The revolution of automation and artificial intelligence (AI) will be one of the most profound economic shifts of this century. It’s already well underway and public policy must meet the challenge.

Populists on the left and right often blame free trade for North American job losses in manufacturing, though automation is actually the main culprit. A 2015 study by the Center for Business and Economic Research at Ball State University concluded that of the 5.6 million manufacturing jobs lost in the U.S. between 2000 and 2010, 85 per cent were attributable to developments in technology, namely automation.

Automation and AI-induced job losses are certain to spread to other sectors.

A 2016 World Bank study found that in the coming decades, 47 per cent of all jobs in the U.S., 57 per cent of all jobs in the Organization for Economic Co-operation and Development countries (OECD, essentially the developed world), and 77 per cent of all jobs in China are susceptible to automation and AI.

Atlantic Canada won’t be immune to this shift.

For example, a 2016 study by the Brookfield Institute found that 41.5 per cent of New Brunswick’s employed labour force members are in jobs that are at “high risk of being affected by automation.” That means they have a 70 to 100 per cent probability of being affected over the next 20 years.  

The disappearance of so many jobs in such a short period could severely strain our existing social welfare models.

In the five fiscal years between 2012-13 and 2016-17, New Brunswick’s personal income taxes accounted for an average of 28 per cent of the province’s own-source budget revenue and an average of 18 per cent of all revenue. Federal transfers contributed an average of 36 per cent of the province’s revenue during the period.

An average of nearly 49 per cent of the federal government’s total revenue during these five years came from personal income taxes. What happens when potentially 40 per cent of the province’s and country’s income taxpayers suddenly cease to earn income?

Nor will it only be low-wage jobs hit by automation. The Brookfield study estimates the average income of workers with a high risk of job loss due to automation was $33,411, about $3,000 more than New Brunswick’s 2015 median income. An individual earning that amount still would have paid $2,756 in federal income tax and $2,065 in provincial income tax this year.

The impact on provincial and federal income tax revenue (and thus federal transfers) would push New Brunswick ever closer into fiscal free-fall.

So what to do about the impact of automation on the provincial job market and fiscal health?

The current model of social program provision would become unsustainable, both due to expanded demand and severely diminished revenue sources. Moreover, the social disruption caused by automation, if not mitigated, could produce unsavoury political outcomes.

Donald Trump’s presidential election victory in the United States was partly driven by anxious working class voters who attributed free trade to diminished job prospects that were far more rooted in automation. The rise of populist movement in Europe also drawing support from such anxieties.

One idea that demands serious consideration is a guaranteed annual income (GAI). A GAI, broadly speaking, works like this: citizens are provided with regular payments from the government that adjust according to a sliding scale based on the employment income. These payments would cease once a certain employment income threshold is reached. The aim would be to top up existing earned income for all citizens, whether employed or not, and would be free of conditions.

Such a scheme could only be affordable, especially in a low-employment future, if it replaced the suite of existing government programs, such as employment insurance, social assistance and public pensions. It may even demand the full or partial conversion of public health care funding into a personal health insurance fund, as some GAI proponents suggest. Existing programs like student loans could also be folded into the GAI program.

A GAI could also remove many of the perverse incentives and inefficiencies of our current social welfare model and liberate individuals to pursue more entrepreneurial lifestyles and regularly upgrade their skills via education.

Replacing our social welfare with GAI wouldn’t be without difficult adjustments for some. However, GAI may mitigate and diminish the much more severe consequences that an automation-induced jobs massacre could bring. It’s a policy discussion worth launching.

Patrick Webber is a research associate with the Atlantic Institute for Market Studies. He lives in Fredericton.


gai automation

The views, opinions and positions expressed by columnists and contributors are the author’s alone. They do not inherently or expressly reflect the views, opinions and/or positions of our publication.