IWH study in construction sector suggests unionized firms are safer

First industry-wide study in Ontario by Institute for Work & Health finds unionized construction workers report more claims overall but fewer claims that result in time off work

Published: November 10, 2015

It’s a question that has come up over the years: are unionized workplaces safer? In a new groundbreaking study, Institute for Work & Health (IWH) researchers found unionized construction workers were, overall, more likely to file work-related injury claims than their non-unionized counterparts. However, they were less likely to file lost-time claims (i.e. claims that resulted in missed days of work).

The study was published online in September as an open-access paper by the Journal of Occupational and Environmental Medicine (doi:10.1097/JOM.0000000000000562).

The study finds rates of lost-time claims at unionized construction workplaces in Ontario were 14 per cent lower than at non-unionized workplaces. When taking into account all types of injury claims filed with Ontario’s Workplace Safety and Insurance Board (WSIB), unionized workers were 13 per cent more likely to make a claim. Claims that did not result in time off, known as no-lost-time claims, accounted for this difference; these claims were 28 per cent higher at unionized workplaces.

These findings suggest that unionized workers may be more likely to report injuries, including injuries that don’t require time off work, at workplaces where managers and supervisors are committed to safety, says IWH Senior Scientist Dr. Ben Amick, co-lead investigator on the study.

While unionized workers may be more inclined to make work-related injury claims, these findings suggest that their claims are less likely to be of a serious nature, he adds.

The lower rates of lost-time claims might also suggest that unionized workplaces are safer, says Associate Scientific Director Dr. Sheilah Hogg-Johnson and project co-lead.

It could be they do a better job educating workers, in part through apprenticeship training. They may have more effective health and safety programs and practices. They may give workers more voice to influence the health and safety of their work environments, and to report not only injuries, but also near-misses, she adds.

Studies comparing unionized and non-unionized workplaces are very difficult to do because of a lack of data. To examine whether unionized workplaces are safer, one would need data from all union-certified employers, as well as all non-unionized workplaces. Including just a sample in either category would raise questions about selection bias (for example, an over-representation of firms with untypically good or untypically bad health and safety records).

The involvement of the study funder, the Ontario Construction Secretariat (OCS), helped overcome this barrier. Comprising 25 building trade unions and employers in the province’s industrial, commercial and institutional (ICI) construction sector, the OCS gave the research team access to a very comprehensive database of unionized firms in this sector. To ensure research independence, the OCS agreed that study results would be published, whatever the outcome.

The team used WSIB data to compile a full list of construction firms in the ICI sector. The painstaking portion of this research involved linking the OCS and WSIB datasets to determine which of these firms were unionized and which were not. All companies with at least two employees during the study’s seven-year window were included. Any company that had only one full-time employee or fewer the entire seven years had to be removed, as these companies were likely owner-operator businesses.

The final sample from the seven years included 5,800 unionized firms employing 720,000 full-time equivalent workers and 39,000 non-unionized firms employing 810,000. Because the unionized firms tended to be larger on average than the non-unionized firms (hence, more likely to have more resources for health and safety), the researchers took firm size into account when reporting their findings. They also took into account industry subgroup (because of the different hazard levels linked to different types of work in each subgroup), and postal area (because of regional variations in safety culture) and business complexity as defined by the number of industry subgroups a company worked in (for the same reason that size was taken into account).

After these influences were taken into account, the findings showed that:

  • total claim rates were 13 per cent higher at unionized firms;
  • no-lost-time claim rates were 28 per cent higher at unionized firms;
  • lost-time claim rates were 14 per cent lower at unionized firms; and
  • claim rates for musculoskeletal injuries were eight per cent lower at unionized firms.

Rates of critical injury claims, a type of lost-time claim for severe injuries like amputations, burns, blindness, etc., were 29 per cent lower at unionized firms when taking regional variation, business complexity and industry subgroup into account. Due to the small number of critical injuries, however, the researchers could not account for firm size when looking at critical injury rates.

What might be an explanation for this “union safety effect”—this apparent link between union certification status and safety outcomes? In addition to the possible reasons above, it may be that unionized workers are older and more experienced, or maybe unionized firms have less turnover, says Hogg-Johnson.

We can only speculate on the reasons, adds Amick. But we’re now examining the organizational policies and practices at unionized construction firms. This research hasn’t been done before, and we hope it will help us understand how construction workplaces are different, and what role unions play in producing safety outcomes.