- Productivity growth has been declining since the 1960s and is now near historic lows. - Three waves have contributed to the decline in productivity growth since the 2000s: the waning of a mid-1990s productivity boom, financial crisis aftereffects like weak demand and uncertainty, and slowing capital intensity and total factor productivity growth in some countries. - Unlocking demand growth and promoting digital diffusion could help boost annual labor productivity growth above 2% in Western countries through both digital and non-digital opportunities.

1. SOLVING THE PRODUCTIVITY
PUZZLE: THE ROLE OF
DEMAND AND DIGITAL
CONFIDENTIAL AND PROPRIETARY
Any use of this material without specific permission of McKinsey & Company is strictly prohibited
JAANA REMES
June 28, 2018

2. McKinsey & Company | 1
Focus of the
study
Productivity growth has been declining since the 1960s and
today stands near historic lows
Trend line of labor productivity growth, total economy
% year-over-year
NOTE: Productivity defined as GDP per hour worked. Calculated using Hodrick Prescott filter. Drawn from similar analysis in Martin Neil Baily and Nicholas Montalbano, “Why is productivity
growth so slow? Possible explanations and policy responses,” Brookings Institution, September 2016
SOURCE: Bergeaud, A., Cette, G. and Lecat, R. (2016): "Productivity Trends in Advanced Countries between 1890 and 2012," Review of Income and
Wealth, vol. 62(3), pages 420–444.; McKinsey Global Institute analysis
1 Simple average across France, Germany, Italy, Spain, Sweden, UK.
2020
7
1
6
-1
5
4
2
3
-2
40801870 5090 2010 90 20007060 10301900 80
0
Europe1 USA
Great
Depression
Great
Recession
WWI WWII

3. McKinsey & Company | 2
We find patterns of a productivity-
weak, job-rich recovery with too
few accelerating sectors

4. McKinsey & Company | 3
In many countries, exceptionally low productivity growth post-recession
reflects slowing value-added growth despite faster growth in hours-worked
Compound annual growth rate, %
Germany
Labor productivity Value-added Hours-worked
SOURCE: The Conference Board (May 2017 release); McKinsey Global Institute analysis
3210 4-1
2010-20161985-2005
1 4320-1 4-1 30 1 2
Sweden
United Kingdom
France
United States
Italy
Spain
Average1
1 Weighted average using the 2016 share of real PPP GDP.
NOTE: Ordering based on fastest to slowest productivity growth in the 2010-2016 period

5. McKinsey & Company | 4
Shifts in aggregate productivity growth are the result of individual sectors
accelerating and decelerating; today we have too few jumping sectors
SOURCE: EU KLEMS; BLS; McKinsey Global Institute analysis
(1) sector is classified as "jumping" in year Y if its compounded annual growth rate of productivity for years Y-3 through Y is at least 3 pp higher than it was
for years 1995 to 2014 as a whole.
Time periods with top two and bottom two number of jumping sectorsUnited States example
4
0
1212
23
15
8
23
31
42
50
19
8
1515
19
15
031998 012000 0299 09
Ø 18
20140806 1310 1207 110504
Jumping
sectors1;
Share of
total; Total
sectors = 26
21 21 16 14 12 14 29 13 5 14 17 1124 18 8 0 4
Share of
value-
added2
% of total
nominal VA

6. McKinsey & Company | 5
-1.5
-1.2
-0.1 0.2
-0.7
-0.4
-0.2
-1.2
0.5
0.3
-0.5
-1.2-2.3
-0.5
0.0
0.5
0.1
-0.4
0.2
0.2
-0.9
0.0
-0.2
0.0
0.8
0.0
-0.21.4
Contribution to the decline in labor productivity growth, 2010–14 vs. 2000–04, Percentage
points1
Slow productivity growth was accompanied by a decline in capital intensity growth, as
well as declining total factor productivity growth in some countries
1 Analysis based on the Solow growth accounting framework. We have also calculated the contribution from productivity growth of each sector (a “within” effect, which weights the contribution of
a sector’s labor productivity growth by its share of nominal GDP) and the impact of labor movements across sectors with different productivity levels (a “mix” effect). 2 EU KLEMS data on TFP
presents a relevant discrepancy with other data sources such as Conference Board or Penn World Tables. Hence, we take the average TFP of the three databases and calculate L quality as
a residual 3 In Italy, the period analyzed is 2010-2013 instead of 2010-2014 due to data limitations 4 Data for US is only for the private business sector.
SOURCE: EU KLEMS (2016 release); BLS Multifactor Productivity database (2016 release); McKinsey Global Institute analysis
1.5
1.0
2.3
-0.2
2.9
0.9
1.7
0.9
3.6
-0.2
0.0
1.4
0.0
0.6
Increases productivity growthDecreases productivity growth
Labor productivity
growth,
2000–04 (%)
NOTE: Ordering of countries based on fastest to slowest productivity growth in 2010–14. Numbers may not sum due to rounding
Change in capital
intensity growth
Change in labor
quality growth
Change in sector
mix shift
2010–14 (%)
2 3
Change in total
factor produ-
ctivity growth
4

7. McKinsey & Company | 6
Three waves collided to drag
down productivity growth

8. McKinsey & Company | 7
Three waves explain these patterns and the low productivity growth of
today
Contribution to the decline in productivity growth
from 2010-14 vs 2000-04, Percentage points
(Average across France, Germany, Sweden ,UK
and US)
0.5
2.4
-0.2
-0.9
-0.8
2000-04 productivity growth
2010-14 productivity growth
Wave 1: Waning of a mid-1990s
productivity boom
Wave 2: Financial crisis aftereffects
including weak demand and uncertainty
Residual1
SOURCE: EU KLEMS (2016 release), BLS Multifactor Productivity database (2016 release), McKinsey Global Institute analysis
Wave 2
Sectors experiencing a boom/bust (finance, real estate, construction)
Excess capacity, slow demand recovery, uncertainty
Financial crisis-related hours contraction and expansion
First ICT revolution
Restructuring and offshoring
Wave 1
1 Includes impact of labor movement across sectors (‘mix effect”) and sectors not considered in our analysis. May include some of the impact from
transition costs of digital.
Wave 3: Digital disruption ???

9. McKinsey & Company | 8
Percentage point contribution to the decline in productivity growth from 2010-14 vs. 2000-04
-0.8
-0.1
-0.2
-1.1
-0.4
-2.0
-0.3
-0.9
-1.1
-0.9
-1.2
-1.3
The impact of each wave varies across countries
First ICT revolution
Restructuring and offshoring
-0.5
-2.5
-0.7
-3.8
-1.9
-2.0
Financial crisis
aftereffects
Mfg., Retail, Utilities,
Finance, Real estate,
Construction
Waning of a mid-1990s
productivity boom
Manufacturing, ICT, Retail,
Utilities Residual1
Total change in
productivity growth
Average
-0.5
-0.9
0.7
0
-0.2
-0.1
Other drag on productivity growth
Other support on productivity growth
0.1
0.2
-0.04
-0.1
0.01
-0.4
Impact of labor
movement across
sectors
(“mix effect”)
SOURCE: EU KLEMS (2016 release), BLS Multifactor Productivity database (2016 release), McKinsey Global Institute analysis
Financial crisis-related hours contraction & expansion
Excess capacity, slow demand recovery, uncertainty
Sectors experiencing a boom / bust (finance, real estate, construction)
1 Includes impact of sectors not considered in our analysis NOTE: US data includes only private business sector

10. McKinsey & Company | 9
United States and Western Europe, Productivity growth potential, Percentage points
Productivity growth
potential (2015-2025)
Non-digital opportunitiesDigital opportunities
~1.2+
2.0+~0.8+
SOURCE: McKinsey Global Institute analysis
Unlocking demand growth and promoting digital diffusion may deliver annual
labor productivity growth above 2 percent
NOTE: Our estimate for the productivity growth potential builds on extensive past MGI research on sector opportunities for improving productivity through
technologies that are already implemented today or have a clear path to deployment at scale by 2025. These include benefits from digitization (e.g., Big
Data, Internet of Things, automation, AI) as well as non-digital opportunities such as mix shifts in products and channels, continued consolidation, etc.

11. 10McKinsey & Company
How are AI and automation
transforming work?

12. 11McKinsey & Company
Our approach focuses on activities and the capabilities of
currently demonstrated technologies
SOURCE: Expert interviews; McKinsey analysis
Occupations
Retail
salespeople
Social1
Linguistic2
Cognitive3
Sensory perception4
Physical5▪ ...
▪ …
▪ …
~800 occupations
Teachers
Health practitioners
Food and beverage
service workers
Activities
Greet customers
▪ ...
▪ …
▪ …
Process sales and
transactions
~2,000 activities assessed
across all occupations
Clean and maintain work
areas
Demonstrate product
features
Answer questions about
products and services
?
Capabilities
Based on currently
demonstrated technology
capabilities as of 2016

13. 12McKinsey & Company
Less than 10% of jobs can be fully automated, but nearly all jobs will
be impacted
SOURCE: McKinsey Global Institute analysis
~50%
of current work
activities could
be automated
of jobs involve
tasks that are
>90% automatable
10%
But less than

14. 13McKinsey & Company
There are jobs which are more and less automatable
SOURCE: McKinsey Global Institute Global Automation Impact Model; McKinsey Global Institute analysis
% of automatable activities
Example occupations
>0%>10%
Psychiatrists Legislators
>20%>30%>40%
Fashion designers Chief executives
>50%>60%
Bus drivers Nursing assistants Web developers
>70%>80%
Stock clerks Travel agents Dental lab technicians
Sewing machine operators Assembly-line workers
>90%100%
~60%
~30%
of occupations
have
of tasks
automatable

15. 14McKinsey & Company
The automation potential of work activities varies by sector – US example
1 We define automation potential by the work activities that can be automated by adapting currently demonstrated technology
SOURCE: McKinsey Global Institute Global Automation Impact Model; McKinsey Global Institute analysis
FTE weighted percent of technically automatable activities by industry, Percent
31
36
37
38
39
40
40
41
43
44
44
44
47
48
53
53
55
58
64Manufacturing
Accommodation/food services
Transportation/warehousing
Information
Administrative/support/waste management
Management of companies/enterprises
Educational services
Finance/insurance
Arts/entertainment/recreation
Utilities
Construction
Professional, scientific, and technical services
Other services
Retail trade
Wholesale trade
Mining
Real estate/rental and leasing
Agriculture, forestry, fishing and hunting
Health care/social assistance
~50%
of work
activities
have the
potential to
be
automated

16. 15McKinsey & Company
45 48 56
78
55 52 44
22
High school or
some experience
Less than
high school
Some post-
secondary
education
Bachelor and
graduate degree
Automatable
Non-automatable
SOURCE: BLS 2014; O*Net; Global Automation Impact Model; McKinsey analysis
Automation affects occupations with lower educational requirements
disproportionately
Example
occupations
▪ Logging
equipment
operators
▪ Fast food
cooks
▪ Stock clerks
▪ Travel
agents
▪ Some
medical
technicians
▪ Lawyer
▪ Doctors
▪ Teacher
▪ Statisticians
▪ Chief
executives
▪ Nursing
assistants
▪ Web
developers
▪ Electricians
▪ Legal
secretaries
Technical automation potential of work activities by job zone in the US, %

17. McKinsey & Company
How can policymakers
help shape the outcome?

18. 17McKinsey & Company
Throughout history, large scale sector employment declines have
been countered by growth of new sectors that have absorbed workers
0.3
2.2
5.9
0.2
0.7
4.9
0.8
9.9
5.0
12.8
6.1
9.3
Employment share change, 1850–2015
Percentage points
Trade (retail and
wholesale)
Construction
Transportation
Agriculture, -55.9
Manufacturing, -3.6
Household work, +2.71
Mining, -1.3
Professional services
Utilities
Business and repair services
Telecommunications
Health care
Entertainment
Education
Government
Financial services
60
5
20
80
70
40
30
25
35
50
10
95
15
85
55
75
45
90
100
65
0
1850 2000
Share of total employment by sector in the United States, 1850–2015
% of jobs
1900 201550
1 Increase from 1850 to 1860 in employment share of household work primarily due to changes in how unpaid labor (slavery) was tracked.
SOURCE: IPUMS USA 2017; US Bureau of Labor Statistics; McKinsey Global Institute analysis

19. McKinsey & Company | 18SOURCE: McKinsey Global Institute analysis
Economic
growth
DemandSupply
Investment leakages due to
▪ Weak consumption
▪ Aging
▪ Rising returns on investment
with higher profit
concentration
Consumption leakages due to
▪ Weak productivity growth and
slowing workforce growth
▪ A partial diverging of real
wages from productivity due
to declining labor share and
rising house / land prices
▪ Rising inequality impacting
consumption
Long-term demand leakages could act as a drag on productivity and job
growth and may be further amplified by digital
Digital
transition may
amplify
leakages

20. 19McKinsey & Company
Strengthen
demand
by growing
purchasing
power,
investment, and
entrepreneurship
Invest in
human
capital
through
education,
training and life-
long learning
Reinvigorate
labor market
dynamism
by enabling more
diverse forms of
work and rethinking
transition support
for all workers
Accelerate
digital
diffusion
from
government to
SMEs
Some priorities to shape the future

21. @JaanaRemes
@McKinsey_MGI
McKinseyGlobalInstitute
DOWNLOAD MGI RESEARCH AT
WWW.MCKINSEY.COM./MGI