Two-fifths of workforce at ‘high risk’ of being affected by automation: Report

AI, advanced robotics entering realm of cognitive, non-routine tasks, occupations

Nearly 42 per cent of the Canadian labour force is at a high risk of being affected by automation in the future.

That’s because the more recent rise of artificial intelligence and advanced robotics means automation is entering the realm of cognitive, non-routine tasks and occupations, such as driving and conducting job interviews, according to a report from the Brookfield Institute for Innovation + Entrepreneurship at Ryerson University in Toronto.

Forty-two per cent is a sobering number, even if just half becomes true, according to Sean Mullin, executive director at the institute.

“It’s a huge restructuring of the labour force over just a short period of time.”

Most of the high-risk occupations are in office support and general administration, sales and services, transportation and distribution, lower skilled technical occupations in health, natural and applied sciences, as well as manufacturing and construction labourers and assemblers, said the report. 

Individuals in these occupations have a much lower average income, at $33,000, which is nearly $29,000 less than occupations at a low risk. They are also less-educated, as 12.7 per cent of those in the high-risk category have a university degree, compared to 45.6 per cent of those in the low-risk group.

Jobs that are considered at a low risk of automation are linked to high skill levels and higher earnings, such as management and jobs in science, technology, engineering and math (STEM). 

To make its predictions, the Brookfield Institute used the 2013 and 2015 findings of both Carl Benedikt Frey and Michael Osborne from the University of Oxford and consulting firm McKinsey & Company. The former estimated the proportion of occupations that can be automated over the next 10 to 20 years while the latter looked at the percentage of work activities that could be automated using existing technologies.

“We thought by using both of these methodologies, it gave you two different approaches to address the same phenomenon,” said Mullin. “Interestingly enough, the results both came out to be 42 per cent of the Canadian labour force.”

On the other hand, 36 per cent of the employed labour force are at a low risk of being affected by automation while 22 per cent are at medium risk.

However, predictions are never perfect, said Sarah Anson-Cartwright, director of skills and immigration policy at the Canadian Chamber of Commerce.

“Many of us would be skeptical to see how that actually plays out. It’s not that we haven’t seen other books and dire predictions, but we also have seen how technology can spawn different types of employment, and the shift that can take place.”

While manufacturing, for example, has seen a huge shedding of traditional assembly line jobs, there has also been a shift towards more advanced manufacturing with different skills, she said.

“That’s something you have to be very focused on: What areas of a sector may then be the new areas of growth?” said Anson-Cartwright.

“We really have to be a little more sanguine when you get a big number like that, that we actually break it down and understand how we should be looking at both the nature of the sector where those occupations may be.”

Not all jobs will be lost, according to Mullin, as many will be restructured while new jobs will be created as the nature of occupations changes due to the impact of technology and computerization.

And using the Canadian Occupational Project System, the Brookfield researchers found occupations at a low risk of being affected by automation are expected to grow much more than those in the other risk profiles. There are expected to be nearly 712,000 new jobs in occupations with a low risk of being affected by automation, compared to nearly 360,000 new jobs in the high-risk group.

“The areas of the economy that we already deemed to be low risk are already projected to be the largest creators of jobs going forward, so that’s heartening,” said Mullin. “It’ll probably be a phenomenon where a large number of those jobs are eliminated, others are restricted, others the nature of the jobs… (is) changed significantly.”

The effect of automation on the high-risk area is going to be a net negative, he said, but the low risk could be an area where automation, computers and artificial intelligence end up being a net creator of jobs, such as computer programmers or AI technicians, where people work hand in hand with sophisticated software.

Moving forward, Canada’s public and private sectors can help to mitigate the negative effects of technology on employment, while ensuring Canadians can leverage technology to innovate and improve productivity and standards of living, said the report.

“It’s kind of all in our collective interest as a society to get our heads around this,” said Mullin. “From the private sector, in particular, if you’re a company that misses out on this trend, what it means is you’re going to be at risk from a competitive business environment that another company that is using labour more effectively, utilizing cutting-edge technology, is going to potentially outcompete you. So I think private companies are going to have to adapt to these changes, particularly if that’s where the market is going.”

But as soon as people talk about automation, there’s a fear it means losing staff, said Mave Dhariwal, director of the NAIT Shell Manufacturing Centre at the Northern Alberta Institute of Technology in Edmonton, which works to link productivity enhancement services with manufacturing solutions.

“We have to make work easier for staff by adopting technology so that the heavier work, the repetitive work, we do that with assistance from some kind of robotic or some kind of automation…. we have to expand the market reach so that we can actually try and get more business, so rather than laying off staff, we should really be looking for more staff, developing more staff because we’ve made the work easier. Because, if you think about it, globally, we want to 
really reach out to markets as far as we can.”

Small companies cannot afford to adopt technologies without testing them first, so to facilitate that, they can test a new product at the centre, he said. And if companies don’t want to shut down production to introduce a new product or system, staff can be trained at the centre first.

“That’s how we… encourage industry to adopt technology without having to invest money in it upfront.”

It’s also very disruptive to get rid of employees in one area while finding more qualified ones in another, said Mullin, so for an employee to be successful in this new world, it’ll be about building a set of skills called “digital literacy.” That means working constantly with computers and sophisticated software, and delivering jobs together with the tools — as opposed to separately, he said.

“There’s many ways we’d love to continue to get into this data and also start to craft what could be potential policy solutions, what education systems should be doing, how we should be preparing our children…. (understanding) what does it mean for our post-secondary institutions, what does it mean for our social safety net if there’s going to be a huge disruption, what does it mean for the private sector, on-the-job training or work-integrated learning-type programs — all these things are going to be critical components of dealing with this in the future.”

Connections between the private sector and post-secondary institutions are critical, along with having employers sending signals to government in terms of labour market information and skills focus, said Anson-Cartwright.

A report like this gives people a chance to take stock and consider how to evaluate and understand the types of skills that are needed to do well, she said.

“Don’t assume because you’re looking at getting into law that some of the more routine tasks might not be more automated, and there’d be less need. They talk about commoditization in a whole range of areas — even the filing of some legal documents that used to be done in a very manual, human-driven process may not need necessarily as much oversight.

“We would all benefit if we thought more in terms of not just knowledge and what you’ve studied but what skills did you require or are you requiring? And how transferable are certain types of skills?”

Latest stories