Wednesday, July 23, 2014

NYT: Technology, Aided by Recession, Is Polarizing the Work World

JULY 22, 2014 New York Times
Claire Cain Miller

The American work force has been growing polarized for decades. On one end, there are highly skilled jobs like writing software or performing surgery, and on the other are service jobs like child care and cutting hair. The jobs in the middle, meanwhile, such as factory work, sales and bookkeeping, are shrinking — one of the reasons for the economy’s slow climb out of the recession.

Where did those jobs go? Part of the answer lies in Silicon Valley. It is no coincidence that many of those jobs entail the same repetitive tasks that computers, robots and other machines are uniquely suited to perform, fromrobots loading conveyor belts in factories to selling airline tickets.

A new working paper from the National Bureau of Economic Research shows how the recession accelerated the displacement of these midwage jobs. As technology now encroaches on jobs that people assumed would always belong to humans, it is useful to consider those most affected by the job displacement so far: the young, the less educated and men.

A lot of economic research has focused on the polarization of jobs, notably byDavid Autor of M.I.T. He differentiates between routine tasks that follow well-defined procedures — the kind of midwage jobs that computers have become so good at — and nonroutine ones that require flexibility, problem-solving and human interaction.

The trading floor of the New York Stock Exchange in July 2013. Stock trading is still done by humans, to some extent, but much more by computers than in past decades. CreditSpencer Platt/Getty Images
The new study, which analyzed data from the Current Population Survey from 1976 to 2012, illustrates that the recession had a disproportionately large effect on routine jobs, and greatly sped up their loss. That is probably because even if a new technology is cheaper and more efficient than a human laborer, bosses are unlikely to fire employees and replace them with computers when times are good. The recession, however, gave them a motive. And the people who lost those jobs are generally unable to find new ones, said Henry E. Siu, an associate professor at the University of British Columbia and an author of the study.

Young people and those with only a high school diploma are much more likely to be unemployed and replaced by a machine, he said. And to the authors’ surprise, men are more vulnerable than women.

“When you look at data, women who would otherwise be finding middle-paying routine jobs tend to be moving up the job ladder to these higher-paying brain jobs, whereas men are much more likely to just be moving from blue-collar jobs into not finding a job,” said Mr. Siu, who wrote the study with Guido Matias Cortes of the University of Manchester, Nir Jaimovich of Duke University and Christopher J. Nekarda of the Federal Reserve in Washington.

The changing demographics in the United States play a small role in the loss of midwage jobs, as do policies related to offshoring, unions and the minimum wage. But the study found that two-thirds of the decline in routine jobs is explained by a drop in the number of unemployed people who can get these jobs, and an increase in the number of people who had these jobs and lost them.

And the driver behind those shifts is technology.
“Over the very long run, technological progress is good for everybody, but over shorter time horizons, it’s not that everybody’s a winner,” Mr. Siu said. “Certain demographic groups like the young and less educated in another world would be doing fine, but in today’s world are not.”

The line between jobs that are considered routine and able to be done by a machine and those that require a human brain is a blurry one and becoming blurrier, said Erik Brynjolfsson and Andrew McAfee of M.I.T., authors of “The Second Machine Age.”

“There are examples up and down the spectrum,” Mr. Brynjolfsson said. “It’s a process of scientific discovery. It’s not like we know exactly which task will be next to automate.”

Already, machines are learning to do certain jobs that once seemed confined to humans, from elder care to wealth management to art. The question is what will happen if these jobs also disappear.

No comments:

Post a Comment