The survey finds that U.S. experts rate U.S. innovation in integrated circuit design as high, while Chinese experts rate China's innovation as low. Chinese experts believe China's most innovative teams are 4-5 years behind the current frontier. U.S. experts believe China could catch up within 4 years if they devote sufficient resources. Both groups identify talent and intellectual property protection as key obstacles. The survey provides an initial measure of the innovation gap but has limitations such as a small sample size and potential differences in how experts define innovation.
Measuring U.S.-China Innovation Gap in Semiconductor Design
1. Measuring the U.S.-China Innovation Gap:
Initial Findings of the UCSD-Tsinghua
Innovation Metrics Survey Project
Peter
Cowhey
Tai
Ming
Cheung
with
Eric
Anderson
July
9,
2013
2. Current Innovation Metrics
Mismeasure Actual Innovation
• Most
common
metrics
are
inputs
and
acAviAes,
not
outputs
(R&D
expenditures,
patent
output,
number
of
STEM
graduates,
journal
citaAons)
• Magnitude
of
impact
innovaAon
inputs
have
on
commercial
innovaAon
varies
across
industries
Innovation
Activities
Innovation
Outputs
Innovation
Environment
3. New Survey to Measure
U.S.-China Innovation
• For
each
industry,
survey
answers:
– What
is
the
gap
in
innovaAon
between
the
United
States
and
China?
– At
what
rate
is
Chinese
innovaAon
catching
up
to
the
United
States?
• Includes
quesAons
on
innovaAon
environment
– DomesAc/internaAonal
financing
and
talent,
government
regulaAon,
collaboraAon,
geography
4. Survey Demographics
• Web-‐based
survey
administered
from
May-‐June
2013
• Received
68
U.S.
responses;
23
China
responses
• China
sample
less
IC
design
experience
than
U.S.
sample
– China
average
9
years;
U.S.
average
22
years
5. Correcting for
Cross-Country Differences
• Concern
that
US
and
China
sample
may
approach
term
“innovaAon”
with
different
concepts
and
definiAons
• Created
scenarios
depicAng
varying
levels
of
innovaAon
to
correct
for
bias
• This
report
only
uses
respondents’
assessments
of
their
own
country
Methodology
used:
h_p://gking.harvard.edu/files/gking/files/vign.pdf
6. Survey Results:
Innovation Level
• AcAviAes
of
5
most
leading-‐edge
IC
design
teams
– 67%
of
US
experts
rate
US
innovaAon
high
– 81%
of
Chinese
experts
rate
China
innovaAon
low
7. Chinese Experts’ Assessment
of China’s Innovation Gap
• China’s
most
innovaAve
IC
design
teams:
– Average
49
months
to
current
fronAer
– Average
57
months
to
advancing
fronAer
– 35
percent
of
Chinese
experts
say
China
will
never
catch
up
8. U.S. Experts’ Views
of Innovation Gap
“China
will
catch
up
in
about
4
years
Ame.
There
is
nothing
that
moves
fast
enough
in
most
IC
design
spaces
that
is
so
unique
that
it
can’t
be
overcome,
as
long
has
you
have
enough
money
to
pour
into
it.”
-‐-‐Execu've
Vice
President
of
Engineering,
Fabless
Semiconductor
Company
“The
duraAon
of
IC
design
development
depends
on
the
number
of
engineers
‘thrown’
at
a
problem.
Owing
to
costs,
companies
in
India/China
can
reduce
cycle
Ames
if
they
perceive
it
is
needed.”
-‐-‐Associate
Professor,
Integrated
Circuits
9. Survey Results:
Innovation Obstacles
• U.S.
three
largest
obstacles:
venture
capital,
qualified
talent,
foreign
compeAAon
• China
three
largest
obstacles:
weak
IP
protecAon,
qualified
talent,
lack
of
high-‐quality
IP
10. Survey Results:
Government Impact
• Chinese
view
industrial
policy
as
posiAve,
But
only
22%
rate
it
highly
posiAve
• Both
U.S.
and
China
have
balanced
view
of
public
services
(i.e.
tax
and
business
administra'on,
customs
services,
immigra'on
processing,
public
infrastructure)
11. Conclusion: Policy Implication
• Fair
consensus
that
China
is
4-‐5
years
behind;
substanAal
view
that
it
will
never
catch
up
• Biggest
problem
for
the
Chinese,
in
their
own
view,
is
intellectual
property
• China’s
industrial
policy
is
viewed
as
posiAve
but
not
as
highly
posiAve—not
the
crux
of
their
compeAAve
advantage
• Conclusion:
Devote
more
resources
to
IP
protecAon
and
fewer
resources
to
subsidized
R&D
• Need
to
extend
survey
to
other
sectors
and
conAnue
to
refine
measurements
• Survey
provides
measure
of
innovaAon
gap
at
leading
edge,
which
may
differ
from
gap
at
trailing
edge
12. FYI: Limitations and
Potential Mismeasurement
• Conceptual
limits:
– InnovaAon
definiAon
includes
products
and
processes
but
excludes
markeAng
and
organizaAonal
methods
– Survey
focuses
on
leading-‐edge
innovaAon,
capturing
the
fronAer
but
not
measuring
the
whole
industry
– InnovaAon
inherently
a
dynamic
concept—likely
to
vary
within
IC
design
subsectors
• Survey
Response
risks:
– Low
survey
response
rates
of
~5%
could
mean
non-‐representaAve
sample
– Pretested
quesAon
wording
but
experts
may
sAll
differ
in
interpretaAon
(i.e.
by
subsectors)