Automation and artificial intelligence are reshaping the economy.
But how disruptive will their long-term impact be? Will the jobs available now still be around when today’s students hit the workforce? What skills will give kids the best chance to thrive in tomorrow’s uncertain labor market?
The country’s top economists, technologists, and policymakers are sharply divided.
To help K-12 educators and policymakers make sense of the debate, Education Week talked to more than two dozen experts and reviewed dozens of reports and research studies seeking to address these questions.
Here are 10 studies you need to know:
1. Artificial Intelligence, Automation, and the Economy (Executive Office of the President of the United States, December 2016)
Former President Barack Obama’s top advisers on technology and the economy sought to better understand the impact of artificial-intelligence-driven automation on the U.S. job market. They concluded that such advances “will open up new opportunities,” but also “have the potential to disrupt the current livelihoods of millions of Americans.” The report says workers without a high-school diploma, making less than $20 per hour, will be the most vulnerable. Without the right policies, Obama’s advisers warned, technology’s march could lead to a “winner-take-all” economy marked by increasing inequality.
Still, though, the White House sought to increase investment in artificial intelligence, while arguing that improved education and job training and a stronger social safety net could help ease the massive transition already underway in the economy.
2. Information Technology and the U.S. Workforce: Where Are We and Where Do We Go From Here? (National Academies of Science, Engineering, and Medicine, 2017)
Under Obama, the National Academies listed above also convened in 2015 a Committee on Information Technology, Automation, and the U.S. Workforce to examine the impact of technology on the job market. A decorated panel of experts reached consensus that big disruptions are coming.
“Computers have become increasingly competent at both physical and cognitive tasks that have previously been done primarily by humans, such as speech recognition, identifying faces and other objects in images, interpreting text, analyzing medical data, driving cars, and many other tasks,” the report reads. “The workforce will be increasingly affected as more and more cognitive tasks become fully or partly automatable.”
Put in simpler terms: Robots and algorithms are coming for middle-skill, middle-class jobs, too. It’s unclear whether the economy will be able to create enough new jobs to keep up. Either way, there will likely be significant shifts in the way work is organized and wealth is distributed.
3. AI, Robotics, and the Future of Jobs (Pew Research Center, August 2014)
As part of its study of technology and the internet, the Pew Research Center interviewed 1,896 experts about the economic impact of advances in robotics and artificial intelligence. Those experts were sharply divided.
Half envisioned a “future in which robots and digital agents have displaced significant numbers of both blue- and white-collar workers,” and many expressed “concern that this will lead to vast increases in income inequality, masses of people who are effectively unemployable, and breakdowns in the social order.” Many of these more-pessimistic experts were worried about automation in white-collar work and the shortcomings of the education system.
The other half of experts, though, expected that technology would not displace more jobs than it creates by 2025. The nature of work will likely change substantially, these experts believed, but they also expressed faith that human ingenuity will lead to new types of jobs and new ways for people to participate in the labor market. After all, these experts said, that’s what’s happened during other eras of big technological change.
4. The Future of Employment: How Susceptible Are Jobs to Computerisation? (Carl Benedikt Frey and Michael A. Osborne, September 2013)
This influential paper set off alarm bells and helped kickstart the future of work conversation. Using statistical techniques, Frey and Osborne estimated that about 47 percent of total employment in the United States was at high risk of automation in the coming two decades. That projection was so high, they wrote, because “algorithms for big data are now rapidly entering domains reliant upon pattern recognition and can readily substitute for labor in a wide range of non-routine cognitive tasks.”
Among the most vulnerable sectors, according to Frey and Osborne: transportation and logistics, office and administrative support, the service industry, and manufacturing and production occupations.
5. The Risk of Automation for Jobs in OECD Countries: A Comparative Analysis (Melanie Arntz, Terry Gregory, and Ulrich Zierahn, June 2016)
Frey and Osborne’s work sparked a flurry of follow-up studies, including this report by the Organization for Economic Cooperation and Development. The authors argued that focusing on automation’s impact on entire occupations, rather than specific tasks, had led to “overestimation” of the number of jobs at risk.
Arntz, Gregory, and Zierahn looked at data for 21 OECD countries, including the U.S. They found that, on average, about 9 percent of jobs were automatable.
While “automation and digitalization are unlikely to destroy large numbers of jobs,” the researchers argued, “low-qualified workers are likely to bear the brunt” of the changes that will come.
As a result, they said, the challenge ahead lies in addressing economic and social inequality, in part through improved job training and re-training.
6. A Future That Works: Automation, Employment, and Productivity (McKinsey Global Institute, January 2017)
Analysts at the McKinsey Global Institute have also embraced the look-at-specific-skills-not-whole-occupations approach.
According to their analysis, about half the work currently done by U.S. employees is automatable. But that work is spread out across a wide variety of occupations. So while fewer than 5 percent of jobs are likely to be automated away entirely, McKinsey Global predicted, existing technologies are already capable of taking over a third or more the tasks required in the majority of jobs.
Among the work-related tasks most at risk of automation: anything requiring routine physical labor, operating machinery in predictable physical environments, or collecting and processing data.
7. The Growth of Low-Skill Service Jobs and the Polarization of the US Labor Market (David H. Autor and David Dorn, August 2013)
The premise of this seminal study: While technology and automation have helped drive wages up for highly educated workers, they’ve also helped push poorly educated workers into service-sector jobs. The result is a more-polarized, more-unequal labor market.
Here’s how Autor and Dorn put it, in their own words:
We hypothesize that recent computerization has substituted for low-skill workers in performing routine tasks while complementing the abstract, creative, problem-solving, and coordination tasks performed by highly educated workers. As the declining price of computer technology has driven down the wages paid to routine tasks, low-skill workers have reallocated their labor supply to service occupations, which are difficult to automate because they rely heavily on dexterity, flexible interpersonal communication, and direct physical proximity.”
8. Robots and Jobs: Evidence From U.S. Labor Markets (Daron Acemoglu and Pascual Restrepo, March 2017)
Here, MIT economists Acemoglu and Restrepo sought to determine what happens to jobs in a given labor market when the companies in the area adopt industrial robots. They looked at the time period between 1990 and 2007, finding that each new robot costs about six jobs while also leading to declines in wages.
The auto-manufacturing areas of the Rust Belt and Appalachia have been hardest hit, Acemoglu and Restrepo found. But in general, U.S. companies have been relatively slow to adopt industrial robots, they said, and overall, only between 360,000 to 670,000 jobs have been lost.
If the pace of such automation accelerates, as many experts predict, that number could grow quickly, they warned.
9. Skill Demands and Mismatch in U.S. Manufacturing (Andrew Weaver and Paul Osterman, January 2016)
Despite widespread fears of automation and constant talk of U.S. workers lacking the skills that employers need, Osterman and Weaver reached different conclusions. Through extensive surveys and interviews with U.S. manufacturers, they found that demand for advanced math, computer, and problem-solving skills was “generally modest.” As a result, Osterman said in an interview with Education Week, schools would be wise to focus first and foremost on cultivating students’ basic skills—including reading, algebra and “community-college level math,” and interpersonal abilities, such as communicating and working well in teams.
10. The Zombie Robot Argument Lurches On (Economic Policy Institute, May 2017)
“There is no evidence that automation leads to joblessness or inequality,” according to EPI.
So instead of worrying about a potential robot apocalypse that may never happen, the economic think tank contends, we should focus on the problems the U.S. labor market already has.
“Rather than debating possible problems that are more than a decade away, policymakers need to focus on addressing the decades-long crisis of wage stagnation,” their report reads. “And as it turns out, policies to expand good jobs and increase wages are the same measures needed to ensure that workers potentially displaced by automation have good jobs to transition to.”
That involves more than just education and job training, EPI believes. Instead, it will take lifting the minimum wage, restoring lost collective-bargaining rights for workers, a fairer tax system, and other policy changes.
BONUS: Technology and the American Economy, Vol. 1 (National Commission on Technology, Automation, and Economic Progress, February 1966)
Just in case you thought any of this was new, here’s the report prepared by a blue-ribbon panel created by President Lyndon Johnson more than 50 years ago.
Since the early 1960’s, civil-rights leaders such as Martin Luther King, Jr. had been warning of the impact of “automation and cybernation” on workers, especially those who were poor and/or Black. In a 1962 speech to the United Automobile Workers, for example, King warned that “labor today has grave problems of employment, shorter hours, old age security, housing and retraining against the impact of automation.”
The committee created by Johnson to investigate the challenge recognized the rapid pace of technological change in the economy.
But this was the committee’s bottom-line conclusion:
“The basic fact is that technology eliminates jobs, not work.”
See also:
- COMPLETE COVERAGE: Schools and the Future of Work
- The Future of Work is Uncertain. Schools Should Worry Now
- Preparing Students for Tomorrow’s Jobs: 10 Experts Offer Advice for Educators