Globe and Mail reporter Tavia Grant was searching for Canada’s deadliest job.
In 2012, ProPublica and PBS Frontline found that tower climbing—scaling communications towers for maintenance—was a dangerous vocation. Although 93 climbers died on the job in the previous eight years, the fatalities weren’t listed in the U.S. Occupational Safety and Health Administration’s workplace accident database. They found the tower industry’s death rate was about 10 times higher than the construction industry’s between 2003 and 2010.
Grant wanted to replicate the research in Canada.
Follow her steps to complete the story.
In January 2017, Grant ordered data from the Association of Workers’ Compensation Boards of Canada (AWCBC) on annual worker fatalities, broken down by occupation, from 2006-2015. The data didn’t have every worker death in Canada, but the occupations with the most fatalities were truck drivers and firefighters, among others. Total worker deaths were nearly 10,000.
Grant sent the data to Statistics Canada, which produces customized reports upon request, to calculate the annual fatality rate. By the end of the month, Grant had the first set of information.
Throughout February, Grant analyzed the data, knowing the story had significant public interest.
The data suggested truck drivers had the absolute highest number of deaths—773 between 2006 and 2015. However, when converted into proportional fatality rates, it showed firefighters were one of Canada’s most dangerous occupations. After about two weeks reporting on firefighters, Grant discovered they were statistically incongruous. Unlike other professions, special legislation in many provinces automatically claim cancer as a work-related illness for firefighters. The studies Grant wanted to replicate focused on sudden on-the-job deaths—called traumatic deaths—and not long-term illnesses.
In order to accurately compare Canada to the U.S., Grant needed the same type of data ProPublica had analyzed. She had to order more data about traumatic deaths.
Around April, Grant contacted the Transportation Safety Board of Canada (TSB) and kept in touch every month. She sent a request for occupational fatality data from the AWCBC with a focus on traumatic deaths.
The new data showed that there were 1,758 traumatic deaths between 2011-2015. Truck drivers, construction trades helpers and labourers, material handlers, and air pilots were among the occupations with the most fatalities. She sent the data to StatsCan to convert the proportional risk of traumatic death rates.
The returned data showed fishing and trapping as the industries with the highest traumatic death rate proportional to the number of paid workers in the field. Statistically, logging was the deadliest job, but of the six most dangerous occupations, three were related to fishing.
For a small industry, fishing had shockingly high death rates. Grant then verified the calculations with Globe data scientist Shengqing Wu and cross-checked the information with the TSB data which illustrated that fishing industry fatality rates hadn’t improved since 1999.
It became clear that Canada was lacking data, but Grant’s team persisted. She knew the investigation could create positive change if done properly. Now, the story needed a human interest element.
Grant believed meeting sources in person led to better details and more emotional storytelling. Although the story was still data-driven, Grant knew connecting with readers’ emotions would resonate.
She visited a fishing community in Nova Scotia with the highest death toll of fishers of all provinces. Provincial fatality rates weren’t available, so Grant looked at sheer numbers. The Nova Scotia government was actively trying to decrease the number of deaths. Grant talked to fishers and families who’d lost loved ones. The families were “remarkably open” to her investigation.
After returning from Nova Scotia, Grant was on vacation for August and put the story aside. In September and October, Grant analyzed and double-checked the data, did follow-up interviews, fact-checked, and wrote. Despite the lack of available data, and after months of rigorous work, Grant’s story was finally published on October 27, 2017.
Days after the article was published, the federal Ministry of Innovation, Science and Economic Development released a statement. “Statistics Canada is prepared to explore collection of data from the provincial and territorial ministries of labour and the workers’ compensations boards.” The last national analysis was produced in 1996.
“I felt encouraged that our journalism had an impact in influencing federal policy and in raising public awareness about an underreported topic,” Grant said in an email. It took a year, but the results speak for themselves.