Automation, AI and Anxiety: Policy Preferred, Populism Possible

Who is fearful of automation and what do they want politicians to do about it? This paper finds a correlation between Canadians’ fear of job losses from automation and populist and nativist views—but also that Canadians favour traditional government policy approaches to job disruption, such as retraining, more than radical measures such as reducing immigration.

READ THE FULL REPORT

Together, automation and artificial intelligence (AI) have the potential to fundamentally reshape economics and social life. How will these trends affect politics and public policy? Will they expand or lessen the appeal of populism? Will they make it easier or more difficult for governments to shape public policy?

This report explores the potential for automation and AI to lead to widespread political and policy unrest and change in Canada. To examine this, we consider four related questions about automation and AI:

How knowledgeable are citizens about automation and AI?
What do they expect its effects to be for themselves, for employment and the economy, and for society?
How worried are they about the potential effects of automation and AI?
What kinds of politics and bundles of policy responses are citizens willing to support to confront the challenges (and opportunities) of automation and AI?

To understand citizens’ views on automation and AI and their related policy preferences, we surveyed 1,995 Canadians in May and June 2019. Our survey sample was drawn from multiple panels with quotas for age, gender and region, providing a representative sample of the population. Our goal is to understand how subjects’ objective exposure to automation and AI and their own beliefs about them—which may not align—relate to their preferences for various policy responses to the challenges of automation and AI.

Findings

We found that individuals have not aligned their expectations of the job loss effects of automation and AI with their own actual exposure. Individuals regularly underestimate their own exposure. When the time for reskilling comes, it will often be a shock, and a dislocating one at that. Policymakers will only be able to encourage people to take advantage of reskilling opportunities if those in the labour market understand their need for such reskilling. Accordingly, understanding how individuals can improve their knowledge of the particular skills they need for the future of their work is a pressing matter.

The political implications of our findings are as important as the technical policy implications:

Citizens—especially those who are worried about job loss from automation and AI—are willing to support a large bundle of policies to address the downsides of these technologies. Politicians have substantial room to manoeuvre to address their concerns. They should link policy offerings directly to the reasonable apprehensions individuals have about automation and AI.
Citizens’ concerns about automation and AI are not limited to job loss. People are also concerned about social and economic mobility and inequality. Politicians should pursue policy solutions which recognize not only automation and AI’s short-term disruptions, but also their potentially longer-term effects on social and economic dynamism. Politicians can recognize the transformative economic potential of automation and AI while also arguing that the gains from these technologies should be widely and reasonably shared.
Finally, despite our findings linking fear of job loss to populism and nativism, we do not find a direct and clear relationship between citizens’ fear of job loss and their exposure to automation to how they intend to vote in the upcoming election. Likewise, we do not find that fear of job loss is linked to citizens’ evaluations of which political party is best at managing technological change.

The playing field is open for enterprising parties who wish to take up this policy challenge. Our political parties should do so with a depth and thoughtfulness equal to the challenge.

READ THE FULL REPORT

APPENDIX: SEE THE DATA

Watch: Peter Loewen presents findings at Brave New Work conference

Thank you to our lead partner:

Thank you to our supporting partners:

About the Authors

Peter Loewen & Benjamin Allen Stevens

Peter Loewen is a Professor in the Munk School of Global Affairs & Public Policy and the Department of Political Science at the University of Toronto. He is interested in how politicians can make better decisions, in how citizens can make better choices, and how governments can address the disruption of technology and harness its opportunities.
He has published in leading journals of political science, economics, psychology, biology, and general science, as well as popular press work in the Washington Post, San Francisco Chronicle, Globe and Mail, National Post, and Ottawa Citizen. His research has been funded by SSHRC, the European Research Council, the Government of Ontario, and other organizations. He regularly engages in public debate and acts as a consultant to several public and private organizations.
Previously, he served as the Director of the School of Public Policy & Governance and the Centre for the Study of the United States at the Munk School of Global Affairs. He was previously a fellow at the Center for Advanced Study in the Behavioral Sciences at Stanford University and has held visiting positions at Princeton University and the University of Melbourne.

Benjamin Allen Stevens is a Research Associate at the Munk School of Global Affairs and Public Policy, and the Project Manager & Lead Programmer for the Canadian Election Study. His research interests are in political behaviour, public opinion, survey research methods, and technology policy. He has extensive experience writing, programming, and running public opinion surveys, as well as analyzing the resulting data. He was the Project Coordinator and survey programmer for the Local Parliament Project, the largest election study fielded in Canada. He has an MA in Political Science from the University of Toronto, and a BA from Ryerson University.