The American labor market was once built on routine work -- jobs in factories and offices that required human bodies to perform repetitive tasks, whether it was stamping widgets or making phone calls.
Factory jobs are long gone, for the most part, but new research finds that in the last recession, many routine white-collar jobs were also automated, and they are not coming back.
A study from Duke University and the University of British Columbia found that routine physical jobs like forklift operators and welders were automated, as were "cognitive" white-collar tasks.
"Think bank tellers being replaced by ATMs, and secretarial work being replaced by personal computers," said Henry Siu of the University of British Columbia, the report's co-author.
"Historically these occupations rebounded, said Siu. "But this time, they're not coming back." This helps explain why jobs aren't snapping back after the recession, he says, and we may be ushering in a new labor landscape.
Can a robot do your job?
A machine operator has a routine job, one that has well-defined instructions or procedures that a computer can reproduce, but so do bookkeepers, filing clerks, and paralegals.
"If the bulk of what you do could be summarized in a flow chart, a robot could take your job," says Siu.
David Autor, an economist at MIT and a leading scholar studying job automation, calculates that middle-skill occupations made up 60 percent of all jobs in 1979. By 1970, about 35 percent of the American workforce had jobs that primarily consisted of routine tasks, and since then, that number steadily declined until the recession in 2008, when it dipped dramatically to just over 25 percent. It hasn't budged since.
According to Siu and his co-author, all job growth since 2001 has come from non-routine work. Many of the jobs that have bounced back since the recession involve using discretion or solving unanticipated problems. On the low end, this includes low-paying service jobs that require "physical flexibility," jobs like janitors, dishwashers, home health aids and fast food work.
The high-end jobs that have bounced back involve creative problem-solving skills and "cognitive flexibility," including jobs like engineering, computer programming, and public relations.
What has been lost are largely middle-class jobs. "The loss of routine jobs in recent recessions has given rise to jobless recoveries," said Siu. "Employment has struggled to rebound for middle-wage jobs."
Job polarization
The loss of middle-wage jobs contributes to the "hollowing out of the middle" or "sagging middle," that economists have wrung their hands over since American factory workers started losing jobs in the 70s.
Economists call this split "job polarization," which is what happens when job growth occurs at the bottom (the rise in low-wage service jobs) and at the very top (high-paying engineering and programming jobs). It can lead to mobility and income gaps.
Autor's research shows that high-skilled workers who are essential in a tech economy have growing opportunities, while low-end service jobs make up a disproportionately wide base of dead-end jobs.
"This is not an overall improvement in job quality," says Autor's report. Low-skill jobs mean workers have little negotiating power. "These are not, in general, attractive jobs."
According to the Department of Labor, the top employment opportunity in the U.S. is retail sales, which currently has about 200,000 openings.
In the second spot is fast food workers, followed by cashiers and waiter and waitress jobs. None of these offer a family friendly wage until the fifth place job -- registered nurse.
Future earnings
The proliferation of low-wage jobs raises the question of what kind of existence they provide compared to three decades ago when a manufacturing job could pay for a house and college.
Now, two family members working minimum jobs will often hover around the poverty level, according to research from the Economic Policy Institute.
The Department of Labor also ranks the fastest-growing occupations, and "industrial-organizational psychologist" tops the list, paying $80,000 a year. The demand is projected to double by 2022, although with only 1,600 openings available annually.
Siu still has hope for "skilled" middle jobs, for example, he visited a tool manufacturing plant in North Carolina where people left on the job are the ones that understand the computers and know how to calibrate the machines in order to get the right seal, or change the angle a little from the valve they were making the week before."
Those skills dont require a PhD or even a BA in physics, you can learn them in a tech school," he says. As for degrees in medical record transcription or paralegal work? "Those will lose value," he says.
He points out that it's hard to say what will happen with machine work -- computer software can already write a formulaic newspaper story about stock market trends, and legal discovery work that used to require a team of lawyers and assistants can now be performed using optical character recognition and algorithms. Even his own job -- university professor -- is being simplified and condensed using technology in the form of Massive Open Online Courses, or MOOCS.
Siu is hopeful that an economy that relies on creativity and problem-solving will create jobs that don't have names yet. "We don't know yet what jobs will look like in the future."
Factory jobs are long gone, for the most part, but new research finds that in the last recession, many routine white-collar jobs were also automated, and they are not coming back.
A study from Duke University and the University of British Columbia found that routine physical jobs like forklift operators and welders were automated, as were "cognitive" white-collar tasks.
"Think bank tellers being replaced by ATMs, and secretarial work being replaced by personal computers," said Henry Siu of the University of British Columbia, the report's co-author.
"Historically these occupations rebounded, said Siu. "But this time, they're not coming back." This helps explain why jobs aren't snapping back after the recession, he says, and we may be ushering in a new labor landscape.
Can a robot do your job?
A machine operator has a routine job, one that has well-defined instructions or procedures that a computer can reproduce, but so do bookkeepers, filing clerks, and paralegals.
"If the bulk of what you do could be summarized in a flow chart, a robot could take your job," says Siu.
David Autor, an economist at MIT and a leading scholar studying job automation, calculates that middle-skill occupations made up 60 percent of all jobs in 1979. By 1970, about 35 percent of the American workforce had jobs that primarily consisted of routine tasks, and since then, that number steadily declined until the recession in 2008, when it dipped dramatically to just over 25 percent. It hasn't budged since.
According to Siu and his co-author, all job growth since 2001 has come from non-routine work. Many of the jobs that have bounced back since the recession involve using discretion or solving unanticipated problems. On the low end, this includes low-paying service jobs that require "physical flexibility," jobs like janitors, dishwashers, home health aids and fast food work.
The high-end jobs that have bounced back involve creative problem-solving skills and "cognitive flexibility," including jobs like engineering, computer programming, and public relations.
What has been lost are largely middle-class jobs. "The loss of routine jobs in recent recessions has given rise to jobless recoveries," said Siu. "Employment has struggled to rebound for middle-wage jobs."
Job polarization
The loss of middle-wage jobs contributes to the "hollowing out of the middle" or "sagging middle," that economists have wrung their hands over since American factory workers started losing jobs in the 70s.
Economists call this split "job polarization," which is what happens when job growth occurs at the bottom (the rise in low-wage service jobs) and at the very top (high-paying engineering and programming jobs). It can lead to mobility and income gaps.
Autor's research shows that high-skilled workers who are essential in a tech economy have growing opportunities, while low-end service jobs make up a disproportionately wide base of dead-end jobs.
"This is not an overall improvement in job quality," says Autor's report. Low-skill jobs mean workers have little negotiating power. "These are not, in general, attractive jobs."
According to the Department of Labor, the top employment opportunity in the U.S. is retail sales, which currently has about 200,000 openings.
In the second spot is fast food workers, followed by cashiers and waiter and waitress jobs. None of these offer a family friendly wage until the fifth place job -- registered nurse.
Future earnings
The proliferation of low-wage jobs raises the question of what kind of existence they provide compared to three decades ago when a manufacturing job could pay for a house and college.
Now, two family members working minimum jobs will often hover around the poverty level, according to research from the Economic Policy Institute.
The Department of Labor also ranks the fastest-growing occupations, and "industrial-organizational psychologist" tops the list, paying $80,000 a year. The demand is projected to double by 2022, although with only 1,600 openings available annually.
Siu still has hope for "skilled" middle jobs, for example, he visited a tool manufacturing plant in North Carolina where people left on the job are the ones that understand the computers and know how to calibrate the machines in order to get the right seal, or change the angle a little from the valve they were making the week before."
Those skills dont require a PhD or even a BA in physics, you can learn them in a tech school," he says. As for degrees in medical record transcription or paralegal work? "Those will lose value," he says.
He points out that it's hard to say what will happen with machine work -- computer software can already write a formulaic newspaper story about stock market trends, and legal discovery work that used to require a team of lawyers and assistants can now be performed using optical character recognition and algorithms. Even his own job -- university professor -- is being simplified and condensed using technology in the form of Massive Open Online Courses, or MOOCS.
Siu is hopeful that an economy that relies on creativity and problem-solving will create jobs that don't have names yet. "We don't know yet what jobs will look like in the future."