Transcript Cliometrics and Technological Change

Cliometrics and
Technological Change
Nick Crafts
Centrality of Technological
Progress
• The defining characteristic of Modern
Economic Growth (Kuznets)
• “Our knowledge of economic history, of
what production looked like 100 years ago,
and of current events convinces us
beyond any doubt that discovery,
invention, and innovation are of
overwhelming importance in economic
growth” (Paul Romer, 1994)
What Is Distinctive?
• Long-run perspective in which want to measure
and to explain success and failure in economic
growth
• Invention is partly endogenous
• Localized technological progress
• Divergence Big Time
• Historical approach (no tabula rasa)
• NB: demands more than neoclassical
economics has to offer, cf. cliometric
schizophrenia
Economic History and Economics
Solow (1985)
The danger is that “economic theory learns
nothing from economic history, and
economic history is as much corrupted as
enriched by economic theory”
And
“an economic historian is merely an
economist with a high tolerance for dust”
Keywords
• Labor Scarcity
• Macro vs. Micro Inventions
• New Institutional Economic History
• Path Dependence
• Social Capability
• Social Savings
Key Players
• Moses Abramovitz
• Robert Allen
• Paul David
• Robert Fogel
• Joel Mokyr
• Douglass North
What Sort of Growth Theory?
• Exogenous
• Endogenous
• AK
• Unified
• Appropriate (endogenous innovation)
• Better menu now but some items still
indigestible
Macro vs. Micro Inventions
Mokyr (1990)
• Macro Inventions: radical new and
economically important ideas that come along
occasionally and unpredictably
• Micro Inventions: incremental improvements
that account for most TFP growth and reflect
economic incentives
• Implies that explaining timing of technological
breakthroughs (new GPTs) problematic and
time-series predictions of endogenous growth
models unreliable
Solow’s Seminal Paper
“Gross output per man hour
doubled over the interval (190949) with 87.5% of the increase
attributable to technical change
and the remaining 12.5% to
increased use of capital.”
(1957)
Crude TFP vs. Capital Deepening
• US not the same before or since the Solow
period
• Neither “Take-Off” Europe nor “Golden-Age”
Europe match the 7/8 proportion
• Crude TFP growth ≠ technical change generally
• Solow’s famous result, while still correct, is an
outlier
US: Crude TFP Growth, % per year
(Percentage of Labour Productivity Growth)
1800-55
0.2
(50)
1871-90
1.0
(55)
1890-1905
1.3
(74)
1905-48
1.4
(85)
1948-66
1.9
(74)
1966-89
0.3
(50)
Source Abramovitz & David (2001)
Europe: Crude TFP Growth, % per year
(Percentage of Labour Productivity Growth)
Germany
Netherlands
UK
Sweden
1871-1913
0.95 (66)
0.61 (53)
0.50 (67)
0.72 (42)
1950-73
4.32 (72)
3.35 (75)
2.14 (67)
1.86 (51)
1970-90
1.09 (58)
0.90 (56)
1.06 (56)
0.63 (48)
Sources: Crafts (2008); Bosworth & Collins (2003)
TFP Growth ≠ Technical Change
• Maddison’s ‘educated guesses’ suggest technical change
only a subset of TFP in 1950-73 Europe; later econometric
analysis confirms that efficiency improvement a substantial
part of Golden Age TFP growth, e.g. Jerzmanowski (2007)
• TFP growth underestimates technical change in 19th
century USA because standard assumptions inappropriate
…… elasticity of substitution <1 and not Hicks-Neutrality
(Abramovitz & David, 2001) and quite possibly in East Asian
tigers (Rodrik, 1997) and controversially in Soviet Russia
(Weitzman, 1970)
• Intertemporal comparisons of TFP growth are not a
reliable indicator of differences in rate of technical change
Unified Growth Theory
Galor & Weil (2000)
• 3-stage model; first stage is the
Malthusian economy
• Onset of ‘modern economic growth’ is
endogenous and triggered by population
growth
• Positive feedback from population size
means technological progress increases in
the late Malthusian phase
Implications for Unified Growth Theory
• Scope for considerable population growth in preIndustrial Revolution England but cannot see the
GW feedback (Crafts & Mills, 2009)
• Maybe a ‘peso problem’; need technological
progress to materialize in sectors with
favourable demand conditions
• Or perhaps technological change deserves to be
taken seriously and other aspects of
endogenous growth literature have more
potential
First Industrial Revolution
• Not just a scale effect
• Not ‘take-off’
• Ingenuity rather than abstention
• Putting acceleration in TFP growth into
proper perspective
• Endogenous innovation gives some
insights, mutatis mutandis (Allen, 2009;
Mokyr,2002)
Growth in Britain (% per year)
1760-1801
1801-31
1831-60
GDP
1.0
1.9
2.5
TFP
0.1
0.4
0.8
GDP
1.1
2.7
2.5
TFP
0.2
1.3
0.8
Crafts (1985)
Feinstein (1981)
Industrial Revolution and
Relative Factor Prices (Allen, 2009)
• Key characteristic of Britain is high-wage/cheap energy
economy by international standards (cf. the Habakkuk
controversy)
• Adoption of famous inventions initially only had
acceptable payoff at British relative factor prices
• More likely to invent what pays, so Britain was in pole
position
• Eventually, continuing technological progress (e.g.
reduced coal consumption in steam engines) means that
British technology dominates and diffuses internationally
Steam Power and British Economic Growth
• Did not dramatically raise the growth rate
• Peak impact about 100 years after James
Watt
• Technology not well understood and not
cost-effective initially
• GPT growth-accounting arithmetic
reconciles co-existence of slow growth in
the Industrial Revolution and famous
inventions
Total Steam Contribution
0.45
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
1760-1800
1800-30
1830-50
Source: Crafts (2004): includes railway, steamships, steam engines
1850-70
1870-1910
Social Savings of Railways
• Transport benefits (consumer & producer surplus) =
total economic benefits; OK with perfect competition
and CRS in transport-using sector
• Fogel (1964) aimed to measure contribution of new
technology to growth and to destroy the ‘myth of
indispensability’; freight SS = 4.7% US GDP in 1890
• Upper bound measure: (palt – prail)qrail
• Compared with growth accounting includes own TFP
contribution but no spillovers and no capital deepening;
in absence of railway other normal-returns investments
Transport Benefits May Not Capture
All the Economic Benefits
• Imperfect competition in the transport-using
sector
• Agglomeration benefits from industrial relocation, bigger cities etc = TFP spillovers; NEG
might claim key to emergence and ‘lock-in’ of
manufacturing belt (cf. Fogel’s counterfactual)
• Fogel (1979) argued upper bound bias enough
to compensate for these omissions but depends
on strong (ahistorical?) assumptions
The Habakkuk Hypotheses (1962)
• Labor scarcity in 19th-century US meant (1)
more machines, (2) better machines and (3)
more labor-saving bias in technological change
relative to UK
• (2) and (3) not very neoclassical but may fit with
endogenous innovation model (cf. Acemoglu, 2009)
• David (1975) provided interpretation based on
localized learning (cf. Atkinson & Stiglitz, 1969)
• A key corollary is that imposing conventional
growth accounting assumptions underestimates
US rate of technological progress
Choice of Technology: The role of factor prices
P1
KM/Y
A
Po
B
APF
Po
P1
KH/Y
Source: Broadberry, 1994
Technical progress: Localised change
α
KM/Y
A
APF
β
A′
B
APF′
Source: Broadberry, 1994
KH/Y
Did Victorian Britain Fail?
• NEH said NO!; different choice of technology was profit
maximizing based on different cost conditions
• Micro-level neoclassical analysis undermines the
neoclassical growth model’s universal technology
assumption
• Implies an ‘inappropriate technology’ (Acemoglu & Zilibotti,
2001) interpretation of lower British TFP
• Labor-scarcity and mass-market driven US technology
lacked ‘technological congruence’ for Europe in early but
not in mid-20th century (Abramovitz & David, 1996)
Decomposition of output difference between
countries 1 & 2 under appropriate technology
y
2
•
•
y = A2kα
due to k
y2
y′1
due to TFP
•
•
1
y1
k
Source: M. Jermanowski, European Economic Review 51 (2007)
Decomposition of output difference between
countries 1 & 2 under appropriate technology
y
2
•
•
y2
y′1
y′′1
•
y = A2kα
due to k
y = A(k)k α
due to T
due to E
•
•
1
y1
k
Source: M. Jermanowski, European Economic Review 51 (2007)
20th vs 21st Century
• “The restoration of inter-society
income equality will be one of the
major economic events of the century
to come” (Lucas, 2000)
• So divergence will be superseded by
convergence according to neoclassical
ideas but NIEH thinks differently
The Lucas Model
• Countries that start growing later have faster
initial growth and then experience growth
proportional to income gap with leader
Dy/y = a + bx
where x is the number of 50 year periods
since 1800. Hazard rate for beginning
modern economic growth (l) evolves from
0.001 to 0.03
World Growth Rate and Income
Variability
3.5
3
2.5
Parameter Values:
a = .02
b = .025
l min = .001
l max = .03
 = .5
Annual Growth Rate
2
1.5
Log Standard Deviation
1
0.5
0
1800
1850
1900
1950
2000
2050
2100
Lucas’s Underlying Argument
• Obstacles to growth removed through
imitation of good policies, institutions
• In globalized world, capital mobility and
financial liberalization relax the savings
constraint
• K/L and TFP gaps are rapidly reduced
Social Capability (Abramovitz, 1986)
• Catch-up not automatic; need good
incentive structures to assimilate
technology effectively
• Institutions and policies need to evolve
as move from far-from-frontier to close-to
frontier (cf. Aghion & Howitt, 2006; Gerschenkron,
1962) but status quo bias may intervene
• Social capability not independent of
technological epoch (cf. Europe and ICT)
The NIEH View (North, 1990)
• Institutions are the underlying determinants of
economic performance
• Institutions are formal and informal
constraints that structure behaviour
• Institutions are persistent; expect multiple
equilibria and path dependence
• Institutional reform is opposed by vested
interests and subject to government failure
• So divergence may continue indefinitely
Path Dependence
• ‘History matters’; irreversible changes in
outcome probabilities and possible ‘lock-in’
• Not really about QWERTY-type examples
(David, 1985)
• Institutional and geographical aspects
much more important; divergence not
convergence (North, 1990; Krugman, 1991)
• Problematic for social savings (cf. David,
1969)
Conclusions
• Cliometrics needed more than
mainstream economics had to offer;
source of tension
• Scope of historical questions is very
demanding and counterfactuals
difficult to quantify
• Growth economics has moved on and
perhaps is a bit less corrupting