From Catch-Up to FrontierInnovation Growth
Philippe Aghion
Questions
How can China avoid the middle
income trap and succeed in
transition from “catch-up economy”
to “frontier innovator”?
 Does this also require institutional
change, not just policy change?


Rethinking the role and size of the
state?
Schumpeterian Paradigm


Long-run growth is driven by (frontier)
innovations
Innovations result from entrepreneurial
investments (R&D…) which are
themselves motivated by the prospect of
innovation rents


Policy of growth
Creative destruction: new innovations
make old technologies become obsolete

Political economy of growth
Schumpeterian Paradigm
A first prediction of the paradigm is
that (frontier) innovation requires
turnover, i.e reallocation and exit!
 A second prediction is that
competition enhances frontier
innovation and thereby growth
 A third prediction is that growthenhancing policies depend upon
stage of development

Frontier innovation vs catch up growth
Catch up growth in China



Partial market reforms and yardstick
competition between provincial leaders
Reallocation from agriculture to industry
and from SOEs to (credit-constrained)
new TVEs and private enterprises
(Song-Storesletten-Zilibotti)
Technological catch-up taking advantage
of FDI
From catch-up growth to innovationbased growth in China

More reallocation-based growth can be
achieved from liberalizing labor flows
from rural to urban areas and from
developing financial sector
From catch-up growth to innovationbased growth in China

Yet, several reasons for expecting a
slowdown:


Gains from reallocating resources from
agriculture to industry and from
absorption of imported technologies have
exhausting effects
Wage increases will reduce comparative
advantage of China in what it currently
exports
Pillars of an innovation-based
economy
First pillar: Competition
Competition/entry is more growthenhancing for countries or sectors that
are closer to technological frontier
 Competition/entry is more growth
enhancing in countries or states with
less regulated labor markets

Three fallacies about competition
policy



Competition policy would counteract effects of
patent policy: in fact the two policies are
complementary
Competition policy goes against any form of
industrial policy: in fact the two are
complementary
Competition policy works independently of
institutions: in fact corruption limits competition
Second pillar: education


Need good primary/secondary
education...importance of good PISA
performance
To have good graduate education is
more growth-enhancing closer to
technological frontier....importance of
good Shanghai rankings
Primary/secondary education

Quality, not just quantity, of
investment matters

Two illustrations
PISA and growth
 Investing more in more autonomous
universities, is more growth-enhancing

Autonomy of universities
Autonomie
20
Source : The Governance and Performance of ResearchUniversities: Evidence from Europe and the U.S. – P. Aghion et alii – NBER avril
2009
Third pillar: Labor market flexibility:
“flexsecurity”


Labor market flexibility is more growth
enhancing the closer a country is to the
technological frontier
Need to combine labor market flexibility with
reasonable unemployment benefits conditional
upon training for new jobs: flexsecurity!
EPL
Variable
Leader MFP growth
Gap to Leader
EPL
eq1
eq2
0.02949
0.02996
-0.00858***
-0.00836***
eq3
eq4
0.02830
0.02813
eq5
-0.00000
EPL, for highest tercile
0.00002
-0.00009**
-0.00011**
-0.00015***
EPL, for middle tercile
0.00004*
0.00002
0.00001
0.00001
EPL, for lowest tercile
0.00004
-0.00005
0.00002
0.00003
-0.01261***
-0.00816
-0.00547
Gap, for middle tercile
-0.00276
-0.00174
-0.00210
Gap, for lowest tercile
-0.00901***
-0.01095***
-0.01173***
EPL*Gap, for highest tercile
-0.00017
-0.00029*
EPL*Gap, for middle tercile
-0.00004
-0.00003
0.00012*
0.00014**
MFP Gap, for highest tercile
EPL*Gap, for lowest tercile
Leader growth, for highest tercile
Leader growth, for middle tercile
Leader growth, for lowest tercile
0.13600***
0.00817
-0.02597
legend: * p<.1; ** p<.05; *** p<.01
Flexsecurity
Flexsecurity
Flexsecurity
Flexsecurity
Fourth pillar: Finance


As country moves closer to frontier,
needs to rely more on equity finance and
stock markets
Reason is that innovative investments
are more risky and therefore investors
require both, to get a share of upside
returns and to get control rights (AghionBolton, 1992; Kaplan-Stromberg 2002).
Preliminary results
Finance, Growth and Distance to Frontier
Value Added Growth, 1980-1990
OLS
IV
OLS
IV
Stock Market * Financial Dependence
0.065
0.035
-0.008
-0.139
[.026]**
[.023]
[.058] [.069]**
Stock Market * Fin Dep * Dist to Frontier
0.289
1.072
[.327] [.448]**
Private Lending * Fin Dep
0.059
0.029
0.059
0.036
[.036]*
[.028]
[.034]
[.027]
Private Lending * Fin Dep * Dist to Frontier
-0.528
-0.919
[.164] [.243]***
Observations
972
661
887
638
R-squared
0.3
0.3
0.38
0.36
Country & Sector Dummies included.
* significant at 10%; ** significant at 5%; *** significant at 1%
Fifth pillar: Democracy


Democracy is more growth-enhancing for
industries that are closer to the technological
frontier
This is not surprising for at least two reasons:


Frontier innovation requires free thinking
Frontier innovation requires creative destruction,
but lack of democracy favors corruption and in
particular collusion between incumbents and
(local) leaders.
Reducing corruption or
increasing trust enhances growth
Free press reduces corruption
Sixth pillar: countercyclical
macroeconomic policy
Two Contrasted Views of How to
Conduct Macrooeconomic Policy
Keynesian view (non-discriminatory
increase in public spending)
 Conservative view (tax and spending
cuts)

A Third Way

There is a third way between
keynesian and conservative
approaches

namely, countercyclical fiscal and
monetary policy to partly circumvent
credit market imperfections and
thereby help firms maintain their
growth-enhancing investments over
the cycle.
Fiscal Policy Over the Cycle
17 OECD countries, 45 manufacturing
industries
 Period 1980-2005
 Finding: Countercyclical fiscal
policy enhances growth more in
sectors that are more dependent on
external finance or in sectors with
lower asset tangibility

Fiscal countercyclicality across
OECD countries
From fiscal to monetary policy
More countercyclical monetary policy,
i.e with lower short-run real interest
rates in recessions and higher rates in
booms...
 ....is more growth-enhancing in more
credit constrained or more liquidityconstrained sectors

Conclusion 1:

Moving towards frontier-innovation
growth requires:





Competition
Investing efficiently in education and
universities
Labor market flexibility and training
Stock market finance
Checks and balances to limit corruption
Conclusion 2:

Competitive and independent media
supported by rule of law, will spur
innovation-based growth for at least two
reasons:


It will put checks and balances on (local)
leaders, thereby reducing corruption
which in turn will foster creative
destruction
It will increase China’s attractiveness to
foreign researchers, more generally it will
enhance China’s “soft power” (J. Nye).
Conclusion 3: Seek higher quality
growth

Environment:
• State intervention to foster green
innovation and production

Income distribution:
• Excessive inequality encourages capture
and undermines competition and trust
• The top end stops contributing to public
good provision
Smart State: shouldn’t we all become
Scandinavians?




Targeted and well-governed growth
investments and wise countercyclical
macroeconomic policy
Social dialogue (high unionization rates) favor
external and internal labor market flexibility, and
enhance trust between firms and employees
Fiscal system which helps deliver on budget
balance, growth, inclusiveness, and the
environment
Politicians under strict checks and balances
(“Toblerone” story in Sweden)