Globalization and trade have transformed the American economy. But increasingly, the competition for jobs comes from inside our own borders, with automation, robots and artificial intelligence rapidly moving into the workforce. What can we expect and what can we do about it? Join us in our search for robot-proof jobs.
Take the Quiz
Hear the podcast
Automation At Work
The McKinsey Global Institute analyzed the work activities of more than 800 occupations in the U.S. to determine what percentage of a job could be automated using current technology. It turns out, a small fraction of jobs are either entirely automatable or entirely robot-proof.
0% Automatable
Ambulance Drivers and Attendants, Except Emergency Medical Technicians
Animal Scientists
Animal Trainers
Astronomers
Athletes and Sports Competitors
Clergy
Dancers
Directors, Religious Activities and Education
Historians
Mathematical Technicians
Models
Music Directors and Composers
Religious Workers, All Other
Roof Bolters, Mining
100% Automatable
Aircraft Cargo Handling Supervisors
Dredge Operators
Foundry Mold and Coremakers
Graders and Sorters, Agricultural Products
Logging Equipment Operators
Machine Feeders and Offbearers
Medical Appliance Technicians
Motion Picture Projectionists
Ophthalmic Laboratory Technicians
Packaging and Filling Machine Operators and Tenders
Plasterers and Stucco Masons
Slaughterers and Meat Packers
The vast majority of jobs consist of some portion of tasks that can be automated and some portion of tasks that can’t. Think you know which jobs are most susceptible to automation? Take our quiz below and find out.
Which of these three jobs is most automatable?
Tree-Pruning Specialist
CEO
Real Estate Broker
Make a selection
Correct Answer: Real Estate Broker
Pay doesn’t directly correspond to automatability. Plenty of lower-paid jobs, like tree pruning, which requires physical dexterity and maneuvering in unstructured environments, cannot be easily or cheaply replaced by machines.
Which of these three major occupations is most automatable?
Nurse
Freight-Mover
Waiter
Make a selection
Correct Answer: Waiter
Some of the biggest job categories in America — clerks, drivers and food service workers — are highly automatable. Jobs that require interpersonal skills, like nursing and teaching, tend be to less so.
Combined Food Preparation and Serving Workers, Including Fast Food
86.7% Automatable
3,022,890 U.S. Workers
$9.45 per hour
Stock Clerks and Order Fillers
86.4% Automatable
1,800,430 U.S. Workers
$13.00 per hour
Bookkeeping, Accounting, and Auditing Clerks
85.6% Automatable
1,586,390 U.S. Workers
$17.57 per hour
Cooks, Restaurant
84.0% Automatable
1,057,540 U.S. Workers
$11.02 per hour
Heavy and Tractor-Trailer Truck Drivers
81.4% Automatable
1,585,310 U.S. Workers
$19.31 per hour
Waiters and Waitresses
76.9% Automatable
2,403,940 U.S. Workers
$9.95 per hour
Office Clerks, General
61.5% Automatable
2,832,020 U.S. Workers
$14.37 per hour
Secretaries and Administrative Assistants, Except Legal, Medical, and Executive
53.6% Automatable
2,159,010 U.S. Workers
$15.91 per hour
Cashiers
48.6% Automatable
3,343,380 U.S. Workers
$10.14 per hour
Retail Salespersons
46.8% Automatable
4,485,080 U.S. Workers
$12.70 per hour
Nursing Assistants
44.2% Automatable
1,427,730 U.S. Workers
$12.33 per hour
Security Guards
39.4% Automatable
1,066,730 U.S. Workers
$13.92 per hour
First-Line Supervisors of Retail Sales Workers
32.7% Automatable
1,213,530 U.S. Workers
$20.08 per hour
Customer Service Representatives
29.3% Automatable
2,389,590 U.S. Workers
$15.88 per hour
Registered Nurses
28.9% Automatable
2,661,850 U.S. Workers
$30.98 per hour
Maintenance and Repair Workers, General
25.2% Automatable
1,249,080 U.S. Workers
$18.25 per hour
Personal Care Aides
23.6% Automatable
1,134,940 U.S. Workers
$10.26 per hour
General and Operations Managers
22.8% Automatable
1,973,700 U.S. Workers
$51.45 per hour
Janitors and Cleaners, Except Maids and Housekeeping Cleaners
22.4% Automatable
2,101,820 U.S. Workers
$12.10 per hour
First-Line Supervisors of Office and Administrative Support Workers
22.1% Automatable
1,366,530 U.S. Workers
$24.94 per hour
Sales Representatives, Wholesale and Manufacturing, Except Technical and Scientific Products
21.1% Automatable
1,403,200 U.S. Workers
$29.40 per hour
Teacher Assistants
19.9% Automatable
1,190,630 U.S. Workers
$11.88 per hour
Elementary School Teachers, Except Special Education
15.4% Automatable
1,344,080 U.S. Workers
$25.55 per hour
Accountants and Auditors
12.2% Automatable
1,168,320 U.S. Workers
$32.24 per hour
Laborers and Freight, Stock, and Material Movers, Hand
6.7%
2,284,660 U.S. Workers
$13.21 per hour
Which of these three high-growth jobs is most automatable?
Web Developer
Personal Financial Advisor
Massage Therapist
Make a selection
Correct Answer: Web Developer
Not all high-growth or high-tech jobs require human know-how. Repetitive work done by web developers or technicians can be done efficiently and inexpensively by computers.
Phlebotomists
89.40% Automatable
108,060 U.S. Workers
24.9% projected growth (2014)
$14.67 per hour
Medical Secretaries
57.48% Automatable
512,560 U.S. Workers
20.5% projected growth (2014)
$15.36 per hour
Medical Assistants
54.36% Automatable
571,640 U.S. Workers
23.5% projected growth (2014)
$14.47 per hour
Health Technologists and Technicians, All Other
52.89% Automatable
88,910 U.S. Workers
23.1% projected growth (2014)
$21.57 per hour
Diagnostic Medical Sonographers
48.06% Automatable
57,710 U.S. Workers
26.4% projected growth (2014)
$31.94 per hour
Cardiovascular Technologists and Technicians
45.33% Automatable
50,970 U.S. Workers
22.2% projected growth (2014)
$26.23 per hour
Web Developers
42.58% Automatable
112,810 U.S. Workers
26.6% projected growth (2014)
$29.19 per hour
Speech-Language Pathologists
40.44% Automatable
125,050 U.S. Workers
21.3% projected growth (2014)
$34.41 per hour
Physical Therapist Assistants
37.11% Automatable
72,540 U.S. Workers
40.6% projected growth (2014)
$25.43 per hour
Emergency Medical Technicians and Paramedics
35.48% Automatable
237,520 U.S. Workers
24.2% projected growth (2014)
$16.42 per hour
Physical Therapists
31.84% Automatable
195,580 U.S. Workers
34% projected growth (2014)
$39.58 per hour
Computer Systems Analysts
29.19% Automatable
507,120 U.S. Workers
20.9% projected growth (2014)
$37.66 per hour
Physician Assistants
26.78% Automatable
88,020 U.S. Workers
30.4% projected growth (2014)
$46.40 per hour
Mental Health Counselors
26.13% Automatable
115,560 U.S. Workers
19.6% projected growth (2014)
$21.29 per hour
Occupational Therapists
25.20% Automatable
108,390 U.S. Workers
26.5% projected growth (2014)
$37.37 per hour
Substance Abuse and Behavioral Disorder Counselors
Chart 1 represents all U.S. occupations, as captured in the McKinsey Global Institute’s automation report.
Chart 2 represents the top 25 largest U.S. occupations as measured by total employees.
Chart 3 represents the top 25 fastest-growing U.S. occupations as measured by percentage growth, excluding occupations with less than 50,000 people employed.
Behind the Numbers
A robot has just moved into the cubicle next to you, and you’re wondering whether it will soon take your place. Experts say humans are better at jobs that require empathy, creativity or physical dexterity than our robot competitors. For instance, in the graphs above, empathetic nurses, creative CEOs and dextrous tree pruners are all on the low end when it comes to automatability, as measured by McKinsey. These core characteristics may determine how resistant a job is to automation more than pay, industry sector or skill level.
Are you the radiologist or the taxi driver?
Automatability is just a number. While it's useful to know how much of your job could be done by machines, it may be more useful to understand what piece of your job is robotizable. Is it the part that takes skill or the part that is rote?
For example, radiology and taxi driving — two jobs transformed by technology. Computers are starting to read medical images just as well as radiologists. But radiologists add value in other ways machines can’t: by communicating with patients and integrating medical information into diagnoses and treatment plans. This leaves radiologists with a skilled portion of work that cannot be automated, giving them a better shot at keeping their exclusive high-paying jobs.
On the other hand, we have the taxi driver whose job consists of two basic parts: navigating and driving. Years ago, taxi drivers had to study and memorize entire city maps, a specialized skill that allowed only the qualified few to make money. With the advent of GPS and smartphone apps, the navigation aspect has gone digital, allowing more people to become drivers, and in turn, drive down wages. Contrast that with the other piece of their job — pushing the gas pedal, hitting the brakes and turning the wheel — which doesn’t take much expertise. Most anyone can drive — hey, even computers are learning.
And then there’s demand
Let’s go back to our radiologist. One might assume new imaging machines will mean fewer radiologists, because a single radiologist can get more work done assisted by technology. This is where demand comes in. As technology allows radiology services to become faster and cheaper, more people might be able to access and pay for them. In other words, the demand for radiologists could go up and offset any lost work brought on by technology.
A similar demand dynamic took place with bank tellers from the 1970s to the early 2000s. ATMs became widespread, which could have spelled trouble for human bank tellers. Instead, the number of tellers grew. Because of ATMs, banks could afford to open new branches staffed by fewer tellers. But overall, more tellers were hired to serve more customers at more locations. If demand had stayed flat however, the ranks of bank tellers would have declined.
Of course there are other factors that determine whether a human or machine will be hired for the job. One is relative cost, meaning how much the boss has to pay a human versus a machine to get the work done. Even if a robot can feasibly do a task, it still has to make economic sense to install and use it. Another is social acceptability — whether society is willing to automate a job. For example, it may be a long time before we are comfortable with robo-judges, robo-legislators or robo-priests. Maybe. Maybe not.
Economic theory happily acknowledges that technology eradicates lots of jobs. Take, for example, the 98 percent of all farming jobs that were annihilated by machines over the last two centuries. We are taught that’s OK, because technology created more productive, higher-paying, less back-breaking and more interesting jobs. The farm workers got better work in factories and American standards of living went up. That’s the classic promise of technology, but does it hold up today?
Armed with this question and data from McKinsey, Marketplace Morning Report host David Brancaccio and producer Katie Long drove across the Midwest — a region that has seen waves of competition from both globalization and automation — to search for jobs that technology can’t easily replace. They found a cop who says you’re better off getting pulled over by him than a robot, a composer who gets inspiration from a sofa and a woman who makes her living by continually evolving.
Along the way, they looked at how to train people for these robot-proof jobs in the near term. Plus, what happens in the long term when technological advancement goes from incremental to radical, putting most of us out of work?
Click on the player above to hear their journey or scroll through the photos below to meet the voices from their reporting.
Forward
Back
James Manyika, McKinsey Global Institute
-Courtesy of McKinsey
Erik Brynjolfsson, Massachusetts Institute of Technology
