Download PowerPoint-presentatie

Why is productivity growth so low
(in the 'Second Machine Age')?
Prof. Dr. Alfred Kleinknecht (Emeritus)
Visiting Fellow of WSI, Hans-Böckler-Stiftung,
Düsseldorf
www.alfredkleinknecht.nl
The labor productivity growth slowdown in advanced
OECD countries
(Source: https://www.conference-board.org/data/economydatabase/
Why is the productivity crisis relevant? (1)
You can achieve economic growth by:
1. Working more hours
2. Working more productive hours (i.e. more value added per
hour = higher labor productivity (through better technology
and organization)
If Option 2 fails, you need to use option 1 → thanks to the
productivity crisis, the US, Japan and West-Europe may get a
highly labor-intensive growth!
… as far as there is still growth (?)
Why is the productivity crisis relevant? (2)
Value added per working hour is (by definition) equal to National
Income
Hence, with low labor productivity growth, there is little to be
distributed (extra) between labor, capital and the public sector
Possible consequences:
 Pressure on public expenditures
 Tendency towards ‘precarization’ of work (working poor,
many atypical jobs)
 Erosion of the middle class
Three explanations of the productivity slump
(1)
A Keynesian explanation: weak effective demand
and investment ('balance sheet recession' after
2008)
Richard C. Koo: The world in balance sheet recession: causes,
cure, and politics, in: Real-world Economics Review, 1911
(issue 58) S. 19-36;
http://www.paecon.net/PAEReview/issue58/Koo58.pdf
Three explanations of the productivity slump
(2)
A low technological opportunities explanation:
Exhaustion of the American IT Boom
David Gordon, The rise and fall of American growth,
Princeton Univ. Press, 2016
The labor productivity growth slowdown in advanced
OECD countries
(Source: https://www.conference-board.org/data/economydatabase/
US ITBoom
Three explanations of the productivity slump
(3)
A varieties-of-capitalism explanation:
The Emergence of Supply-side Economics
After the 'Golden Age of Capitalism' (1948-1973) growth rates
declined substantially: rise of unemployment, high inflation
('stagflation' etc.).
This was a breeding ground for a revolution in the economics
discipline: from 'demand-side economics' (Keynes) to 'supplyside economics' (Hayek, Friedman) and a sharp attack on the
Keynesian welfare state.
What does supply-side economics mean?
 Sobering of the welfare state
 Privatization & deregulation an a broad scale
 No more anticyclical fiscal policy; only monetary policy
 More inequality of incomes ('performance must pay!')
 No full employment but a (sufficiently high) level of ‘natural‘
unemployment
 Deregulation of labor markets – Easy firing and (downwardly)
flexible wages
"There is no such thing like society …"
(Margret Thatcher)
But Supply-side Economics had a limited
reach:
P. Hall & D. Soskice, Varieties of capitalism, Oxford University
Press, 2001 distinguish two types of capitalism:
1. Anglo-Saxon ‘Liberal Market Economies’ (US, UK, Canada,
New Zeeland, Australia) with 'flexible' labor markets (weak
protection of labor)
2. 'Coordinated Market Economies' ('Old-Europe' + Japan)
with ‘rigid’ labor markets (strong protection of labor)
Complaint by supply-siders:
Old Europe is too slow in adopting
'structural reforms' of labor markets!
A comparison between Liberal Market
Economies (LME) and Coordinated Market
Economies (CME)
Comparing some core variables for
 Liberal Market Economies (US, UK, Canada, New
Zealand, Australia) versus:
 Coordinated Market Economies (proxy: EU-12)
Labour market deregulation leads to moderate wage growth!
Development of real wages,
1960-2004; 1960 = 100
450
EU-12 excl.
Luxemburg
Development of real wages, 1960 = 100
400
350
300
250
200
150
100
50
0
Anglo-Saxon
countries: US,
UK, Canada,
New Zealand,
Australia
In the long run: there is little difference in GDP growth …
Development of real GDP,
1960-2004; 1960 = 100
500
Anglo-Saxon
countries: US, UK,
Canada, New
Zealand, Australia
450
400
350
300
EU-12 excl.
Luxemburg
250
200
150
100
50
0
Year
… but GDP/working hour is growing faster in Old Europe
Development of labor productivity,
1960-2004; 1960 = 100
500
450
400
EU-12 excl.
Luxemburg
350
300
250
200
150
100
50
US, UK, Canada, New
Zeeland, Australia
0
Labor productivity = GDP per working hour
… and this results in job-poor growth in Europe and jobrich growth in Anglo-Saxon countries …
200
Development of total hours worked
1960-2004 (1960 = 100)
180
US, UK,
Canada, New
Zeeland,
Australia
160
140
EU-12 excl.
Luxemburg
120
100
80
60
40
20
0
Eurosclerosis!?
Core question: Does wage growth influence
labor productivity growth?
Traditional argument:
 Growth of labor productivity → wage growth (end of story)
My argument (to be proven):
 There must also be a causal link: wage growth → labor
productivity growth
Doubts:
W. J. Jansen: 'Kleinknechthypothese mist empirisch bewijs!', Economisch
Statistische Berichten, Vol. 89 (Sept. 2004), p. 418.
Rationale?
•
•
•
•
Speed of capital – labour substitution (Hicks)
Vintage effects (Tjan & Hartog)
Induced Innovation (van Weizsäcker; Kennedy)
Creative destruction (Schumpeter)
Slow
diffusion of
automation
technology
Key finding from panel data estimates: A one-per cent
wage decline (increase) leads to 0.35 ̶ 0.45% lower
(higher) growth of value added per labour hour
Sources:
Vergeer & Kleinknecht 2011: The impact of labor market deregulation on
productivity, Journal of Post-Keynesian Economics, Vol. 33(2): 369-404.
Vergeer & Kleinknecht 2014: Does labor market deregulation reduce labor
productivity growth? International Labour Review, Vol. 153(3): 365-393.
Evidence at macro-level is consistent with
findings from firm-level studies:
High shares of flexible labour in a firm are negatively related to:
 Labour productivity growth
 The probability to innovate
… and flexible labour enhances the growth of management
bureaucracies for monitoring & control
Recent paper: Alfred Kleinknecht, Flore N. van Schaik & Haibo Zhou (2014):
Is flexible labour good for innovation?, in Cambridge Journal of Economics,
Vol. 38(5): 1207-1219.
Summarizing:
 Deregulation of labour markets changes power relations
between capital and labour → lower wage growth
 … which has no influence on GDP growth, but …
 … causes a slower diffusion of advanced process
technology → lower growth of GDP per working hour
 … and hence a labour-intensive growth path
 ... but with low productivity growth there is little to be
(extra) distributed and this causes pressure towards:
– 'Precarization' of work
– Pressure for budget cuts in the public sector
– Problems financing demographic change
A similar story holds for the Netherlands …
As a response to the Dutch Disease (ca. 1975-85), trade
unions made very modest wage claims during the past 35
years …
… problems with effective demand were solved by means of
high current account surpluses …
… and The Netherlands experience a strong job growth!
Broad national consensus in the Netherlands in favour of
"Loonmatiging" (= very low wage claims)
Low wage strategy brought no higher GDP
growth
(other than expected by the mainstream)!
Growth of GDP
EU-15
NL
1950-60
4.5
4.6
1960-73
5.2
4.9
1973-80
2.6
2.4
1981-90
2.4
2.2
1991-00
2.5
2.8
≈1% extra GDP growth
through a housing bubble
Source: Groningen Growth and Development Centre
www.ggdc.net / Total Economy Database
… but a highly labor-intensive GDP growth!
Growth of GDP
Growth of labour
hours per 1% BBP
growth
EU-15
NL
EU-15
NL
1950-60
4.5
4.6
0.07
0.10
1960-73
5.2
4.9
− 0.09
0.07
1973-80
2.6
2.4
− 0.15
- 0.05
1981-90
2.4
2.2
0.12
0.57
1991-00
2.5
2.8
0.13
0.61
Source:
Groningen Growth and Development Centre
www.ggdc.net / Total Economy Database
Highly jobintensive
growth!
At a given GDP growth, growth of labor input
depends on labor productivity growth!
Growth of GDP
Growth of GDP per
labor hour
Growth of labor
hours per 1% GDP
growth
EU-15
NL
EU-15
NL
EU-15
NL
1950-60
4.5
4.6
4.2
4.2
0.07
0.10
1960-73
5.2
4.9
5.7
4.5
− 0.09
0.07
1973-80
2.6
2.4
3.0
2.5
− 0.15
− 0.05
1981-90
2.4
2.2
2.1
1.0
0.12
0.57
1991-00
2.5
2.8
2.2
1.1
0.13
0.61
Source:
Groningen Growth and Development Centre
www.ggdc.net / Total Economy Database
Remember: Above we had 3 neoclassical
arguments of why weak wage growth
would hamper labor productivity growth:
(1) Factor substitution (Hicks)
(2) Induced Innovation (Kennedy & von Weizsacker)
(3) Vintage effects (Tjan & Hartog).
… but there are more (not-so-neoclassical) arguments why
'structural reforms' of labor markets can impede innovation
and productivity …
Why are CMEs outperforming LMEs in
labor productivity growth? (1)
Deregulation of labor markets leads to “more dynamism”,
i.e. higher job turnover, which in turn reduces loyalty and
commitment of workers:
→ more
leaking of trade secrets and technological
knowledge (i.e. stronger externalities)
→ more need for monitoring & control
→ thicker management bureaucracies (next sheet)
Share of managers in the working population
(19 OECD countries, 1984-1997)
Norway
Spain
Greece
Sweden
Italy
Switzerland
Belgium
Ireland
Germany
Portugal
Japan
Denmark
Finland
Austria
Netherlands
U.K.
Australia
USA
Canada
According to De Beer (2001),
the Dutch figure increased
from 2% to 6% during 19781998 (the period of
‘flexibilization’ of the Dutch
labour market)
0
5
10
15
Managers as a percentage of the non-agrarian working population
Why are the CMEs outperforming LMEs
in labor productivity growth? (2)
Deregulation of labor markets and higher job turnover
lead to:
 Lower investment in firm-financed training
 Lower benefits from “learning by doing” and weaker
“organizational memories” (learning from past
failures)
 More power for top management and less critical
feedback from the shop floor → more autocratic
management
Why are the CMEs outperforming LMEs in
labor productivity growth? (3)
Flexible firing increases risk-aversion on the shop floor: in the
selection of innovative solutions, people that are easy to fire
will choose less risky options → too little progression!
Empirical support using patent and patent citation data:
Acharya, Viral V., et al. (2010): Labor laws and innovation,
NBER Working Paper 16484. Cambridge, MA: NBER
Why are the CMEs outperforming LMEs in labor
productivity growth? (4)

People on the shop floor possess much of the (tacit)
knowledge required for implementation of process
innovations. People threatened by easy firing have
incentives to hide knowledge relevant to labour-saving
process innovations (Lorenz, 1992, 1999)
More generally, people that are easy to fire have strong
incentives hiding information about how their work can be
done more efficiently (exploiting information asymmetry
between management and the shop floor)
→ in a hire & fire regime you make poor use of the (tacit)
knowledge of your workers
Why are the CMEs outperforming LMEs in
labor productivity growth? (5)

CMEs tend to have more centralized wage negotiations;
wage increases force technological laggards to modernize
equipment

Other than a “Garage Business” model of innovation, a
“Routinized Innovation Model” requires a continuous
accumulation of (often: tacit) knowledge → need for
continuity of personnel
"Tacit knowledge" = poorly documented, ill-codified
knowledge from experience (Polanyi 1966)
Much depends on the type of innovation model:
Flexible labor has no
influence on innovation
Schumpeter I innovation model:
'Entrepreneurial model': start-ups, inventor-entrepreneur ('Garage
business').
Knowledge base: General (and generally available) knowledge
Schumpeter II innovation model:
'Routinized Innovation': Professional R&D labs in larger firms.
Continuous improvement of products, processes or systems requires
continuous accumulation of knowledge.
Knowledge base: Historically accumulated, firm-specific and often 'tacit'
knowledge (i.e. ill-documented knowledge from personal experience)
Flexible labor has a strongly negative influence on
innovation and productivity growth
Conclusion:
Supply-side economists correctly argue:
Deregulation of labor markets (weak trade unions, no
minimum wages, poor social benefits, downwardly flexible
wages etc.) leads to more employment …
… but they forget to tell us that this is at the cost of low labor
productivity growth …
… and hence somebody needs to sacrifice income …
… in real life we get a growing class of working poor on
precarious jobs …
… and an erosion of the middle class (Trump effect!)
General conclusion:
What is useful from a neoclassical view ('how to allocate
scarce resources efficiently') …
… can be counter-productive from a Schumpeterian view
('how to make resources less scarce through innovation?')
A legacy of Schumpeter: (Static) allocative
efficiency is no good milieu for innovation!
"Perfect competition … is a condition for optimal allocation of
resources … But … introduction of new methods of
production and new commodities is hardly conceivable with
perfect … competition … And this means that the bulk of …
economic progress is incompatible with it. As a matter of fact,
perfect competition is and always has been temporarily
suspended whenever anything new is being introduced …".
Joseph A. Schumpeter: Capitalism, Socialism and Democracy, London: Allen
& Unwin 1943, p. 104-105.
Rounding up:
Two alternative strategies for capitalism
Supply-Side Economics: Through ‘Structural Reforms’ to flexible
labor markets: easy firing + poor welfare state:
 A much more unequal income distribution
 Overall poor labor productivity growth → many (but often
precarious) jobs!
European Social Model: Rigid labor markets + strong welfare
state + tough investments in education and research:
 High speed of labor-saving technical change → Many highly
productive jobs for protected insiders, but:
 Poor overall growth of labor input, and therefore …
 Shorter working times rather than wage claims  Reduce
labor supply!
More reading?
www.alfredkleinknecht.nl