Changing Japan’s National Innovation System
and Its Policy Implications
Kazuyuki Motohashi
University of Tokyo & RIETI
http://www.mo.rcast.u-tokyo.ac.jp/
Outline of Presentation
• Characterizing Japanese System: Large firm dominated
system
• But it is changing. RIETI Survey on R&D collaboration
• Econometric analysis on R&D collaboration on research
productivity
• Policy implications for facilitating changes of J-NIS toward
network model
2
Business Expenditure of R&D in Japan
Toyota
6%
Matsushita
5% Sony
5% Honda
4%
Nissan
3%
Rest of firms
44%
Hitachi
3%
Toshiba
3%
NEC
3%
Other top 30
firms
19%
Cannon
2%
NTT
3%
3
Japan’s national innovation system
4
Changes in Large Firm Dominated
System?: Possible Factors
• Intense innovation competition: globalization and
catching up of Korea, Taiwan and China
• Necessity of innovation speed in order to
appropriate rents from R&D
• Importance of scientific knowledge for industrial
innovation: particularly the case for biopharmaceuticals
• Institutional changes in science sector: PRIs and
national university reforms
5
RIETI’s Survey on R&D Collaboration
• Firm level survey on external R&D collaboration: business
to business networks and university and industry linkages
• Data for 2003, 556 samples
• Survey items
– Recent trend of external R&D collaboration and IPR licensing
– Factors behind R&D collaboration decision
– Managing the boundary of firm in R&D, internal R&D vs
outsourcing
• Detail results are found as the following site
http://www.rieti.go.jp/jp/projects/innovation-system/H15.html
6
Japanese system is changing?
From RIETI survey
C ollaboration w ith Large firm s
0%
10%
20 or under
12.2
21-100
13.0
101-300
20%
30%
40%
50%
60%
11.6
70%
5 year s ago
Pr es ent ( i ncr eas ed)
24.4
6.1
38.8
2.3
46.0
1001-2000
2001 or over
12.8
68.1
4.3
C ollaboration w ith sm all firm s
20 or under
21-100
101-300
301-1000
1001-2000
2001 or over
10%
14.9
20%
30%
50%
5.7
16.5
25.2
22.0
60%
C ollaboratio w ith universities
70%
5 year s ago
Pr es ent ( i ncr eas ed)
4.2
11.0
40%
80%
90%
20 or under
21-100
4.1
101-300
10.3
301-1000
13.3
51.4
90%
5.6
301-1000
0%
80%
1001-2000
5.2
2001 or over
0%
10%
23.0
20%
30%
40%
50%
10.4
21.0
60%
70%
80%
5 year s ago
Pr es ent ( i ncr eas ed)
15.3
27.6
14.4
46.6
9.5
54.0
20.5
81.9
7
1.0
90%
Management of firm’s boundary in R&D
80%
70%
O w n R&D
W ith large firm s
W ith sm all firm s
W ith university
72.0
64.1
60%
50%
40%
52.8
51.2
47.6
47.0
46.5
37.7
32.9
30%
32.4
32.2
31.3
30.3
29.3
28.4
20.2
18.7
20.5
20%
16.9
16.2
13.5
10%
5.9
5.9
9.3
7.9
10.6
8.3
6.7
5.5
7.1
4.9
3.3
2.3
2.2
2.2
5.9
0%
Learning technology
Cost reduction
project
Basic science
Core technology
Technology
frontier project
R&D needing speed
R&D related to
core technology
New R&D subject
Product Development
project
8
Factors behind R&D outsourcing
Reasons why increasing R&D collaboration
36.7
38.1
21.3
20%
29.4
24.8
24.5
27.2
41.2
41.8
40%
48.6
51.4
48.5
50%
Universities
60.3
59.6
60%
Small firms
67.9
Large firms
70%
PRIs
55.1
10.1
5.5
5.9
3.0
1.5
8.2
5.5
2.9
2.0
10%
6.4
5.9
8.3
2.9
2.2
11.0
9.7
6.7
6.4
4.4
3.7
0%
16.9
14.9
29.9
29.9
30%
Easy to access
counterpart information
Important for technology
standard
Success in past
collaboration projects
Respond to intense R&D
competition
Upgrade own basic
technologycapability
Cost reduction of R&D
Needs to access to
basic science
Policy push for
industry-science linkage
Shortage of R&D
fund by own
Upgrading technology level
of counterparts
9
Motivation and underlining hypotheses
for econometric analysis
• Factors behing external R&D collaboration
– Intense innovation competition?
– Increasing complexity of innovation and the role of
scientific knowledge
– Selection and concentration of R&D projects, but it
needs wider technological scope
• UIC’s impacts on research and production
productivity: greater impact for small firms
– Less Not-Invented-Here Syndrome
– Focusing on more concrete project (short term benefit)
and greater pressure for commercialization
10
Important factors behind R&D
0%
10%
20%
30%
40%
50%
60%
66.9
M eet m arket needs
59.2
Shoten lead tim e
37.9
Explore new R &D them e
29.5
Focus R &D them es
22.3
C ost reduction
16.7
C om m ercialize ow n tech seeds
Shift to applied R &D
11.9
U pgrade basic R &D skills
11.5
Staff reduction
70%
4.5
11
Collaboration and R&D strategy
Shorter development lead time
Focusing R&D theme
Reduce R&D cost
Reduce R&D staffs
Explore new technology seeds
more R&D for application and
development project
large firms
SME
LF
++
--
Universities
SME
LF
++
++
+
+++
Improving basic technology capability
Market needs for R&D
Commercialization of tech seeds
SME and startups
SME
LF
++
++
++
++
+
-++
12
Research Productivity by Firm Age
all
(1)
0.276
(7.81)**
0.250
(6.08)**
-0.030
(0.23)
0.377
(3.21)**
all
(2)
lrd
0.260
(7.19)**
lemp
0.246
(5.41)**
cord
-0.056
(0.45)
univ1
0.355
(3.05)**
lage
-2.402
(4.81)**
lage2
0.360
(4.86)**
Constant
-1.683
2.302
(7.10)** (2.57)*
Industry Dummies
yes
yes
Observations
450
438
R-squared
0.62
0.64
Absolute value of t statistics in parentheses
* significant at 5%; ** significant at 1%
-1950
(4)
0.434
(5.61)**
0.397
(3.72)**
-0.131
(0.53)
0.203
(0.95)
1951-70
(5)
0.183
(3.05)**
0.315
(3.30)**
0.146
(0.67)
-0.077
(0.33)
1971(6)
0.109
(2.29)*
0.131
(2.84)**
-0.169
(1.06)
0.348
(2.09)*
-4.257
(8.51)**
yes
168
0.77
-1.188
(2.83)**
yes
134
0.55
0.439
(1.30)
yes
136
0.49
13
Research Productivity and R&D Project Type
Research Productivity
New Product Development
Scientific Research
Firm Age
14
Implications for J-NIS
Commercialization of
new product
Commercialization of
new product
Research Lab.
Scope of UIC
Close to the market
Product development
Systemic barriers
Large firm
SMEs
Start-ups
Scope of UIC
Scientific knowledge and fundamental science at
universities and PRIs
15
Synthesis and policy implications
• Growing trend of R&D external collaboration
• Reflecting firms’ R&D strategy for innovation speed and
wider technological scope
• Research productivity is higher for young and small firms
as compared to old and large firms
• The role of SMEs and start-ups for Japan’s NIS reform
toward network type system
• SMEs and start-ups: facilitates system’s transformation. In
addition, it may be beneficial for large firms to have strong
high-tech startups
• Policies for facilitating network type NIS system are
important, such as IPR, labor mobility, VC finance
16