Automation and a Changing Economy
April 2020
Alongside Automation, U.S. Workforce Has Grown Over Time
0
20
40
60
80
100
120
140
160
180
Emp
loym
en
t (i
n m
illio
ns)
Civilian Employment Level
2007iPhone Released
1990World Wide Web
Invented
1977Personal Computer
Introduced
Source: analysis of Current Population Survey and Employment Projections data, U.S. Bureau of Labor Statistics.
151K 524K
2.9 MILLION
12.2 MILLION
TOTAL US JOBS CREATED BY PERSONAL COMPUTERS
Source: IPUMS; Moody’s; IMPLAN; US Bureau of Labor Statistics; FRED; McKinsey Global Institute analysis
15.8 MILLIONNET JOBS CREATED
DIRECT (Computer equipment
manufacturing, 1970–2015)
▲ Assorted managers and administrators 31
▲ Computer software developers (in-industry equipment)
27
▲ Computer scientists 18
▼ Office machine manufacturers (typewriters)
-61
INDIRECT (Computer suppliers, 1970–2015)
▲Managers 42
▲ Semiconductor manufacturing occupations
31
▲ Semiconductor manufacturing occupations
26
▼ Typewriter indirect occupations -79
ENABLED (Computer software and service
industries, 1970–2015)
▲ Software developers (software and apps)
768
▲ Computer scientists 686
▲Managers 416
▼ Typewriter repair -32
UTILIZERS (Computer-utilizing industries,
1980–2015)
▲ Customer service reps 3,205
▲ Computer scientists(not in computer industry)
1,873
▲ Stock and inventory clerks 1,517
▼ Bookkeepers and auditing clerks -881
▼ Secretaries -823
▼ Typists -562
REPRESENTATIVE OCCUPATIONS (in thousands)
Automation Has Created Jobs: Personal Computers
Automation & the Manufacturing Industry
For every additional industrial robot introduced into a local labor market, on average, 6.2
workers in that labor market lost their jobs.
These losses include both direct factory job losses as well as indirect losses elsewhere,
particularly in the construction, business services, wholesale, service, and retail industries.
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
All Employees: Manufacturing/All Employees: Total Nonfarm Payrolls
1943-2018
Source: U.S. Bureau of Labor Statistics. Retrieved from FRED, Federal Reserve Bank of St. Louis. Source: Acemoglu and Restrepo. 2017. “Robots and Jobs.” National Bureau of Economic Research. http://www.nber.org/papers/w23285.
Source: Katz and Margo. 2014. “Technical Change and the Relative Demand for Skilled Labor.” In Human Capital in History. https://www.nber.org/chapters/c12888.pdf.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
1920 1930 1940 1950 1960 1970 1980 1990 2000 2010
Distribution of Jobs Within White Collar Work, 1930-2010
Professional- technical Manager Clerical/ sales
Automation Has Eliminated Jobs: Routine Office Jobs Have Declined
Automation Has Changed Jobs: Bank Tellers & Travel Agents
Despite the introduction of 400,000+ ATMs in the U.S., the
number of bank tellers has increased over time
With the introduction of ATMs, the responsibilities of bank tellers have changed: more
customer-facing, new responsibilities
0
100000
200000
300000
400000
500000
600000
1970 1975 1980 1985 1990 1995 2000 2005 2010 2015
Bank Tellers vs. Travel Agents Over Time
Travel Agents Bank Tellers
Source: U.S. Census
Janitors: 2.2 million in the workforce, digital skills
requirements have increased by45% over the last 10 years
Personal care aides: digital skills requirements have increased by
189% over the last 10 years
56%
30%
40%
48%
5%
23%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
2002 2016
Employment by Levels of Job Digitalization
Low Medium High
Source: Brookings Institution
Automation is Changing Skills: Increasing Digital Skills Needed
Jobs are increasingly higher-skill professional, managerial, and
technical jobs
In 2019, American manufacturers were on track
to employ more college graduates than workers with a high-school education or less.
More than 40% of manufacturing workers have a college degree,
up from 22% in 1991.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
1920 1930 1940 1950 1960 1970 1980 1990 2000 2010
Manufacturing Industry Employment by Skill Levels
Lower Skill (Operatives/Laborers/Service) Middle Skill (Clerical/Sales/Craft)
Higher Skill (Prof/Tech/Manager)
Source: Katz and Margo. 2014. “Technical Change and the Relative Demand for Skilled Labor.” In Human Capital in History.https://www.nber.org/chapters/c12888.pdf.
Automation is Changing Skills: Manufacturing Industry
Source: Author’s analysis of Employment Level data, U.S. Bureau of Labor Statistics. Retrieved from Federal Reserve Bank of St. Louis. https://fred.stlouisfed.org/graph/?g=43U0#0.
0
10,000
20,000
30,000
40,000
50,000
60,000
70,000
1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017
Employment Levels of Routine and Nonroutine Occupations, 1983-2017(in thousands)
Nonroutine Cognitive
Routine Cognitive
Routine Manual
Nonroutine Manual
Automation Contributing to a Divided Labor Market: Nonroutine Jobs Growing, Routine Jobs Remain Flat
Source: OECD. 2017. “Job polarization by country.” In OECD Employment Outlook 2017. https://www.oecd-library.org/employment/oecd-employment-outlook-2017/job-polarization-by-country_empl_outlook-2017-graph39-en.
-20
-15
-10
-5
0
5
10
15
20
Percent Change in Share of Total Employment, OECD Countries, 1995-2015
Low Wage Middle Wage High Wage
Automation Contributing to a Divided Labor Market: Middle-Wage Jobs in Decline Across OECD Countries
Source: Katz and Margo. 2014. “Technical Change and the Relative Demand for Skilled labor.” In Human Capital in History. https://www.nber.org/chapters/c12888.pdf.
0
10
20
30
40
50
60
70
1920 1930 1940 1950 1960 1970 1980 1990 2000 2010
Occupational Distribution of U.S. Employment, 1920-2010Middle- and High-Skilled White Collar Work Grew Significantly as a Share of All U.S. Jobs
White Collar
Skilled Blue Collar (Craft)
Operative/Laborer/Service
Agricultural Occupations
Automation Contributing to a Divided Labor Market
Operative/Laborer/Service
Skilled Blue Collar (Craft)
Potential for Increased Automation: Undergraduate Enrollment Trends
Source: AI Index, University provided data
2010 2012 2014 2016 2018
% O
f u
nd
ergr
adu
ate
s
PERCENT OF UNDERGRADUATES ENROLLED IN INTRO TO AI
(2010-2018)
Berkeley Stanford UIUC UW
13%
10%
8%
5%
3%
0%
2010 2012 2014 2016 2018
% O
f u
nd
ergr
adu
ate
s
PERCENT OF UNDERGRADUATES ENROLLED IN INTRO TO ML
(2010-2018)
Berkeley Stanford UIUC UW
13%
10%
8%
5%
3%
0%
Source: AI Index, University provided data
Potential for Increased Automation: AI Image Detection Outperforming Humans
Source: ImageNet 2010-2018; see appendix, AI Index
2010 2012 2014 2016
Acc
ura
cy
ImageNet Competition Test Set Accuracy Human Performance
100%
90%
80%
70%
Projections of Automation Disruption
Projections about what jobs are at risk and how many workers could become displaced:
» McKinsey Global Institute • Up to 32 percent of workers may need to transition to entirely different occupations
by 2030 as a result of automation
» Brookings Institution• Over the next few decades, approximately 25 percent of U.S. employment will have
experienced high exposure to automation (with greater than 70 percent of current task content at risk of substitution).
THE ASPEN INSTITUTE | FUTURE OF WORK INITIATIVE 13
Who is Most at Risk of Automation?» Low-wage, routine jobs
• The jobs that appear most vulnerable are those that involve routine cognitive and manual tasks: repetitive, predictable activities like operating machinery, preparing fast food, and collecting and processing data.
» Women• Studies of jobs in Phoenix and Indianapolis show that women in certain markets may be more likely than men to be employed
in jobs at highest risk of automation, such as cashier, office clerk, secretarial and administrative positions.
» People of color• A recent study by the Joint Center for Political and Economic Studies found that larger shares of Latino (31 percent) and
African American (27 percent) workers are employed in the 30 jobs identified as being at highest risk of automation, particularly cooking, cashiering, and food preparation, than are White (24 percent) or Asian American (20 percent) workers.
» Disabled• A recent study by Source America shows that workers with disabilities are, on average, more likely to be employed in jobs at
high risk of automation, such as in grounds and building maintenance, food service, retail, warehousing, and administrative work..
» Young people• A Brookings Institution study found that workers between the ages of 16 and 24 face a notably high average automation
exposure due to their dramatic overrepresentation in automatable jobs, such as those in the food services..
» Rural vs. urban• Automation risk varies across U.S. regions, states, and cities. For example, the “American Heartland” states, which have a
longstanding and continued specialization in the manufacturing and agricultural industries, are expected to face heightened automation risk.
THE ASPEN INSTITUTE | FUTURE OF WORK INITIATIVE 14
Who is Most at Risk of Automation?
44%
19%
8%6%
1% 0%0%
10%
20%
30%
40%
50%
Less than High School High School Degree orEquivalent
Trade School Certificate Associates Degree Bachelors Degree Graduate Degree
Shar
e o
f Pe
op
le w
ith
Hig
h A
uto
mat
ibili
tyb
y Ed
uca
tio
n
Percent of Workers in Highly Automatable Jobs by Education Level
Calculations based on the Survey of Adult Skills (PIAAC) 2012.Source: Arntz, Gregory, and Ulrich Zierahn. 2016. “The Risk of Automation for Jobs in OECD Countries.” OECD Social, Employment and Migration Working Papers. https://www.oecd-ilibrary.org/social-issues-migration-health/the-risk-of-automation-for-jobs-in-oecd-countries_5jlz9h56dvq7-en
Policy & Employer Response: Limited Supports for Workers
19.4
16.7
12.411.2
13.111.7
8.6 8.4
0
5
10
15
20
25
1996 2001 2004 2008
Percent of Workers Receiving Job-Related Training
Employer-Paid Training On-the-Job Training
0.01%
0.01%
0.01%
0.03%
0.07%
0.12%
0.17%
0.20%
0.37%
0.60%
0.0% 0.1% 0.2% 0.3% 0.4% 0.5% 0.6% 0.7%
Australia
United Kingdom
Japan
United States
Canada
Spain
Italy
Germany
France
Denmark
% of GDP
Total Public Spending on Worker Training, 2015
Note: Fraction of workers ages 18-65 receiving training of any duration last year.Source: Census Bureau, Survey of Income and Program Participation (Employment and Training Topical Module); CEA calculation Source: McKinsey Global Institute analysis of OECD data, December 2017
DIGNIFIED RETIREMENT
STABLE CAREER
AFFORDABLE EDUCATION
RETIREMENT?• Social Security Trust Fund exhausts by 2035 • 401(K)
CAREER• Multiple jobs over career• Rise of on-demand economy
and non-traditional work• Increased financial insecurity
EDUCATION• Student debt• Value of 4-year vs. 2-year• Alternative pathways
Social SecurityMedicareEmployer Pension
Universal high school movementHigher Education ActGI Bill
Employer benefits (health care, retirement) Government protections (minimum wage, collective bargaining, overtime, anti-discrimination)
Policy Development in the 20th Century
Policy Recommendations for the Future of Work
» Expand apprenticeship programs
» Promote worker voice
» Create Worker Training Tax Credit
Encourage Employers to Lead Human-Centric
Approach to Automation
1
» Establish lifelonglearning & trainingaccounts
» Improve data ontraining outcomes
» Promote job quality
Enable Workers to Access Skills Training, Good Jobs, and New
Economic Opportunities
2
» Modernize Unemployment Insurance
» Provide wage insurance to older workers
» Develop place-based policies
Help People and Communities Recover from Displacements
3
» Develop better data on local and regional economies
» Create new metrics fortracking technologicalprogress & automation
» Improve occupationalprojections
Understand the Impact of Automation on the
Workforce
4
Encourage Employers to Lead a Human-Centric Approach to Automation
THE ASPEN INSTITUTE | FUTURE OF WORK INITIATIVE 19
» Promote employer engagement and investment through a worker training tax credit, expansion of apprenticeships, and new sector and regional workforce partnerships.
Automation changes workforce skill needs, yet employer investment in workforce development has declined.
Employers are making decisions about adopting automation, but may not take into account potential impacts on workers and communities.
» Encourage employers to adopt a multi-stakeholder approach to automation decisions by promoting new forms of worker voice and ownership and developing proactive strategies to identify and address impacts in advance.
» CHALLENGES » SOLUTIONS
Enable Workers to Access Skills Training, Good Jobs, and New Economic Opportunities
THE ASPEN INSTITUTE | FUTURE OF WORK INITIATIVE 20
» Improve access to effective and affordable skills training and develop a culture and system of lifelong learning.
The labor market is constantly evolving, with automation contributing to changing jobs and skill needs, but supports for worker training and adult education are limited.
Many workers struggle to make ends meet, and while automation has the potential to improve job quality, it also may lead to more low-wage jobs and greater economic insecurity.
» Increase wage subsidies and the minimum wage, while creating more economic opportunities by improving labor market flexibility and promoting entrepreneurship.
» CHALLENGES » SOLUTIONS
Help People and Communities Recover from Displacements
THE ASPEN INSTITUTE | FUTURE OF WORK INITIATIVE 21
» Strengthen supports for unemployed workers through retraining, reemployment services, and Unemployment Insurance to help displaced workers transition to new jobs and careers.
Workers displaced by automation face significant economic challenges.
Communities that are severely impacted by automation require targeted and comprehensive strategies to recover and transition.
» Support local economic development and improve regional competitiveness through sector-based development strategies and investment in digital infrastructure.
» CHALLENGES » SOLUTIONS
Understand the Impact of Automation on the Workforce
THE ASPEN INSTITUTE | FUTURE OF WORK INITIATIVE 22
» Provide key stakeholders with better information on the impact of automation by collecting data on technological advancements, adoption rates, and workforce impacts.
Policymakers, communities, workers, businesses, educators, and other stakeholders struggle to understand how automation is changing the economy because federal, state, and local data on the impact of technology on work is inadequate.
» CHALLENGE » SOLUTIONS
Automation and a Changing Economy
www.aspeninstitute.org/programs/future-of-work/automation/
April 2020