The Question with AI Isn’t Whether We’ll Lose Our Jobs — It’s How Much We’ll Get Paid

jan18-31-822800904-anucha-sirivisansuwan
anucha sirivisansuwan/Getty Images

The basic fact is that technology eliminates jobs, not work. It is the continuous obligation of economic policy to match increases in productive potential with increases in purchasing power and demand. Otherwise the potential created by technical progress runs to waste in idle capacity, unemployment, and deprivation. —National Commission on Technology, Automation and Economic Progress, Technology and the American Economy, Volume 1, February 1966, pg. 9.

The fear that machines will replace human labor is a durable one in the public mind, from the time of the Luddites in the early 19th century. Yet most economists have viewed “the end of humans in jobs” as a groundless fear, inconsistent with the evidence. The standard view of technical change is that some jobs are displaced by the substitution of machines for labor, but that the fear of total displacement is misplaced because new jobs are created, largely due to the technology-fueled increase in productivity. Humans have always shifted away from work suitable for machines and to other jobs. This was true in the 1930s, when the shift was away from agriculture, through the 1990s and early 2000s, when the shift was largely out of manufacturing.

However, the expansion of what can be automated in recent years has raised the question: Is this time different?

It doesn’t have to be. Yes, there are reasons for concern, both technical and political. Machines are now able to take on less-routine tasks, and this transition is occurring during an era in which many workers are already struggling. Nonetheless, with the right policies we can get the best of both worlds: automation without rampant unemployment.

Is This Time Different?

To date, automation has meant industrial robots and computer hardware and software designed to do predictable, routine, and codifiable tasks requiring physical strength and exertion, and the repetition of logical tasks, such as calculation. With robotics, artificial intelligence, and machine learning, what we call automation seems poised to take on a greater share of high-productivity jobs and a range of tasks that were previously the domain of humans. These are tasks requiring problem solving, decision making, and interaction within a less-than-fully-predictable environment. Automation of this sort includes self-driving cars and diagnosing disease.

Insight Center

Automation anxiety is made more acute by a labor market that has tilted against workers over the last 30 years, with increasing income inequality and stagnant real wages. Wage growth has not kept up with productivity growth; labor’s share of GDP has fallen and capital’s share has risen. The social contract established after World War II, where hard work and loyalty to the firm were met with rising wages, benefits, skills training, and economic security from firms no longer characterizes much of the American workplace. The “fissured workplace” — where firms focus on their core competencies and contract out everything else — results in low pay, few benefits, and job insecurity for workers. The share of workers in alternative work arrangements, as independent contractors, franchisees, and in the gig economy, is growing substantially, from 10.7% in 2005 to 15.8% in 2015. The old structures of the postwar labor market are not up to the task of the 21stcentury wave of automation, particularly for the low- and middle-skill workers already disadvantaged by previous skill-biased technological change and globalization. While technology and globalization have spurred competition, efficiency, and dynamism, the gains have not been shared by all. The unequal distribution of the gains is not a technical destiny; it is the work of institutions, business, and governments.

Will Robots Take All the Jobs?

Currently, most automation involves routine, structured, predictable physical activities and the collection and processing of data. Generally, these tasks form the basis of occupations in manufacturing, professional and business services, food service, and retail trade. Looking ahead, these tasks will continue to have the highest potential for advanced automation. Currently, less than 5% of occupations are entirely automated, and about 60% of occupations have at least 30% of tasks that can be automated. Based on these estimates, there is considerable potential for the spread of advanced automation. What is less knowable is how many new jobs will be created by automation-related productivity growth and how humans and machines will work together.

It’s likely that humans will continue to dominate machines in a variety of skills, including creativity, interpersonal relations, caring, emotional range and complexity, dexterity, mobility. Luckily, we know there will be ample opportunities in these jobs. The Bureau of Labor Statistics issues periodic occupational growth projections, and in its most recent report, for the time period 2016 to 2026, 11 of the top 25 fastest-growing occupations are health care–related, where human-dominant skills are essential. These occupations include home health aides, personal care aides, physician assistants, nurse practitioners, physical therapy assistants, and aides. Some of these occupations require a four-year degree and post-baccalaureate training (nurse practitioners, physician assistants), but some require on-the-job training and certification with a high school diploma (home health aides, personal care aides, physical therapy aides).

Series

The Economy in 2018

However, even though jobs where humans have absolute advantage may be narrowing, there is little reason to expect an end to human work. The reason stems from a classic idea in economics: comparative advantage.

Even in a world where robots have absolute advantage in everything — meaning robots can do everything more efficiently than humans can — robots will be deployed where they have the greatest relative productivity advantage. Humans, meanwhile, will work where they have the smallest disadvantage. If robots can produce 10 times as many automobiles per day as a team of humans, but only twice as many houses, it makes sense to have the robots specialize and focus full-time where they’re relatively most efficient, in order to maximize output. Therefore, even though people are a bit worse than robots at building houses, that job still falls to humans.

That means that the relevant question is “Will the jobs where humans have comparative advantage pay well and have good working conditions?” As we know from displacement due to globalization and increasing international trade, there is nothing that guarantees that humans displaced from jobs will be reemployed in new jobs that pay as well as their old jobs, or even pay well enough to maintain middle-class status.

What We Can Do

Though there is still much we don’t know about how this wave of automation will proceed, there are several areas of action we can identify now.

Education and training are at the top of the list. Human capital investment must be at the center of any strategy for producing skills that are complementary to technology. The current workforce — including the unemployed — needs opportunities for re-skilling and up-skilling, with businesses taking an active role both in determining the skills needed and in providing the skill training. Workers need opportunities for lifelong learning, and employers will be key. An extensive research literature documents the high returns to workers and firms from employer-based training. Workplace training helps bridge gaps between school learning and the application of these skills in the workplace and to specific occupations.

Schools will have to change too. Anticipating future skill needs and demands adds to the urgency of addressing the many challenges in K-12 and higher education, including achievement and opportunity gaps by race and socioeconomic status in K-12 schooling, and improving access, affordability, and success in post-secondary education. The education system must also do more to produce STEM workers and to ensure that workforce is diverse.

But education alone will not be sufficient. Policy makers should focus on cushioning the necessary transitions following job loss by strengthening the social safety net. In the U.S., this means strengthening unemployment insurance (ensuring benefit adequacy, including durations of eligibility), Medicaid, Supplemental Nutrition Assistance Program, and Transitional Assistance to Needy Families. A wage insurance program for all displaced workers will help encourage people to remain attached to the labor force.

In 1966 the final report of the National Commission on Technology, Automation and Economic Progress stated, “Constant displacement is the price of a dynamic economy. History suggests that it is a price worth paying. But the accompanying burdens and benefits should be distributed fairly, and this has not always been the case.” The Commission recommended responses that manage the overall health of the economy (managing and strengthening aggregate demand), promote educational opportunity, provide public employment, and secure transitional income maintenance. Fifty years later, these areas remain the basic road map for public policy response. The solutions, and any obstacles, are political, not economic or technical.

Powered by WPeMatico

Antiques

AdSense