Tuesday, November 1, 2016

OHS consultation on gas stations and convenience stores

Last week, I attended a government consultation about occupational health and safety for Alberta gas and convenience store workers. Between June 2015 and March 2016, there were 4 fatalities and 1 serious injury caused by violence and gas-and-dash incidents. There were three interesting things about this consultation.

First, the Minister of Labour attended, stayed for the whole presentation, and participated (I don’t ever recall a Tory minister doing that). There were also four ND MLAs, which suggests to me that action on this issue is imminent (perhaps a private member’s bill along the lines of BC’s Grant’s Law).

Second, this sector is profoundly unsafe. The lost-time claim (LTC) numbers provided suggest there were 11,576 LTCs between 2010 and 2014 (inclusive) across about 2200 employers. Using an annual average of 50,950 person-years worked, I think this means the annual LTC rate (claims per 100 person years worked) was 4.54 (versus a provincial average of 1.26 in 2014).

This rate of injury is crazy high (basically it looks like every employer had a LTC each year!), especially since LTCs represent only a fraction (say 10-20%) of all injuries. Around 0.7% of LTC injuries stem from assaults and violence. The most common causes of injury are over-exertion, bodily reaction, and struck by objects.

Third, there was a general sense that gas pre-payment (either at the pump or in store) was an acceptable safety measure for the government to impose (about 77% of stations have pay-at-the-pump technology installed). There are 12 reported cases of fuel theft per day in Alberta and pre-payment would eliminate them. That said, there is little evidence beyond the anecdotal that fuel theft is a significant source of injury.

By contrast, violence seems to be more of an issue in this sector. According to one hand out, there were 83 assaults resulting in LTCs from 2011-2015. More broadly, there appear to have been 750 victims of crime in this sector (victims include both workers and bystanders) in 2012.

Interestingly, there was little interest among employers in installing barriers between workers and customers. Employers cited ineffectiveness, cost, operational disruptions, and the off-putting experience of barriers for customers as reasons not to implement barriers.

I could not find any peer-reviewed literature on the efficacy of barriers in retail establishments. Barriers separating cab drivers from customers seem to reduce injuries. Obviously barriers are not a perfect form of hazard control (workers have to leave the cage eventually) but barriers seem to offer some protection from violence. The one study cited in the presentation was summarized as barriers showed mixed results. This looked like a consultant’s report so the quality of its evidence is unknown.

Violence prevention plans were touted by industry representatives as a better alternative than barriers. The literature seems to suggest that these programs do reduce the chance of robbery. But robbery is only one source of workplace violence.

Indeed, in looking at US convenience store homicides, one study found basically 80% of homicides had no robbery motive. This suggests that robbery prevention does not necessarily result in violence prevention (although it may reduce robbery-related violence).

One of the take-aways from the session for me was that hazard-control costs (both the initial investment and the effect on operations) remain a significant issue for employers. I was also struck by how the broad consensus sat uncomfortably with the data.

Specifically, there was broad acceptance of requiring gas pre-payment but little evidence that gas-and-dash was meaningfully associated with injury. Basically, pay-at-the-pump likely won’t make workplaces much safer (although it will reduce fuel theft).

At the same time, violence appears to be a much more significant risk than gas-and-dash. And there is some evidence barriers work to reduce injuries. Yet, there was little appetite for requiring barriers between workers and the public.

Despite that disappointing outcome, overall, I was heartened by the consultation. It was positive, clearly important the government, and there was a broad diversity of voices at the table.

-- Bob Barnetson


  1. Wow, that's kind of frightful. I'd guess there are probably a lot more injuries in those workplaces that simply go unreported because most of those jobs are likely minimum wage - or certainly below a living wage and many working them could be young or new to the country and not know how or even that they could make any sort of claim and when trying to scrape by on less than a living wage any sort of disruption of that wage could be disastrous, workers might just try to "soldier on" and keep showing up afraid taking time off to recuperate might mean a loss of that wage.

    I've never worked in a gas station or convenience store, but I've worked in a fair share of other small, late-night retail environments and know many others who have worked in gas stations and convenience stores. I'd guess some of the reluctance in wanting to put up any sort of barrier is that there is an expectation of the employee to be out from behind the counter (where, presumably, a barrier would go) doing things constantly - showing customers where to find products they are looking for, re-stocking those products, cleaning the store, and acting as a de facto security guard preventing and sometimes stopping shoplifting (I've had employers that, I swear, expected me to defend their stock with my life... all for minimum-freaking-wage!?). I'm not saying barriers can't be a workable solution, just saying I can see why they could be perceived as not being terribly effective, when workers won't even be behind them for a large part of the time.

    Perhaps it needs to be PART of a solution. Is there any data on when most incidences of violence take place? If most incidents of violence are happening at night maybe any store that wishes to say open past a certain time must have a barrier and employees are not required to leave that protected area during certain hours...?

    Anyway, interesting post. I need to check in on this blog more often.

  2. Thank for your comment, Tim.

    I think you have hit the nail on the head re: employer reluctance to install barriers. If an employer controls the hazard of violence with an barrier, then the employer is obligated to ensure employees are behind the barrier. This, in turn, profoundly affects employers’ abilities to direct employees to do tasks like stocking, cleaning, making coffee, etc. This is very disruptive to employers and may entail financial costs.

    Your question about the timing of violence is an interesting one. This isn’t my field and a quick scan of the literature didn’t kick up much that was applicable. The one article I did find (from Australia) suggests that most robberies (which is only one source of violence) occur at night. This timing likely reflects opportunity (i.e., fewer staff and customers to interrupt).


    In the consultation, employer reps repeatedly attacked the idea that night work was more dangerous. They focused on the arbitrary nature of Grant’s Law (which originally required barriers or two staffers on graveyard shifts). While it is true that the risk of robbery doesn’t simply stop at 6 am, there does seem to be a relationship between robbery and night work thus hazard controls might reasonably be linked to time of work.

    Looking at the arguments made as well as the lobbying that weakened Grant’s Law in 2012 (employers that have violence prevention programs are now exempt from the “two workers or a barrier requirement”), the position of many employers mostly seems to be about minimizing cost.

    It is interesting to see this consultation in some historical context. Looking at Occupational Exposure Limits (how much of a chemical workers can be exposed to “safely”), what we see is that employers typically resist regulation by arguing that there is a lack of scientific certitude (often while suppressing evidence they have collected which would establish this).

    The result is that OELs have historically been set at levels that employers are already achieving. Interestingly, the “safe” level of exposure pretty much always goes down over time. This suggests that “safe” is more of a political calculation by regulators (i.e., what will employers accept) rather than a scientific one.

    In this consultation, we have employers accepting pay at the pump (which is already available in most stations), which seems to have limited potential to reduce injuries (since there are few related to gas-and-dash). But employers are resisting a much more expensive intervention that appears associated with a much greater source of injury (violence).

    Make of that what you will.