The Future of Work Is … Bleak

In considering the societal implications of his decades of work in computer science, Moshe Y. Vardi said he looked for projections on the impact of technology on the job markets, and found "this is a real popular topic. Every month another study comes out, and every study says something different." 

Vardi, a professor of computer science at Rice University (and also editor-in-chief of Communications), wondered why there were so many different predictions. He considered the well-known aphorism "predictions are difficult, especially about the future," but concluded "it’s really the other way around; predictions are easy, especially about the future. Everybody can make a prediction.

"Correct predictions are difficult." 

At the Work Awesome conference on the future of work held in New York City on June 23, Vardi was asked to discuss the outlook for jobs in 10 years. He considered a number of factors over the past 40 years that can provide clues as to how job markets will evolve.

While talk of off-shoring and outsourcing have led many to believe that manufacturing in the U.S., is gone, Vardi pointed out that the statistics show U.S. manufacturing volume in constant dollars "is at an all-time high now. The U.S. is a manufacturing giant. We are roughly comparable to China in volume, in output." 

However, he added, manufacturing employment peaked around 1980, "and has been going down, down, down, down, down, way down into the rabbit hole, and we are today at the level of maybe 1947." The reason? "Over the last 20 years, the productivity of our production employees has doubled, so guess what? They need about half the number of employees." 

Vardi said four macroeconomic factors must be considered to understand the ultimate impact these trends will have on jobs and, ultimately, on people. They are: 

• Productivity. "Productivity growth will drive economic growth." 
• Gross Domestic Product, "growth in GDP." 
• Jobs.
• Wages.

From the end of World War II to roughly 1980, Vardi said, "these four numbers marched together: productivity grew, the economy grew, we generated jobs, and the jobs paid better and better." He referred to this period as the Great Coupling, because all these factors grew together.

Around 1980, however, Vardi said, "we had the Great De-Coupling," in which productivity continued to grow, and GDP along with it, but jobs (private employment) and wages did not. "Real wages, adjusted to inflation, have now been stagnant for 40 years," he said, adding, "if you look at this, you can understand what is happening now in the election: millions of very, very unhappy people … incredibly unhappy with the system."

The reports we typically hear about unemployment being down are misleading, he explained, because they only measure the number of active job-seekers. A more meaningful measure, Vardi suggested, is Labor Force Participation, the percentage of the population that is in the workforce, and that has declined more or less steadily since 1990. As fewer people work, more apply for unemployment benefits that last up to two years; when that is exhausted, he said, many will have a doctor confirm they have some sort of disability and will file for long-term disability benefits. 

The impact of technology on the job market has been job polarization, which Vardi explained as "the jobs in the middle are being eaten by the poles."

He said jobs at the high end of the job spectrum, high-value knowledge worker-type positions, are too difficult and/or expensive to automate, while jobs at the low end "are not worthwhile to automate. Being a busboy in a restaurant, cleaning the tables, it’s going to actually be very, very difficult to automate. It will require a tremendous amount of situational awareness, dexterity, agility, grasping ability. It’s going to be very very expensive to automate, and how much do I pay a busboy? Minimum wage, so it won’t do."

People who have lost jobs in the current economy really have two choices, Vardi said: upscaling, taking a higher-value position with greater knowledge and skill requirements (and higher pay), or downscaling, taking a job that requires fewer skill and pays less. "This is the hollowing of the economy," as the jobs in the middle increasingly will be automated out of existence.

"The group in the middle is shrinking because we have growth in the high end and growth in the low end," Vardi said.

Vardi observed a "gold rush" going on in the autonomous vehicle sector, citing a McKinsey forecast that the sector would be worth about $2 trillion over the next decade. Automakers, technology companies, and others are diving into the market to get their share of that pie. He said there is general consensus that technical, legal, and liability issues will be resolved over the next decade, so we should eventually expect all driving to be automated. 

An immediate benefit to this will be the end of the 1.25 million deaths each year from car crashes, 95% of which are attributable to human error, Vardi said.

However, in the U.S. alone, there are about 4 million professional drivers, as well as many others for whom driving is a significant portion of their jobs. The result would be the elimination of about 10 million jobs in the U.S., Vardi said.

And, keeping in mind that "these are people with typically at best a high school education, they have the option of upscaling or downscaling. What do you think is going to happen? "

Nobody knows specifically what will happen in the future, Vardi said, but "if you look at working-class people, the last 35 years have been very harsh on them. This explains a lot of what’s happening right now politically in this country."

He added, "We can expect the next 35 years to be equally harsh for everybody. Lots of people will do very well; people who are in the business of automated cars will make a fortune, but 10 million people are going to lose their jobs, and people have said there will be new jobs for them. Upscaling? Less likely. Downscaling? More likely."

A conference on "Humans, Machines, and the Future of Work" will be held Dec. 5-6 at Rice University’s Bioscience Research Collaborative in Houston, TX. The conference will focus on issues created by the impact of information technology on labor markets over the next 25 years, addressing questions such as:

• What advances in artificial intelligence, robotics, and automation are expected over the next 25 years?
• What will be the impact of these advances on job creation, job destruction, and wages in the labor market?
• Which skills will be required for the job market of the future?
• Can education better prepare workers for that job market?
• What educational changes are needed?
• What economic and social policies are required to integrate people who are left out of future labor markets?
• How can we preserve and increase social mobility in such an environment?

Lawrence M. Fisher is Senior Editor/News for ACM magazines.