New University of Toronto study looks at soaring cost of building transit in Canada

A new study released by the University of Toronto’s School of Cities is showing how costs to build major transit projects in Toronto and elsewhere in Canada has soared in recent years.

“Somehow in peer countries like Italy, Spain, Turkey, South Korea, the cost per kilometre has been going down while it’s been going up sharply in Canada, and not just Canada (but) the other what we call Anglosphere countries so Australia, U.S., U.K.,” Karen Chapple, the school’s director and an editor of the study, said.

A comparative study called Understanding the Drivers of Transit Construction in Canada examined transit projects in Toronto dating back around 70 years. The study’s authors used data from transportation researcher Stephen Wickens. It showed the per-kilometre cost adjusted to inflation and charted each of the projects in 2023 dollars.

“In an attempt to close the transit infrastructure gap, cities across Canada are investing tens of billions of dollars into capital expansion through the construction of additional mass transit infrastructure,” the study’s authors wrote.

“However, over the past two decades, the cost of building new transit infrastructure in Toronto (and across Canada) has increased significantly, at a rate far beyond inflation.”

While building (in various stages) the Bloor-Danforth line and the Yonge-University-Spadina line (Downsview to Finch stations) between the mid-1950s and mid-1990s was less than $150 million a kilometre, that cost started creeping up after 2000.

The Sheppard line (opened in 2002) was around $200 million a kilometre, the Line 1 subway extension to Vaughan from Sheppard West (opened in 2017) approached nearly $400 million a kilometre, the Line 1 subway extension into Richmond Hill is forecast to be closing in $800 million a kilometre (set to open in 2029 and 2030) and the updated Ontario Line is forecast to be more than $1 billion a kilometre.

“The Ontario Line is 10 times the cost of the original Yonge Street subway line, even controlling for inflation, 10 times as expensive and that’s partly because we’re digging now through established neighbourhoods. There’s lots of mitigation around that, lots of extra costs to protect people, but these other factors at work as well,” Chapple said.

“It is mind-blowing.”

The study’s authors said they looked at data from 60 countries and 1,083 LRT, subway and rail projects to establish average per-kilometre costs for building transit lines. They determined the global average is $242 million per kilometre.

The authors said Canada had the ninth-highest costs and the average came out to $396 million per kilometre. New Zealand was the highest at $1.04 billion per kilometre followed by Qatar and Hong Kong ($949 million per kilometre each). The three countries with the lowest costs were Chile at $89 million per kilometre followed by Spain ($95 million per kilometre) and South Africa ($105 million per kilometre).

When analyzing the cost differences, the authors noted environmental, market, housing, cost of living and labour were often cited as contributing factors.

“Our benchmarking analysis, however, reveals only a slight correlation between average construction cost and contextual factors such as GDP, unionization rates, or cost of living. Even among OECD nations high on the human development index, average per-kilometre costs range across the full spectrum of our database,” the study said, adding there is commonality among Anglophone countries.

“The problem is not so much with English itself; rather, that these nations share a common institutional history, exchange ideas, and learn from one another. As such, our benchmarking investigation concurs with other studies of global transit costs, strongly indicating that national costs are associated most
closely with project delivery practices, policies, and governance.

“In fact, several domestic cases of low-cost transit construction demonstrate that effective project delivery is possible in the Canadian context with a shift away from traditionally Anglophone project delivery practices.”

Chapple added that “soft” costs have become a huge driver in costs in Canada. She defined soft costs as things like contingencies (money set aside for inflation and unforeseen expenses), acquiring land, planning, project management, design and engineering,

“The headline here really is that these transit systems cost more to plan than they do to build,” Chapple said.

“There is a tradition in Canada of over-design. So engineers are designing for the worst-case scenario. Just [a] comparable example, here is how we plan for Black Friday. We plan to build parking lots just for that one day a year. That’s what they’re doing with transit systems.”

She said two other factors are at play too: An eagerness to cater to different groups and not having enough experts on government payrolls.

“We’re … overly responsive to external stakeholders, so when you have a community complain we tunnel deeper and that adds to the cost,” Chapple said.

“There’s sort of a lack of in-house capacity, in-house knowledge, in-house expertise and that’s costing the country and our taxpayers a lot as we rely on professional consultants and then we accept their cost escalations without really much critical thinking.”

CityNews contacted Metrolinx, the provincial transportation agency overseeing many signature transit expansion projects in the Greater Toronto and Hamilton area, to ask about the study’s findings. While a representative acknowledged receipt of the request, a response wasn’t received by the time of publication.

“There is no single driver of transit construction costs, nor is there a silver bullet to bring high-cost projects in line with low-cost counterparts nor is the problem of high costs unique to Metrolinx,” the study’s authors wrote as they reviewed a variety of data gathered by Metrolinx.

“High-cost jurisdictions, like Toronto, experience cost escalation through a series of compounding factors … from planning and construction to institutional and procedural inefficiencies.”

While they said further research is needed, addressing the retention of government staff with expertise, potentially reassessing the over-cautious approach to managing risk, increasing transparency and reducing political micromanagement were cited as issues in the paper.

Meanwhile, Chapple said governments in Canada need to start incorporating techniques and approaches seen in other countries overseas.

“We don’t learn enough from the best practice cases around the world. We’re too insular. We’re too prone to imitating the U.K. and the U.S., and they’re not doing it very well,” she said.