Transcript R&D - CREI

Panel:
Innovation Policies for Developing
Countries – more of the same?
Manuel Trajtenberg
Tel Aviv University, NBER, CEPR
June 2005
1
The Rationale for Government
Support of R&D and Innovation
R&D
Social returns >> private,
market failures
K
TFP
Growth
Government intervention
(e.g. subsidize R&D)
Historically also “demonstration
effects” from success of military
R&D in WWII, cold war; raising
hopes in medicine (post-antibiotics).
2
Government Support to R&D – cont.
Government support in developed countries to:
– Basic research, higher education
– Defense-related R&D
– Health Care/Medical R&D
Also widespread support to commercial, civilian
R&D, menu of policies :
– Tax incentives
– Grants, conditional loans
– Incentives for tech transfer
– Jump-start VCs
3
The emerging view:
R&D/Innovation policy as panacea
• The Lisbon agenda: 3% R&D/GDP, the “K economy”
• Race in emerging economies to develop innovation
policies, set up Gov support to R&D (e.g. eastern
Europe, Central Asia, India, etc)
• Presumed success stories turned into “role models”:
Israel, Finland, Taiwan, Bangalore.
• Bandwagon effect in riding the globalization R&D
wave; rush to attract and set up Venture Capital funds
all over.
Is all this really relevant for development?
Constructive skepticism…
4
Sobering questions about R&D Policy
for Developing Countries
Global
R&D, K
• Why not free ride?
• Is the inflow really larger
than the outflow?
• What exactly do we
mean by “absorptive
capacity?”
R&D
K
TFP
• Need not only formal
R&D, but host of other
factors to get useful K
• What sort of innovations
impact TFP?
• Are these the same in
DCs and in LDCs?
Growth
Is “growth” 
“development” in
this context?
5
R&D Policy: The case of Israel
Background (1970’s):
Israel had little resources, but highly skilled manpower,
scientific and tech prowess – how to mobilize them for
growth? Government Strategic Decision: Jump-start and
breed a “science-based” sector by providing broad
financial support, and making up for market failures.
Hallmark of policies:
• “Neutrality”: respond to market demand/signals, do not
“pick winners.”
• Dynamic/Innovative: create new and varied support
programs according to evolving needs
• Mobility of personnel in and out of Gov. programs; not
6
“self-perpetuating”
Main R&D-Support Programs in Israel
•
Matching grants to commercial R&D projects: criteria:
innovativeness, tech and commercial feasibility, risk,
spillovers; paybacks if success; some strings attached.
• “Magnet” Program for support of generic, precompetitive R&D consortia: corporations + academia;
longer term, higher support. Examples: Nano Functional
Materials, Streaming Media Messaging, Digital Printing.
• Technological “Incubators” Program: from innovative
ideas to start-ups.
1993-97:
“Yozma” Program: Jump-started the Venture Capital
Industry – success, hence discontinued
7
Some indicators of R&D performance
in Israel: 1990-2000
• ICT production grew 4.6 times (16% per year); share of
GDP grew from 5% to 14%
• Exports grew 6.2 times (to $ 15 billion), 1/3 of total exports
• The ICT sector contributed 30% of the growth of GDP.
• 2nd largest VC market after the US
• US Patents per capita: fourth (after US, Japan, Taiwan)
• Major innovations: ICQ, disk-on-key, cardiac stents,
camera/pill for gastro imaging, shopping.com, etc.
• Finding of additionality: every additional dollar of R&D
support => 1.4 dollars of R&D performed (but caution…)
• 2004: R&D/GDP= 4.6% (world highest, but…), ~ 4,000
8
high tech companies.
And yet, sustained growth elusive…
• Wide disparity between fast growing High Tech
sector and the rest of the economy; stagnant
productivity of non-tradable, non-ICT sectors – a
“dual economy”.
• Rising socio-economic inequality: concern in
itself, and narrowing future pool of human capital.
9
Why?
Innovate here, benefit elsewhere…
1. In spite of neutrality, support mostly product
rather than process innovations that could be
applied locally; hence little R&D in e.g. services,
chemicals, etc.
2. Innovations made in Israel mainly for exports,
some spillovers internal, but benefits (e.g. TFP
growth) realized mostly abroad.
no “Wal-Mart effect” in Israel
10
Innovate here, benefit elsewhere – cont.
3. R&D labs of multinationals: absorb local talent, but
where do the benefits of the new K go? (e.g. Intel
Israel designed the Centrino chip for laptops – so
what?)
4. VC-backed star-ups: must exit, mostly by selling off
to US-based corporations – again, who benefits?
(very few large Israeli-based corporations)
Israel: powerhouse in generating innovations,
but not quite in benefiting from them.
And this is a successful case!
11
Some lessons from the
Israeli experience
1. Cannot have sustained growth by relying just on one
fast-growing sector (ICT), while the rest of the
economy stagnates:
Need to encourage and channel innovativeness also
to non-High Tech sectors
2. Cannot have growth while widening gap across socioeconomic segments of the population:
Need policies of inclusion, of reaching out to left-behind
segments, of expanding the pool of human capital.
12
Casting doubts on common perceptions
underlying R&D policies
• Associate R&D/Innovation just with “High Tech” as
perceived in the developed economies; distorted view!
• View that, because of globalization, only one game in
town: bring in MNLs R&D labs, bring in VC’s,
become players in global HT markets.
• Not such a thing as local markets, let alone local
“needs”, to be addressed with local innovation.
• The Internet fallacy: one universal network does not
mean that everybody has equal access, benefits the
same way, or needs the same tools to deal with it.
13
Innovation to accommodate
heterogeneous needs, local markets
Typical examples for LDCs (but many more):
• Health care: very different incidence of diseases;
need for cheap prevention rather than high end
technology, etc.
• In ICT: simpler software packages (not more
features), less demanding on hardware, more
backwards compatibility;
• If population spread over large, unwired areas:
satellite-based broad band internet.
• Cars: emphasize reliability and functional flexibility
14
(as the Ford Model T did…!)
Failure to innovate for local markets
in low-income countries
P
Demand: high
income countries
AC
fixed cost: R&D
Q
Demand: low income countries
15
Challenges for the design of sensible
R&D policies
• Map heterogeneity of countries in terms of what sort
of innovation/R&D required to foster their growth –
not “one size fits them all”
• How to play the global game and yet try to channel
(more of) the benefits of R&D/Innovation towards
the local economy.
• Understand what do we mean by absorptive
capacity, how to develop such capacity, etc.
• Sequencing of policies, given complementarities,
“chicken and egg” problems
16
Importance of sequencing:
some examples
Matching R&D grants :
Works if already large supply of S&T talent + entrepreneurial
drive. But if those are scarce, and supply inelastic, then
passive grants policy will be ineffective.
Neutrality:
Works well if available S&T skills are sufficiently general
and “unbiased” relative to available complementary assets.
Otherwise ineffectual.
Setting up a VC industry:
Works if already supply of start-ups, and binding constraints
are in funding, managerial skills, international connections.
17
A specific suggestion:
share R&D/innovation failures
• In best of cases low success rate of R&D projects
(e.g. in Israeli incubators 1 in 10): inherent to
nature of activity.
• At global scale, that means lots of replication of
failed R&D routes, waste of resources.
• Potentially a severe problem for LDCs, since
catching up and little resources.
Create information pool of failed projects, but need
clear set of incentives, perhaps sponsored by WB
(see also disclosure of failed patent applications)
Other information failures
18
Promote diffusion of ICT in LDCs
by prolonging life cycle of before-last generation
of products
Global acceleration of innovation => extremely fast
obsolescence of Knowledge, and of sophisticated,
expensive equipment (see multi-billion $ fabs).
Before-last generation of ICT “stuff” perfectly functional
(PCs, cell phones, software), but no longer produced.
Prolong operation of e.g. fabs by ~ ½ year, produce at ~
marginal cost (mostly sunk costs); secondary assured
market (like generics); win win (good PR for
MNLs…)
19
Call for genuinely innovative research
on “Innovation and Development”
• Cannot just translate Arrow (1962) to Swahili…
• Cannot just extrapolate from experience in
developed countries, cannot just borrow their
tools and policies
The truth is, we know very little – big payoff to
economic research in this realm!
This conference:
a step in that direction – good job!
20