. Jones, P. and Nunez, I, (2009). The impact of the business cycle on occupational injuries in the UK. Social Science & Medicine, 69(2): 178-182

Houser, D. (2010). Working hard and staying safe: Drilling rig hands in Alberta, in D. C. Wood (ed.) Research in Economic Anthropology. 30: 331-349
Hilgartner, S. (1985). The political language of risk: Defining occupational health. in D. Nelkin (ed.). The language of risk: Conflicting perspectives on occupational health. Beverly Hills: Sage: 25-65.

Iverson, R. and Barling, J. (2005). The current culture of workplace injury. Paper presented at the Association of Workers’ Compensation Boards of Canada public forum. July. Ottawa, Canada.

Kossoris, M. (1938). Industrial injuries and the business cycle. Monthly labor review. 46. 579.

Michaels, D. (2008). Doubt is their product: How industry’s assault on science threatens your health. New York: Oxford University Press.

Robinson, J. and Shor, G. (1989). Business-Cycle Influences on Work-Related Disability in Construction and Manufacturing. The Milbank Quarterly, 67, Supplement 2 (1): 92-113

Rothe, J. (2008). Oil workers and seat-belt wearing behaviour: The northern Alberta context. International journal of circumpolar health. 67(2-3): 226-34.

No comments:

Post a Comment