A 2012 research brief by Erik Brynjolfsson and Andrew McAfee at MIT renewed an old debate over the effect of new technologies on employment levels. They argued that, counter to the prevailing belief that new technologies and automation simply shift jobs into new sectors after a period of disruption, instead rapid improvements in technology over the past decades have left some workers completely behind, a trend that will continue to accelerate as computers capabilities expand. How does the public see “technological threats” to employment, and have their views changed since the days when robots first began replacing line workers in factories?
From the Roper Center for Public Opinion Research archive:
More machinery, fewer jobs?
The U.S. public has been asked about the effect of new technologies and automation on jobs since the early fifties, with pollsters showing particular interest in the issue during the high unemployment in the 1980s. Trends through 1999 show that the country is often evenly split on whether greater use of workplace technology increases or decreases employment, with variation in responses both over time and by question wording. One of the most negative responses was a 1983 poll that found 56% disagreed that computers and factory automation will create more jobs than they will eliminate, while only 39% agreed. In contrast, 34% in 1989 believed that scientific and technological changes cause unemployment because people’s jobs are replaced by machines, while 45% said scientific and technological changes increase the total number of jobs over the long run.
Current public opinion leans slightly towards a positive assessment of the effect of technology. A majority in a 2015 CNBC poll said that technology has more benefits than drawbacks to the economy, because it provides services and products to consumers at lower prices, though a substantial minority say that the drawbacks of replacing workers outweigh the benefits. A 2012 Pew poll found that 40% of Americans believe new technologies have increased the number of jobs in the U.S., while 32% think they have decreased the number, and 21% say they’ve made no difference.
The most recent poll on this issue points to potential shifts in public opinion with future technologies. A 2015 Monmouth poll about artificial intelligence found 72% of the public believe having machines with the ability to think for themselves would hurt jobs and the economy, among the most negative responses in the history of polling on the effect of technology on employment.
Specifically, what about today’s unemployment?
Questions that ask specifically about whether technology is to blame for current unemployment or underemployment have also found the public divided, though in recent years perhaps more inclined to lay more responsibility at the feet of automation. In 2013, a question with four-way response categories found 69% of the public put a lot or some of the blame for good paying jobs being hard to find on technology replacing workers.
Overall importance of high technology
However, despite ongoing concern about the effect of technology on employment in particular, Americans have consistently been positive about the effect of technology on the economy overall, and in fact have seen technological innovation and development as vital to the country’s economic interests. For example, a 1983 Cambridge Reports/Research International poll found that 48% strongly agreed and 40% somewhat agreed that the future prosperity of the United States depended on more and better technology. In a 1996 Washington Post/Kaiser Family Foundation/Harvard poll, 70% of Americans said the increased use of technology in the workplace was good for the economy. In a 2010 Allstate/National Journal poll, 79% said that information technology was extremely or very important to creating economic growth in the U.S.
So do the unemployed just lack skills?
The public’s generally positive views about the effects of technology on the economy do not necessarily conflict with their willingness to blame for current unemployment on workplace automation. The public may share the views of those economists who argue that technology is disruptive in the short-term, but creates jobs in the long run. In this view, unemployment is caused by displaced workers lacking the skills for the newly available jobs, a situation that rights itself with time, training, and education. However, the public does not appear to see this disconnect between skills and jobs as the underlying cause of unemployment.
In a 1982 poll, 51% believed lack of jobs was the main cause of unemployment, only 21% believed it was lack of skills, and 26% that people just didn’t want to work. In 2010, an even greater proportion of the public believed there were no jobs available for the unemployed. Seventy-nine percent of the public said the main cause of unemployment was a lack of jobs, only 12% thought it was people lacking skills, and just 7% that people just don’t want to work.
A robot? Not my job!
Despite concerns about the effects of technology on jobs overall, very few people today are concerned about losing their own jobs to technology. In the 1980s, a substantial minority were concerned. In a 1984 Hearst Corporation poll, 29% said they were very concerned about computers or robotics threatening their job in the future, and 17% said they were somewhat concerned. A 1993 Gallup/CNN/USA Today poll, however, found only 5% thought technology could eliminate their jobs completely in 5 or 10 years. Over half (56%) believed technology could change the nature of their job, and 38% expected no effect.
Only 13% in a 2015 poll were concerned that their own job could be replaced with technology, at least in the near future. But a 2014 poll of the unemployed found that 30% said technology replacing jobs was at least a minor cause of their unemployment.
As computers take on more and more complex tasks, public opinion on this issue will no doubt continue to be monitored – but by whom? Pollsters, take heed: this NPR job automation assessment tool gives survey researchers a 23% chance of being replaced by machines in the next 20 years.
Roper data collections cited include Opinion Research Corporation, National Science Foundation, Cambridge Reports, CNBC, Pew Research Center, Monmouth University, NYT, WP/Miller Center, Cambridge Reports/Research International, WP/Kaiser Family Foundation/Harvard University, CBS News/NYT, Rutgers University, Allstate/National Journal, Hearst Corporation, Gallup/CNN/USA Today, Kaiser Family Foundation/NYT/CBS News.