Economic Insights
Canadians’ Views on Automation and Net Job Creation in 1989: Were They Right?

Introduction

Prior to the COVID-19 pandemic, advances in artificial intelligence and robotics raised concerns that automation might lead to relatively high unemployment rates in the coming years (Brynjolfsson and McAfee 2014; Acemoglu and Restrepo 2019). However, concerns that technological progress might have adverse effects on employment growth are not new.

In 1989, Statistics Canada’s General Social Survey asked 9,338 respondents, “Do you agree or disagree that, on balance, computers and automation will create more jobs than they will eliminate?”

Canadians had differing views on this issue. Roughly one-third of respondents agreed (somewhat or strongly) that computers and automation would end up creating more jobs than they would destroy (Table 1). However, a greater proportion, about half, disagreed (somewhat or strongly).^Note The remaining 15% of respondents had no opinion on the issue.^Note

Less educated Canadians were less optimistic than their more highly educated counterparts. For example, 27% of men with no high school diploma agreed that computers and automation would create more jobs than they would destroy, compared with 50% of men with some postsecondary education.^Note



What happened? Thanks to job creation in the services-producing sector, total employment in Canada grew by about 6 million after 1989, rising from 13.0 million in 1989 to 19.1 million in 2019 (Chart 1).^Note

Not all sectors experienced employment gains. Employment in manufacturing fell from 2.1 million in 1989 to 1.7 million in 2019, largely because of employment declines observed since the early 2000s. Employment also fell in agriculture, forestry and logging, as well as fishing, hunting, and trapping. Nevertheless, the proportion of Canadians who were employed did not fall substantially from 1989 to 2019: in 1989 and 2019, roughly 62% of individuals aged 15 and over had a job.^Note

Assuming that respondents to the 1989 General Social Survey were considering “total employment” when they answered the question on automation, and assuming that employment trends since the late 1980s have been driven solely by computers and automation—a fairly restrictive assumption since many factors (e.g., growth of international trade, plant relocations abroad) likely played a role—the 24% of Canadians who strongly disagreed that computer and automation would create more jobs than they would destroy appear to have been wrong. Their counterparts who strongly held the opposite view (about 13% of respondents) appear to have been correct.

Data table for Chart 1



Data table for Chart 1
Table summary
This table displays the results of Data table for Chart 1 Total employment, Goods-producing sector, Services-producing sector and Manufacturing, calculated using millions units of measure (appearing as column headers).

	Total employment	Goods-producing sector	Services-producing sector	Manufacturing
	millions
1989	13.0	3.8	9.2	2.1
1990	13.1	3.7	9.3	2.1
1991	12.9	3.5	9.3	1.9
1992	12.7	3.4	9.3	1.8
1993	12.8	3.3	9.5	1.8
1994	13.1	3.4	9.7	1.8
1995	13.3	3.5	9.8	1.9
1996	13.4	3.5	9.9	1.9
1997	13.7	3.6	10.1	2.0
1998	14.0	3.7	10.4	2.1
1999	14.4	3.7	10.7	2.2
2000	14.8	3.8	11.0	2.2
2001	14.9	3.8	11.2	2.2
2002	15.3	3.9	11.4	2.3
2003	15.7	3.9	11.7	2.3
2004	15.9	4.0	11.9	2.3
2005	16.1	4.0	12.1	2.2
2006	16.4	4.0	12.4	2.1
2007	16.8	4.0	12.8	2.0
2008	17.0	4.0	13.0	1.9
2009	16.7	3.7	13.0	1.7
2010	17.0	3.7	13.2	1.7
2011	17.2	3.8	13.4	1.7
2012	17.4	3.9	13.6	1.7
2013	17.7	3.9	13.8	1.7
2014	17.8	3.9	13.9	1.7
2015	17.9	3.9	14.1	1.7
2016	18.1	3.8	14.2	1.7
2017	18.4	3.9	14.5	1.7
2018	18.7	3.9	14.7	1.7
2019	19.1	4.0	15.1	1.7
Notes: The goods-producing sector includes manufacturing; construction; utilities; agriculture; forestry and logging and support activities for forestry; fishing, hunting and trapping; and mining, quarrying, and oil and gas extraction. Source: Statistics Canada, Labour Force Survey.

Regardless of how these predictions are interpreted, it is clear that, in 1989, Canadians had varying opinions about the impact of automation on future employment. Although many were optimistic about how automation would affect net job creation, considerably more held less optimistic views.

This fact is a reminder that predictions regarding future employment levels carry a considerable degree of uncertainty. This is true not only in the short term—as can be seen in the wake of COVID-19—but also in the longer term. This fact also raises the possibility that Canadians whose predictions were wrong about the 1989-to-2019 period might end up being right when, 30 years from now, the 1989-to-2049 period is analyzed. Only time will tell.

References

Acemoglu, D., and P. Restrepo. 2019. “Automation and new tasks: How technology displaces and reinstates labor.” Journal of Economic Perspectives 33 (2): 3–30.

Brynjolfsson, E., and A. McAfee. 2014. The Second Machine Age: Work Progress and Prosperity in a Time of Brilliant Technologies. W.W. Norton & Company Publishers.