com8 techno-economic systems institutional innovation (3

1

Chihiro Watanabe

23rd JyvaskylaSummer School

7-9 Aug.. 2013, University of Jyvaskyla, Jyvaskyla

COM8Techno-economic SystemsInstitutional Innovation

(3)

Professor, Department of Industrial Management, Tokyo Seitoku UniversityProfessor Emeritus, Tokyo Institute of Technology

Visiting Professor, National University of SingaporeVisiting Professor, University of Jyvaskyla, Finland

Research Scholar, International Institute for Applied Systems Analysis (IIASA)

[email protected]

COM8 (3)

1. 7 Aug (W) AM2. PM

3. 8 Aug (T) AM4. PM

5. 9 Aug (F) AM6. PM

Technological innovation, growth, diffusion and consumptionProductivity, technological progress, competitiveness

Diffusion of technology, Effects of learningTechnology spillover, Rate of return to R&D investment

Basic concept of institutional innovationNew Stream for institutional innovation

2

COM8: Techno-economic Systems, Institutional Innovation

Chihiro Watanabe ([email protected])

AM: 10-12 am PM: 13-15pm

Identity: SEARCH Systems approach, Empirical approach, Analytical approach, challenge to Rationale, Comprehensive approach, with Historical perspective

3

3. Diffusion of Technology

3.1 Epidemic Function

3.2 Application of Epidemic Function

3.3 Variation of Diffusion Function

3.4 Functionality Development

3.5 Integration of Production Function and Diffusion Function - Innofusion

3.6 Sustainable Functionality Development in Open Innovation

Fig. 1. Comparison between Logistic Function and Epidemic Function.

3. Diffusion of Technology3.1 Epidemic FunctionDiffusion trajectory of innovation, innovative goods and also new products resemble to diffusion process of an epidemic disease which can be depicted by the epidemic function (Verhulst, 1845).

Y（t）

Y（t） : Diffusion level at time t: Upper limit of diffusion (Carrying capacity)

a : Diffusion velocityb : Diffusion level at initial period

N

N

atbeNtY

1

)(Epidemic function :

Logistic function

Epidemic function

t

2N

abln

Pierre Francois Verhulst (1804-1849) Mathematician in BelgiumMathematical Researches into the Law of Population Growth Increase (1845).

bN1 Changed notations

NN

tYtN

)()(

4

(1) Development of Epidemic Function

5

YN N

Y

Level of disseminationof epidemic

disease

Upper limit of dissemination(carrying capacity)

Potential capacityof dissemination

Fig. 2. Diffusion Process of Epidemic Disease.

1. Epidemic disease diffuses proportional to potential capacity of dissemination

2. Increase of epidemic disease N during the unit period dt

3. Develop this balance

4. Solving this differential equation leads to the followinglogistic growth function (epidemic function) illustratinga sigmoid curve:

)(' YNa where a’: coefficient

dtYYNadY )]('[

)1(NYaY

dtdY

Naa 'where

atbeNY

1

where b: coefficient

(21)

(22)

')'(

)'(

2

11,1),'()''(ln,'1

'1,',')(',,)(),('),('

batbat

bat ebbeN

eNY

YN

YYNeXbatbNtaXdtNadX

X

NadtdX

XNa

XXX

NYYXaYNa

YY

YYNX

YYNXYNa

YYYNYaY

dtdY

6

（２）疫学関数

係数を書き直して整理（a’, b’→ a b, a”→ a’）すると、

0dt/)t(dN)NN(bNdt/dN 　　の時、 )t(NN0))t(NN( 　　

(3)

FFlnabt

1

（３）フィッシャ・プライ変換（Fisher-Pry transform）

（４）普及天井の決定

(１)式より、

abteNNF

1

1

btabt eaN

eNN

)NN(bN

dtdN

11

1

N

2/N

aeN

1 t

b: 普及速度 a: 初期時点の普及状況 t -a/b : 逓増 t = -a/b : 転換点

t > -a/b : 逓減

-a/b

累積普及量 N

普及量 dN/dt

4/Na

(2) Structure of Epidemic Function

Fig. 3. Diffusion Trajectories of Epidemic Disease by Increase and Cumulative Stock.

Cumulative stock Y

abln

atbeNtY

1

)(

bN1

)1(NYaY

dtdY

(3) Fisher-Pry Transform

atbeNYF

11

FFbe at

1F

Fbat

1

lnln

(4) Identification of Carrying Capacity

0# dtdY 0))(1)((

##

# NtYtaY

dtdY

time when#t NtY )( #

N

2N

4aN

dtdY

7.

6 0 0 ．

5 0 0 ．

4 0 0 ．

3 0 0 ．

2 0 0 ．

1 0 0 ．

0

1 7 9 6 1 7 7 0 1 7 8 0 1 7 9 0 1 8 0 0 ・・・・・ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

D I e a t h

B o r n

C u m u la t iv e n u m b e r o f c o m p o s it io n s p u b li s h e d

W o lfg a n g A m a d e u s M o z a r t (1 7 5 6 - 1 7 9 1 )

T h e b e s t- f i t t i n g S -c u r v e im p lie s 1 8 c o m p o s i to n s “ m is s in g “ b e - tw e e n 1 7 5 6 a n d 1 7 6 2 . T h e n o m in a l b e g in n in g o f th e c u r v e - - - th e 1 % le v e l -p o in t s a t M o z a r t ‘ s b i r th d a y. T h e n o m in a l e n d th e 9 9 % le v e l in d ic a t e s a p o te n t ia l o f 6 4 4 w o r k s .

1 6 0 ．

1 4 0 ．

1 2 0 ．

1 0 0 ．

8 0 ．

6 0 ．

4 0 ．

2 0 ．

1 8 2 0 1 8 3 0 1 8 4 0 1 8 5 0 1 8 6 0 1 8 7 0 ・・・・・・

. . . . . . . . . . . . . . . . .

C u m u la t iv e n u m b e r o f c o m p o s i t io n s p u b li s h e d

R o b e r t S c h u m a n n ( 1 8 1 0 -1 8 5 6 )

D e a lt h

A t le m p t e d s u ic id e

N e r v o u s b r e a k d o w n

1 %

T h e p u b l ic a t io n o f S c h u m a n n ‘s c o m p o s i t io n s . T h e f i t t e d c u r v e b e g in s a r o u n d1 8 2 6 a n d a im s a t a c e i l i n g o f 1 7 3 . S c h u m a n n ‘s p u b l ic a t io n s r e a c h e d 1 7 0 , s ix te e n y e a r s a f te r h is d e a th .

・・・・・・・・・・・・・・・

2 8 0

2 4 0

2 0 0

1 6 0

1 2 0

8 0

4 0

0

C u m u la t iv e n u m b e r o f p u b lic a l io n s

A lb e r t E in s te in ( 1 8 7 9 -1 9 5 5 )

. . . . . . . . . . . . . . . . . 1 8 8 0 1 8 9 0 1 9 1 0 1 9 2 0 1 9 3 0 1 9 4 0 1 9 5 0 1 9 6 0

E in s te in ‘ s c u m u la t iv e p u b l ic a t io n s a n d th e b e s t - f i t t in g S -c u r v e . T h e f i t in d ic a te s 1 3 p u b l ic a t io n s “ m is s in g “ b e tw e e n th e b e g in n in g o f t h e c u r v e , 1 8 9 4 , a n d E in s tc in ‘ s f i r s t p u b l ic a t io n in 1 9 0 0 . T h e c e i l in g i s e s t im a te d a s 2 7 9 .

モーツアルト、シューマン、アインシュタインの「伝染過程」

資料 T h e o d o r e M o d is , P re d ic t io n (S im o n & S c h u s te r, N e w Y o r k ,1 9 9 2 ) .

D I E D

B o r n 1 %

Fig. 4. Epidemic Process of Mozart, Schumann and Einstein.Source: Theodore Modis, Prediction (Simon & Schuster, New York, 1992).

The social and economic operation of our society can be decomposed in a very large number of sub-diffusionprocesses summing up into an almost inextricable whole (Cesare Marchetti,1996).

3.2 Application of Epidemic Function

8Source: Theodore Modis, Prediction (Simon & Schuster, New York,1992).

Fig. 5. Einstein’s Epidemic Trajectory by Cumulative Number of Papers (1879-1955).

Cumulative number of papers Death (1955)

1880 1890 1910 1920 1930 1940 1950 1960 t

(Carrying capacity)

1%

-

280 -

240 -

200 -

160 -

120 -

80 -

40 -

0

Birth(1879)

YN

atbeNY

1 1894: Published 1% of whole papers in his life at the age of 15 years

old1905: Special relativity

1915: General relativity

3.3 Variation of Diffusion Function(1) Simple Logistic Growth Function (SLF)

9

))(1)(()(N

tYtaYdt

tdY (23)

Y(t): Diffusion level of innovative goods at time t; N: Upper limit of diffusion level (carrying capacity); a: Coefficient.

atbeNtY

1

)( (24)

))(1()()(

)(

)(

NtYa

tYtY

tYdt

tdY

a governs diffusion velocity.

Diffusion level at the initial timing.

bNY

1

)0( 1)0(

YNb

b indicates initial level of diffusion.

)2(

)1(1

)1(1()1(

)1

11(1

)1(

1

1)

atbeat

atb

atat

at

at

atat

beaN

ebeaN

bebe

beaN

bebeaN

NYaY

dtdY

at

dtdY

atat

dtdY

bexaNyx

xy

bebeaN

,12

12

Bi-polarization Diffusion Trajectory

10

1

2

xy 2

xy 1

xxy 12

atebx

dtdY

aNy

abt ln

India

Philippines

Indonesia

ChinaThailand

MalaysiaMexico

Brazil

Russia

Turkey

PolandSlovak Republic

TaiwanKorea

Czech RepublicPortugal

New ZealandGreece

SpainItaly United KingdomIcelandJapanCanadaGermanyFranceSweden BelgiumFinlandAustriaUnited StatesNetherlandsIreland

Denmark

Switzerland

Norway

Luxembourg

Case 1: GDP per capita vs Marginal productivity of investment per capita

11

1 GDP per capita increase

MPi

incr

ease

Fig. 6. Development Trajectory of Investment induced by GDP of 37 Countries (2009).

Xex

MPiy

lnX

Group A: vicious cycle

Group B & C: virtuous cycle

12

Sweden (5.65)＊＊Singapore (5.64)

＊＊

＊

＊

＊

＊

＊

＊＊

＊

＊

＊

＊

＊

＊＊

＊＊

Denmark (5.54)Switzerland (5.48)USA (5.46)Finland (5.44)Canada (5.36)Netherlands (5.32)Norway (5.22)Iceland, Taiwan (5.20)

Germany (5.16)

Luxemburg (5.02)

Austria, New Zealand (4.94)France (4.99)

Japan (4.89)

Ireland (4.82)

Korea (5.14)

Belgium (4.86)

Malaysia (4.65)

Spain (4.81)

Portugal (4.41)

＊＊＊＊

＊＊＊＊India (4.09)

Czech (4.35)China (04.31)

＊＊Hungary (3.98)

Brazil (3.80)＊

＊

＊

Slovak (3.86)

Indonesia (3.72)Turkey (3.68)

＊

＊＊

＊＊

Mexico (3.61)Russia (3.58)

Philippines (3.51)

Poland (3.74)

)746.0060.1( exp Iu

MPIYav

Sweden (5.65)＊＊Singapore (5.64)

＊＊

＊

＊

＊

＊

＊

＊＊

＊

＊

＊

＊

＊

＊＊

＊＊

Denmark (5.54)Switzerland (5.48)USA (5.46)Finland (5.44)Canada (5.36)Netherlands (5.32)Norway (5.22)Iceland, Taiwan (5.20)

Germany (5.16)

Luxemburg (5.02)

Austria, New Zealand (4.94)France (4.99)

Japan (4.89)

Ireland (4.82)

Korea (5.14)

Belgium (4.86)

Malaysia (4.65)

Spain (4.81)

Portugal (4.41)

＊＊＊＊

＊＊＊＊India (4.09)

Czech (4.35)China (04.31)

＊＊Hungary (3.98)

Brazil (3.80)＊

＊

＊

Slovak (3.86)

Indonesia (3.72)Turkey (3.68)

＊

＊＊

＊＊

Mexico (3.61)Russia (3.58)

Philippines (3.51)

Poland (3.74)

)746.0060.1( exp Iu

MPIYav

Australia (5.06)

Italy (4.77)

Greece (4.64)Thailand (3.97)

UK (5.17)

MPI Increase

ICT Increase704.0I1

),( IXFV

Y

VV

IYFD

IVVV

IYFD

Growing economies

YI

IY

XX

XV

VI

IV

XX

XV

VV

)11()1(FD

aYYYaY

IY

Advanced economies

Fig. 7. ICT Driven Growth Trajectory in 40 Countries (2009).

Global co-evolution with emerging economiesMOP (Middle of the pyramid), BOP (Bottom of the pyramid)

Frugal engineering

Marginal Productivity of IC

T

Advancement of ICT

Case 2: Advancement of ICT vs Marginal Productivity of ICT

13

1. National strategy and socio ‐ economic system

2.1 Strategy and Business model2.1.1 Vision and Business strategy

2.1.2 Business model and Market policy

2.1.3 R&D and ICT

2.2 Employment, Promotion and Training2.2.1 Appointment

2.2.2 Promotion

2.2.3 Training

1.1 National strategy

1.1.1 Democracy

1.1.2 Constitution, Law, Regulation, Standard, Manner

1.1.3 Separation of the three powers of Administration, Legislation, and Judicature

1.2 Social system1. 2.1 Education system

1.2.2 Employment system

1.2.3 Infrastructure investment

1.3 Economic system1.3.1 GDP and GDP per capita

1.3.2 Trade‐ based nation, Export and Import

1.3.3 Tech‐ based nation, ICT and Government ICT

2.3 Structure2.3.1 Entrepreneurial organization

2.3.2Affiliated firms

2.3.3 Foreign capital

2.4 Doctrine, Philosophy and Ethics

2.4.1 Business doctrine and Culture

2.4.2 Philosophy and Ethics

2.4.3 Corporate governance



2.1.3 R&D and ICT


2.2.2 Promotion

2.2.3 Training

3. Historical perspectives

2. Entrepreneurial organization and culture

3.1 Geographical structure3.1.1 Geopolitical environment

3.1.2 Population

3.1.3 Homogeneous/Heterogeneous, Gini index

3.2 Culture and Tradition3.2.1 Culture, Custom and Common idea

3.2.2 National spirit, Moral ethic, Manners and Customs

3.2.3 Religion

3.3 State of development3.3.1 Rapid economic growth

3.3.2 Mature economy

3.3.3 Diminishing population and Aging trend

3.4 Paradigm and phase of industrial society3.4.1 Indust . society, Inform. society, Post‐inform. society

3.4.2 Heavy and chemical industrial structure

3.4.3 Knowledge‐ intensified industr ial structure


3.1.2 Population

3.1.3 Homogeneous/Heterogeneous,



3.2.3 Religion








1.1.1 Democracy
















3 Dimensions of InstitutionsHistorical perspectives

National strategy and socio‐ economic system

Entrepreneurial organization andculture

1. National strategy and socio ‐ economic system



2.1.3 R&D and ICT


2.2.2 Promotion

2.2.3 Training


1.1.1 Democracy


















2.1.3 R&D and ICT


2.2.2 Promotion

2.2.3 Training

3. Historical perspectives

2. Entrepreneurial organization and culture


3.1.2 Population




3.2.3 Religion








3.1.2 Population




3.2.3 Religion








3.1.2 Population




3.2.3 Religion








3.1.2 Population




3.2.3 Religion








1.1.1 Democracy
















3 Dimensions of InstitutionsHistorical perspectives



Historical perspectives



World level

National level

Regional level

Industry/Sector level

Firm/Acad./Rese. Inst level

Specific Tech. level Coevolution between hieralchy

ConsumersInnovatorsInventors

Invention becomes innovation when the invention becomes a commercial product, which in turn becomes a successful product when consumers by it.

1. Each respective level of institutions incorporate threedimensional structure.

2. In order for the property emerge,three dimensional components atthe lower level should interact with those at the higher level.

Three Dimensional Constitution and Hierarchical Structure of Institutional Systems

Sources leading to bi-polarization- Institutional innovation

(2) Bass Model (F.M. Bass, 1969) – Diffusion with innovator (leader) and imitator (follower) dynamism

14

)()()( tQtPdt

tdY where P(t): innovator, and Q(t): imitator.

))(1())(()(NtYpNtYNptP

))(1)(()(N

tYtqYtQ

where p: innovation coefficient.

where q: imitation coefficient

))(1))((()()()(N

tYtqYpNtQtPdt

tdY (25)

tqppq

tqp

eeNtY )(

)(

1]1[)(

(26)

taat ebNb

beN

''1'

1

(26’)

a = p + q, a’ = -(p + q), b =q/p, b’ = p/q

Diffusion in open innovation

1) Model Structure4.1 (1)1. Epidemic disease diffuses

proportional to potential capacity of dissemination

3. Develop this balance

)(' YNa

)1(NYaY

dtdY

tqp

tqp

e

epq

)(

)(

1

1

tqp

tqp

epq

eNY

)(

)(

1

)1(FD 0

/

pdqdFD q/p increase FD increase

yxyFD

11q/p x and e-(p+q)t y

（Open innovation）

2

2

2 1

1

11

1

y

yydxdyx

ydxdyxy

y

xydxdy

dxdFD

ytxx

dxdpp

yy

dxdFD 111

1 2

ytx

dxdptxp

yy 2

2 1111

tttqp xpqpey ]11[]1[][ txp 11Since e-(p+q) = y-t and p+q 1,

011

2

2

dxdpt

yxy

dxdFD

0dxdp

p: Innovafor (leader)q: Immitator (follower)

2) Self-propagating Dynamism

15Fig. 8. Declining Nature of FD and Its Sustaining Efforts.

FDthrough

Innofusionby means of

Effective utilization of external resourcesbased on

Coopetition ( q/p increase)

Earlier emergence of FD

Y: production of innovative products, N: carrying capacity

3) Sustainable FD

when

pq increase FD increase

1

FD

T

11 q,p22 q,p

2211 // pqpq

1

FD

T

11 q,p22 q,p

2211 // pqpq

FD: Functionality development (Ability to improve performance of production processes, goods and services by means of innovation)

Home PC 1982-1988 Cellular telephone 1986-1996

Black & White TV Colour TV Video Cassette Recorder 1981-1994

CD player 1986-1996

Speed of technology introduction: The Bass model of technology diffusion

"Diffusion models: Managerial applications and software" (1999), G. Lilien and C. Van den Bulte, Institute for the Study of Business Markets (University of Pennsylvania), ISBN Report 7-1999. Introduction to the Bass Model and its extensions together with several examples of innovation and imitation factors for particular technologies.

http://isbm.smeal.psu.edu/library/working-paper-articles/1999-working-papers/07-1999-diffusion-models.pdf 16

Application of BLF1. Monthly diffusion trajectory of Japan’s mobile phones (MP) over the last decade can be traced by

the bi-logistic growth model.2. This suggests that Japan’s MP diffusion in the last decade was initiated by two waves Y1 and Y2 .

Table 1 Estimation of Japan’s Mobile Phones Diffusion by the Bi-logistic Growth Model

(January 1996-December 2006)

N1 a1 b1 N2 a2 b2 adj. R2

Parameter 35.147 0.074 5.198 65.418 0.036 14.028 0.999

t-value 2.25 4.59 3.26 3.81 6.74 1.33

tata ebN

ebNYYY

212

2

1

121 11

Y(t): cumulative number of MP diffusion at time t;N1, N2: carrying capacities; a1, a2: velocity of diffusion;b1, b2: initial stage of diffusion; and t: time trend by month (Dec. 95 =0, Jan. 96 =1).

IP: Internet Protocol Service

tata ebN

ebNtYtYtY

212

2

1

121 11

)()()(

(27)

(3) Bi-Logistic Growth Function (BLF) – Diffusion of two co-existing innovation (eg. carriage and car)

17

Table 2 Diffusion Parameters in Major Innovative Goods and ServicesN p q adj. R2

Printer aLLBP

(1975-1994)1581

(19.33)5.4310-3

(15.13)5.810-2

(9.94)0.999

LBP/BJ(1987-2005)

97205(166.57)

1.4710-3

(2.27)2.910-2

(37.96)0.999

MP(1990-2006)

MP 138216

(149.45)0.1210-1

(5358.9)0.5810-1

(2616.7)

MP 265741

(170.24)0.2210-2

(1270.1)0.3510-1

(438.3)

LCD(2000-2008)

LCD 1 2.4103

(1654.3)0.310-2

(1654.3)0.210-1

(1654.3)0.999

LCD 2 2.4103

(656.1)0.410-4

(1654.3)0.810-1

(1654.3)

Web(1993-2006)

Web 1.0 2.42105

(145.87)1.3810-5

(8.35)1.0810-1

(58.33)0.999

Web 2.0 2.49105

(75.66)0.2510-5

(2.60)0.5510-1

(22.74)

a Since the period of co-existence of LLDP and LBP/BJ was limited, simple Bas model was used for respective innovation.b Figures in parentheses indicate t-value. All demonstrates statically significant at the 5% level.c LLBP: Large-scale Laser Beam Printer; LBP: Laser Beam Printer; BJ: Bubble Jet Printer; LCD: Liquid Crystal Display; MP: Mobile phone; and Web: Internet dependency

based on the number of co.jp domains.d Sky Walker: Triggered e-mail transmission by mobile phone; RSS 2.0 (Really Simple Syndication): Triggered publishing updated works in a standardized form as blog and video;

NGPVs (Next Generation PV System): Triggered acceleration of customers initiative in PV development and introduction by means of highly advanced next generation technology.

(4) Bi-Bass Model

PV(1976-2007)

PV 1 0.50105

(8.81)19.3610-5

(3.87)2.6610-1

(45.22)0.999

PV 2 12.71105

(8.82)0.0410-5

(5.72)4.1110-1

(47.89)

x =q/p

10.7

19.3

5.0

15.6

2.59

-0.24

0.60

xdpd

lnln

7.3

7.8103

0.1104

105.4104

-0.87

-0.83

-0.92

1.9103

22.0103 -0.89

-0.83

-0.35

0.03

Sky Walker(1997/10)

RSS 2.0(2003/7)

NGPVs(2006)

0.999

0.999

Trigger of new innovation

tqp

tqp

tqp

tqp

epq

eN

epq

eNtY)(

2

2

)(2

)(

1

1

)(1

22

22

11

11

1

)1(

1

)1()(

– Diffusion of successive innovation with leader and follower dynamism

(28)

18

(5) Logistic Growth Function within a Dynamic Carrying Capacity (LGFDCC)- Diffusion with dynamically developing carrying capacity

19

atbetNtY

1

)()( (24’)

tak

kkeb

NtN

1)( where Nk : ultimate carrying capacity (29)

tabatk

taaabaat

kk

aka

kk

k

k ebeN

ebeNtY

1

11)( (30)

20

Application of LGFDCC1) Features Differences between Manufacturing Technology and IT

1980s 1990sParadigm Industrial society Information society

Core technology Manufacturing technology (MT) IT1. Optimization Within firms/Organizations In the market2. Key features formation process Provided by suppliers Formed through the interacting with institutions3. Fundamental nature As given Self-propagating

4. Actors forming features Individual firms/organizations Institutions as a whole5. Objectives Productivity Functionality 6. Development trajectory Growth oriented trajectory Functionality development initiated trajectory

Table 3 Comparison of Features between Manufacturing Technology and IT

Fig. 10. Diffusion Trajectories in Japan’s Fixed and Mobile Phones.

Num

ber o

f sub

scrib

ers

10000

12000

(in 1

0 th

ousa

nd)

Mobile phone

Fixed phone

19961998

1991

93.8 mil.

60.0 mil.6000

8000

2000

4000

(IT)

(MT)

21Fig. 11. Self-propagating Dynamism in Functionality Development of Japan’s Mobile Phones.

ta

k

kat

k

keaa

bbe

NY

/11

Ultimate carrying capacity 9380

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

(year)

(in 1

0 t

housa

nd)

Mobile phone estimates

Mobile phone observations

Carrying capacity

In mobile driven innovation, new functionality emerged in a self-propagating way in a process of diffusion, not at development stage, as from talk to see, see & talk, take a picture, pay and watch.

2) Self-propagating Functionality Development through Innofusion

Parameter Estimate t-value adj. R2

a 1.4810-1 8.94 0.998b 1.44102 9.85ak 3.4310-2 7.63bk 4.21100 8.65Nk 9.38103 10.89

)1(1

)1(k

kat NNNa

dtdN

beNY

NYaY

dtdY

Message exchange

Communication

e-mail

IP

GPS

Music distribution

1968 1980 1999 2001 2005

Camera

TV phone

Network externality

Diffusion

One-seg

Message exchange

Communication

e-mail

IP

GPS

Music distribution

1968 1980 1999 2001 2005

Camera

TV phone

Network externality

Interaction

Diffusion

Functionalitydevelopment

One-seg

Network externality

Self-propagating mechanismDiffusion of IT

Interaction withinstitutional

system

Network externality

Functionality development

Enhancement of carrying capacity)

Acceleration and advancement of IT diffusion

Talk See See & talk Take a picture Pay Watch

2003

3) Contrasting Diffusion Trajectories

)(NYaY

dtdY

1

NK a b aK bK adj. R2

9.38103

(10.89)1.4810-1

(8.94)1.44102

(9.85)3.4310-2

(7.63)4.21100

(8.65)0.998

N dynamic

SLG

LGDCC

I T

M T atbeNY

1

N fixed

N a b adj. R2

6.00103

(5.07)1.7410-1

(28.20)0.64102

(22.96)0.952

Y: Production of innovative goodsN: Carrying capacity (Upper ceiling)a: Velocity of diffusion

MT: Manufacturing technologyIT: Information technology

SLG: Simple logistic growthLGDCC: Logistic growth within dynamic carrying capacity

KNNaN

dtdN 1

0

2000

4000

6000

8000

(in 1

0 th

ousa

nd)

Fixed phoneN

1996

Y(t)

N

Y

t

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

(in 1

0 th

ousa

nd)

Mobile phone estimatesCarrying capacity

KN

tN tY

Y

t

ta

k

kat

k

kea/a

bbe

NY

11

10000

12000

(in 1

0 th

ousa

nd)

Mobile phone

Fixed phone1996

1998

1991

93.8 mil.

60.0 mil.6000

8000

2000

4000

10000

12000

1953 1956

1959

1962

1965

1968 1971 1974

1977

1980

1983 1986

1989 1992

1995 1998 2001

2004

2007 2010 2013

(in 1

0 th

ousa

nd)

Mobile phone

Fixed phone1996

1998

1991

93.8 mil.

60.0 mil.6000

8000

2000

4000

10000

12000

1953 1956

1959

1962

1965

1968 1971 1974

1977

1980

1983 1986

1989 1992

1995 1998 2001

2004

2007 2010 2013

(in 1

0 th

ousa

nd)

Mobile phone

Fixed phone1996

1998

1991

93.8 mil.

60.0 mil.6000

8000

2000

4000

Fig. 12. Self-propagating FD Dynamics.22

Efforts to prolong higher level of FD Sustainable FD Self-propagating FD

atbe 1

23

1. Diffusion trajectory can be depicted by an epidemic function

2. Functionality development (FD)

3. Measurement of FD

4. Functionality development strategy

where Y: Production of innovative goods; N: Carrying capacity; and a: Velocity of diffusion.

3.4 Functionality Development (1) Functionality Development Concept

Y continues to diffuse as far as it incorporates

“Ability to improve performance of production processes, goods and services by means of innovation” = FD

Y terminates to diffuse when it reaches N

(i) Y N (obsolescent stage of FD)0dtdY

(ii) FD can be defined as “Potential capacity before reaching obsolescent stage”

(iii) Degree of FD = N/Y

Functionality development can be depicted by the following diffusion trajectory

Declining nature

atbeNY

1

)NY(aY

dtdY

1

24Fig. 13. Scheme of the Breeding Stage of New Ventures.

Basic research Development Market growthInitial marketing

IPO

Death valley

The sea of Darwin

Start-up

10090

70

50

30

(%)

Later stage

Initial stage

Diffusion ratio 16% line

Emergence of FD

Frequency distribution curve

Accumulative distribution curve

CHASM

Innovator, Early minority (16%)

Early majority (34%) Late majority (34%) Latest (16%)

10

Takeoff period

IPOCHASM: Deep trench

compelling newventures start-up(Moore, 1991).

Timing of FD emergence corresponds to that ofCHASM.

Emergence of FD Overcoming CHASM Take off leading to IPO

FD emergence contributes to overcome CHASM thereby enables IPO (Initial Public Offering) accomplishment.

(2) Timing of Functionality Development Emergence

1) Timing in Overcoming CHASM

25Fig. 14. Level and Timing of Inflection in Diffusion Trajectory.

2) Timing of Functionality Development Emergence by Logistic Growth Function

atbeNY

1

Diffusion model ,

0 t1 t# tt2

Increase in diffusion velocity Decrease in diffusion velocity

accelerate decelerate accelerate decelerate

where t1: inflection point of diffusion velocity in its increasing period; t# : inflection point of diffusion; and t2: inflection point of diffusion velocity in its decreasing period.

Timing of FD emergence

]1

13332[]

1[][ 11

1

000 bN

beNYdS

dtdYS t

attt

Given the initial level of diffusion 1/(1+b) = 0.05, 16.0

NS

CHASM

1. Following Rogers, Mahajan and Moore, timing of FD emergence can be identified as follows:

atbeYNFD 1

2. This time corresponds to CHASM.

2t#t0

a

b32ln abln

dtdY

2

2

dtYd

a

b32ln

t

t

Y)33/( N

2/N

N

)1/(1 b

S

Emergence of FD

t１

Timing of FD emergence

)33/( N

FD

33

233

t

Emergence of FD

t

16% 16%34%34%

Timing of FD emergence corresponds to the timing that maximizes the secondary derivative of the diffusion trajectory, and its level is

(Rogers, Mahajan, Moore)33

26

3) Timing of Functionality Development Emergence by Bass Model

t1t2

3

3

dtYd

t

t

t

t

FD

])32(1[

33

qp

t

qp

qpt 32ln11

qp

qpt

321ln1

2

2

2

dtYd

dtdY

Y

tqp

tqp

e

epq

)(

)(

1

1

tqp

tqp

epq

eNY

)(

)(

1

)1(

FD

Emergence of FD

Shifting to the successive FD

Emergence of FDFig. 15. Timing of the Emergence of Functionality Development.

FD

1t t1

Earlier Emergence of FD

1t FD

27

3.5 Integration of Production Function and Diffusion Function - Innofusion(1) Production Diffusion Integration

As paradigm shifts to an information society, spot where innovation takes place shifts from production site to diffusion processleading to the significance of production diffusion integration: innofusion function.

(i) Production Function

(ii) Diffusion Function (Cumulative Y diffuses as a function of T in high-tech. firms)

(iii) Production Diffusion Integration － Innofusion Function in which Functionality Development is only an option.

),( TXFY YR

TY

XX

YX

XY

YY

FDaY

NYaY

TY 111

Traditional production factors TFP

MPT R&D intensity


Growth rate

Growth

Traditional production factors(Labor, Capital)

TFP Marginal productivity of technology (MPT)

R&D intensity


Diffusion velocityGDP

Growth

Traditional production factors(Labor, Capital)

TFP Marginal productivity of technology (MPT)

R&D intensity


Diffusion velocityGDP

Fig. 16. Integration of Production and Diffusion Functions.

YR

TY

XX

YX

XY

YY

Ability to improve performance of production processes, goods and services by means of innovation

FD

TY aY (1 )1

YNFD

27

increase rate

(R/Y)Production (Y)

(a)

(FD)

Y: Prod. of innovative goodsX: L(Labor), K(Capital),

M(Material), E(Energy)T: TechnologyR: R&D investmentFD: Functionality developmentN: Carrying capacity a: Diffusion velocityb: Diffusion at the initial stage

YN

(31) In high-tech. firms, T is proportional to tand (see next page)

TY

dTdY

(2) Selection of Development Trajectory

)11()1()1(

)1(

,,

FDYa

NYYa

NYYa

TY

NYaY

TY

dTdT

TY

dTdY

dtdT

dTdY

dtdY

tT

Source: Watanabe et al. (2003).Fig. 17. Scheme Leading Japan to Lose Its Institutional Elasticity.

Economic growth dependent model: Depend on Y- Growth Oriented Trajectory

New functionality development model: Stimulate FD - New Functionality Development Initiated Trajectory

Japan (1980s, 1990s) US (1980s) US (1990s)

System conflict System match

MPT: Marginal productivity of technology

Y: Prod. of innov. goods L: Labor, K: CapitalM: Material, E: EnergyT: Technology stockR: R&D investmentFD: Functionality

developmenta: Diffusion velocity

（MPT）

Contribution to labor and capital

Marginal productivity of technology

TFP growth rate (technological progress)

Marginal productivity of technology

EMKLX YR

TY

XX

YX

XY

YY

,,,

FDaY

TY 11

EMKLX YR

TY

XX

YX

XY

YY

,,,

)FD

(aYTY 11

Y: Production of innovative goods;X: Labor, Capital, material and energy T: Technology stock

28

,T)(RT tmtt 11 )g/(RT m 10

),g/(RT )m(tt 1 )g/(RT mtt 1

.1)1(10 g

ggRRg

RT mtmtmtt

tgggtmgRTt )/()1)(1(0

ggmgR

1)1(0 gggR

1

0

Tt: technology stock at time t; Rt: R&D investment at time t; m: time lag between R&D and commercialization; : rate of obsolescence of technology; and g: growth rate of R&D investment at the initial period.

)())((),( TFTXFTXFY

NY

Yadt

Yd1

NYaY

dtdY 1

Production function in high-tech firms can be depicted as follows:

(3) US Locomotive Leveraging Transformation to FD Initiated Trajectory

Fig. 18. Trends in the US Policy for Enhancement of Competitiveness (1980-2004).

Young Report (1985) Global Competition: The New Reality

New Young Report (1987) America’s Competitive Crisis:Confronting the New Reality

29

InformationHighway (93)

1. System conflict led to an institutional less-elasticity in an information society resulting in a dramatic decrease in MPT.2. MPT decrease led to TFP decrease resulting in a decrease in innovation contribution to growth.3. Thus, co-evolution changed to disengagement in an information society. MPT: Marginal Productivity of Technology

(i) Dramatic Decrease in Marginal Productivity of Technology TFP: Total Factor Productivity

Inst

itutio

nal e

last

icity

(ii) Consequent Decrease in Innovation

USA System match

Fig. 6-2. Marginal Productivity of Manufacturing Technology(1975-1999) - Index: 1990 =1.

V = F (L, K, T)

L: labor, K: capital, T: technology stock

V/

MPT

( ∂∂

T)

0.65

0.75

0.85

0.95

1.05

1.15

1.25

1.35

1.45

1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999

USA

Japan0.65

0.75

0.85

0.95

1.05

1.15

1.25

1.35

1.45

1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999

1990

MPT

Fig. 6-1. Institutional Elasticity of Manufacturing Technology - Elasticity of the Shift to an Information Society to Marginal Productivity

of Technology （1980-1999） - Index：1990＝100.

19901980 1999

∂ln

MPT

/∂

lnD

x

Industrial society Information society

Japan System conflict

GDP

Fig. 6-5. Marginal Productivity of Technology(1960-2001)．

Fig. 6-3. TFP Growth Rate (1960-2001)． Fig. 6-4. R&D Intensity (1975-2001).

TFP change rate (TFP/TFP) = R&D intensity (R/V) × Marginal productivity of technology (MPT)Innovation to GDP growth

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1975-1985 1985-1990 1990-1995 1995-2001

Mar

gina

l pro

duct

ivity

of t

echn

olog

y-1.0

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

１９６０－１９７5 １９７５－１９８５１９８５－１９９０１９９０－１９９５１９９５－２００１

Gro

wth

rate（％

p.a.）

2.0

2.2

2.4

2.6

2.8

3.0

3.2

1975-1985 1985-1990 1990-1995 1995-2001

Rat

io %

JP US JP US

JP US6.2

1.5 1.41.0

2.8

0.9-0.3

0.9 0.2

1.5

(4) System Conflict in an Information Society

30

31

(5) Swell of Japan’s Institutional MOT toward a Post-information SocietyAs a consequence of hybrid management fusing “East” and “West”, Japan’s indigenous MOT is again responding to a co-evolutionary dynamism between innovation and institutional systems corresponds to a ubiquitous economy.

Fig.19 . Swell of Japan’s New Innovation.

Industrial society(Materialization economy)

Information society(Digital economy)

Post-information society(Ubiquitous economy)

Manufacturing technologybased growth oriented trajectory

IT driven new functionalitydevelopment initiated trajectory

Service oriented manufacturing

Product oriented Function oriented Solution oriented

a. Seamless,(1) Increasing digitalization of manufacturing process,

(2) Advanced digital infrastructure or alliance, and

(3) Timely correspondence to the customer’s potential desire in the digital economy

1 Just in time cell production(synchronization of JIT and cell production system),

2 Super-flexible manufacturing system,

3 Full automation manufacturing ,

4 Co-evolutionary development between suppliers and customers,

5 Broad integrated consortia business,

6 On-demand manufacturing(full variety of on-demand products production), and

7 “J-sense” (sense of Japan) goods and system.

b. All actors participation and cooperation, and

c. On demand

d. Open-sourcing

e. Long tail

32

(6) Governing Factors to Functionality DevelopmentUnder the competitive circumstance where firms aim at maximizing their profits, equation (31) should be equivalent to relative prices as follows:

Y

TP

PPTY

(32)

P TP YPwhere : relative prices of technology; : technology prices of innovative goods; and : prices of innov. goods.

Equation (31) can be developed as follows:

PN

YaaYNYaY

TY

TY

2

)1( (33)

where dtdYY .

Differentiate equation (32) by time t,

)21()21(2FD

TaPNYYa

NYYaYaP

)FD

(TaPP 21

(35)

where a: diffusion velocity; and : price elasticity to technology (PET).

aTTP

aTTTPP

aTPP

Ta

FD 11

2

lnln11

211

211

2

Functionality development (FD) can be depicted as follows:

(34)

TP

PT

PT

lnln

FDaY

NYaY

TY 111 (31)

33

(7) Prices of Technology and Functionality Development

.

(35)

: price elasticity to technology (PET).

aTTP

aTTTPP

aTPP

Ta

FD 11

2

lnln11

211

211

2

TP

PT

PT

lnln

2 0 21 FDFD

T

YT PP

0 0

)0 ( )0 (

Pric

es

Technology

Products

Functionality-seeking

Mass-production

Technology stock

YYTT PPPPTT

PPTT

PT

///

//

lnln

0 when YYTT PPPP //

34

0

100

200

300

400

500

600

700

800

900

1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998

Pric

es o

f prin

ters

(PV

)

0

500

1000

1500

2000

2500

3000

3500

4000

1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998

Pric

es o

f prin

ters

tech

nolo

gy (P

T)

Trends in Prices of Printers Corresponding to their Technology Stock in Canon Printers (1976 – 1998).

Trends in Prices of Printers Technology Corresponding to their Technology Stock in Canon Printers (1976 – 1998).

LLBP (1976-1985):

LBP/BJ (1985-1998): .

0 lnln

TPV 0 ln

TPT

0 ln

TPT

LLBP

LBP/BJ

Pric

es o

f prin

ters

(Pv)

Pric

es o

f prin

ters

tech

nolo

gy (P

r)

LLBP

LBP/BJ

0// YYTT PPPP

0// YYTT PPPP

0

00 < FD < 2

2 < FD

35

4.6 Sustainable FD in Open Innovation (1) Timing of the Functionality Development

t1t2

3

3

dtYd

t

t

t

t

FD

])32(1[

33

qp

t

qp

qpt 32ln11

qp

qpt

321ln1

2

2

2

dtYd

dtdY

Y

tqp

tqp

e

epq

)(

)(

1

1

tqp

tqp

epq

eNY

)(

)(

1

)1(

FD

Emergence of FD

Shifting to the successive FD

Emergence of FDFig. 15. Timing of the Emergence of Functionality Development.

FD

1t t1

Earlier Emergence of FD

1t FD

Earlier FD emergenceSustainable FD

36

(2) Requirement for Earlier Functionality Development Emergence

03

3

dt

Yd ]32

ln[32

1ln11

xy

qp

qpt

xyx

dxdy

dxdt

pdqdt

]32

ln[/

11

where .2)]1([

])1[(,

)1(1,

)1(1

xp

pdxdpx

dxdy

xxpxy

xpy

Therefore, can be developed as follows:

)1(

])32(ln[])1[(1

)1(1

)1(1]

)32(ln[

)]1([

])1[(

/ 21

xp

xx

pdxdpx

xpxxpxx

xp

pdxdpx

pdqdt

where and .yqp

xpq

1/

pdqdt

/1

In case when ,1)1(

])32(ln[])1[()(

xp

xx

pdxdpx

xW 0/1 pdq

dt Necessary condition for earlier functionality development emergence

q/p increase induces earlier FD emergence leading to sustainable FD

p: innovatorq: imitator

(3) Boundary Satisfying Earlier FD Emergence

1)1(

])32(ln[])1[(0

/1

xp

xx

pdxdpx

pdqdt

1)1(

])32(ln[

)1(

])32(ln[)1(

xp

xx

p

xp

xxdx

dpx

1)1(

])32(ln[])32(ln[

x

xx

p

xxdx

dp

Aiming at identifying the boundary condition satisfying earlier FD emergence in the Bass model, the following analysis is attempted:

Table 4 Conditions of Boundary Function by Areas

321 x

)1(])32(ln[

1

)1(1

])32(ln[

110/1

xx

xpx

dxdp

xx

xpdx

dppdq

dt

32 x

)1(])32(ln[

1

)1(1

])32(ln[

110/1

xx

xpx

dxdp

xx

xpdx

dppdq

dt

If , then , and so

where : x elasticity to p.

If , then , and we get

0])32(ln[ x

0])32(ln[ x

Area x

I + +

II + -

III - -

IV - -

V - +

dxdp

dxdt1

321 x

32 x

1.11x

321 x

32 x

0

0

01

0

0

-4

-3

-2

-1

0

1

2

3

4

1 4 7 10 13 16 19 22 25 28 31 34 37 40

x (q/p)32

Ⅰ Ⅱ

Ⅲ

ⅣⅤ

x# = 11.1

)(px

dxdp

010

B(x)

B(x)

)1(])32(ln[

1)(x

x

x

xB

32,1 xxwhere

Fig. 20. Areas Satisfying Earlier FD Emergence.

(the crossing point of boundary function)

37

Table 5 Diffusion Parameters in Major Innovative Goods and ServicesN p q adj. R2

Printer aLLBP

(1975-1994)1581

(19.33)5.4310-3

(15.13)5.810-2

(9.94)0.999

LBP/BJ(1987-2005)

97205(166.57)

1.4710-3

(2.27)2.910-2

(37.96)0.999

MP(1990-2006)

MP 138216

(149.45)0.1210-1

(5358.9)0.5810-1

(2616.7)

MP 265741

(170.24)0.2210-2

(1270.1)0.3510-1

(438.3)

LCD(2000-2008)

LCD 1 2.4103

(1654.3)0.310-2

(1654.3)0.210-1

(1654.3)0.999

LCD 2 2.4103

(656.1)0.410-4

(1654.3)0.810-1

(1654.3)

Web(1993-2006)

Web 1.0 2.42105

(145.87)1.3810-5

(8.35)1.0810-1

(58.33)0.999

Web 2.0 2.49105

(75.66)0.2510-5

(2.60)0.5510-1

(22.74)

a Since the period of co-existence of LLDP and LBP/BJ was limited, simple Bas model was used for respective innovation.b Figures in parentheses indicate t-value. All demonstrates statically significant at the 5% level.c LLBP: Large-scale Laser Beam Printer; LBP: Laser Beam Printer; BJ: Bubble Jet Printer; LCD: Liquid Crystal Display; MP: Mobile phone; and Web: Internet dependency

based on the number of co.jp domains.d Sky Walker: Triggered e-mail transmission by mobile phone; RSS 2.0 (Really Simple Syndication): Triggered publishing updated works in a standardized form as blog and video;

NGPVs (Next Generation PV System): Triggered acceleration of customers initiative in PV development and introduction by means of highly advanced next generation technology.

(4) Diffusion Dynamics in Major Innovation

PV(1976-2007)

PV 1 0.50105

(8.81)19.3610-5

(3.87)2.6610-1

(45.22)0.999

PV 2 12.71105

(8.82)0.0410-5

(5.72)4.1110-1

(47.89)

x =q/p

10.7

19.3

5.0

15.6

2.59

-0.24

0.60

xdpd

lnln

7.3

7.8103

0.1104

105.4104

-0.87

-0.83

-0.92

1.9103

22.0103 -0.89

-0.83

-0.35

0.03

Sky Walker(1997/10)

RSS 2.0(2003/7)

NGPVs(2006)

0.999

0.999

Trigger of new innovation

tqp

tqp

tqp

tqp

epq

eN

epq

eNtY)(

2

2

)(2

)(

1

1

)(1

22

22

11

11

1

)1(

1

)1()(

(28)

38

Fig. 21. Sustainable Functionality Development Condition.

1. While latest high-technology products as LBP/BJ, MP 2, LCD 2, Web 1.0, Web 2.0, PV 1 and PV 2 satisfy conditions for sustainable functionality development, LLBP (1976), MP 1 (1996) and LCD 1 () do not satisfy these conditions resulting in being substituted by LBP, BJ, MP 2 and LCD 2.

2. This can be considered as substitution from ‘resistance to innovation’ in the early introduction to market,to supra-functionality with customers own initiative.

(5) Sustainable FD by Major Innovation

-1

-0.5

0

0.5

1

1.5

2

5 10 50 100 500 1000 5000 10000 15000 20000 2500032 x (q/p)

px

dxdp

LLBPx# = 11.1

)1(])32(ln[

1)(x

x

x

xB

MP 1

LCD 1

Supr

a-fu

nctio

nalit

yR

esis

tanc

e to

inno

vatio

n

Sky Walker

LBP/BJ

Web 1.0 Web 2.0

q substitutes for pMP 2

LCD 2

RSS 2.0 NGPVs

PV 2PV 1

)(lnln0 xB

xdpd

px

dxdp

(x = q/p)

p, q dynamism forEarlier emergence of FD

0/1 pdq

dt

39

(6) New FD Frontier

Fig. 22. New Functionality Development Frontier Leading to Supra-functionality.

Explore new FD frontier

Instills customers“exciting story with their initiative as heroes/heroines” and thrills them gratification

Incorporates innovation new social,cultural and aspirational value beyond economic value

Printers

Print office

Mobile phone

Communication Watch

LCD WebPassive

information

PV

Electric power company

Home print Internet IPTV Blog Clean energy

Depend on

only dependent

Purchase from

-1

-0.5

0

0.5

1

1.5

2

5 10 50 100 500 1000 5000 10000 15000 20000 2500032 x (q/p)

px

dxdp

LLBPx# = 11.1

)1(])32(ln[

1)(x

x

x

xB

MP 1

LCD 1

Sup

ra-f

unct

iona

lity

Res

ista

nce

to in

nova

tion

Sky Walker

LBP/BJ

Web 1.0 Web 2.0


LCD 2

RSS 2.0 NGPVs

PV 2PV 1

)(lnln0 xB

xdpd

px

dxdp

(x = q/p)

p, q dynamism for Earlier emergence of FD

0/1 pdq

dt

-1

-0.5

0

0.5

1

1.5

2

5 10 50 100 500 1000 5000 10000 15000 20000 2500032 x (q/p)

px

dxdp

LLBPx# = 11.1

)1(])32(ln[

1)(x

x

x

xB

MP 1

LCD 1

Sup

ra-f

unct

iona

lity

Res

ista

nce

to in

nova

tion

Sky Walker

LBP/BJ

Web 1.0 Web 2.0


LCD 2

RSS 2.0 NGPVs

PV 2PV 1

)(lnln0 xB

xdpd

px

dxdp

)(lnln0 xB

xdpd

px

dxdp

(x = q/p)

p, q dynamism for Earlier emergence of FD

0/1 pdq

dt

by own design

Explore new world

with as preferred as hero/heroines by own interior design

only

multimedia services

interactivity

40

a RSS: Really Simple Syndication; and NGPVS: New Generation Photovoltaic System.

MP[Sky Walker]

Web[RSS 2.0]

PV[NGPVS]

E-mailtransmission

Blog, video, news headline

User orientedPV systems

New FD frontier which instills in users an “exciting story on their own initiatives as heroes/heroines” thrills them with gratification beyond economic value

Only communication

E-mailtransmission

New communication

community

Onlyvisit and watch RSS a

Embraced by Web publisher

Expensive,inconvenient NGPVS a

House designinterior

curtain, wall

41

42

(7) Strategy for Substitution

Table 6 Concept of Imitator Substitutes for Innovator

Customer substitutes for supplier

Open innovation

Coopetiotion (Cooperation and competition)

Co-evolutionary domestication

q (imitator) substitutes for p (innovator) Assimilation of spillover technology

Hybrid management of technology

Knowledge transfer

Sustainable FD

Innovator ImitatorLeader FollowerSupplier CustomerProducer ConsumerFear to new products Gratification of consumption

43

4. Effects of Learning4.1 Learning by Doing

4.2 Theory of Learning by Doing

4.3 Learning Curve

4.4 Market Learning

4.5 Hybrid Management of Technology

4.6 Necessity for Effective Learning

44

),,( TKLFY 1KATL

1KLAe t

,iijj qwpvXY

jj

jjj

ii

iii Xp

Xpv

YqYqw ,

))(,( TKLFY

/)]1([ kf

LKkLYf /,/

)/(/ kkgff

nbccK

LaKY nn

1/],)1(1[ 1/11

nbKKL )(

Solow (1956)

Denison (1962)

Solow/Johansen (1960)

SMAC (1961)

Kaldor (1962)

Arrow (1962)

Solow residuals → Technological

progress

TFP

Technology embody into capital

CES (Constant elasticity of substitution)

Technological progress function

Learning by doing

(1) Chronology to Endogenous Technological Improvement

4. Effects of Learning4.1 Learning by Doing

Fig. 23. Chronology to the Initial Efforts to Endogenous Technological Improvement.

Kenneth J. ArrowAug. 23, 1921 (age 91)United StatesStanford UniversityMacroeconomicsGeneral equilibrium theorySocial choice theory

1972 Novel prize in Economics

Solow, Minhas, Arrow, Chenery

Technical change was assumed to change

exogenously

Why firms innovate?

Endogenous growth theoryInnovation and

technical change are determined endogenously

45

David (1975)

Kaldor (1962)

Arrow (1962)

US technology is more Capital intensive

than UK’s

Habbakuk’s postulate (1962)

Learning

Hicks (1932)

Kennedy (1964) Weisaecker (1966) Phelps (1966)

Induced invensions

Labor saving ← Labor saving bias

Unbalance of tech.

Improve quality

Induce Smaller labor

Diffuse

Growth Resources dependency

Capitalto industry

Laborto agriculture

[Plausible]

1. Unbalance of tehcnology

2. Labor saving bias

3. Resources scarcity inducement

Rosenberg (1976) Binswanger (1978)

[Inplausible]

Technological learning Induced innovation

(2) Learning, Diffusion and Induced Innovation

Fig. 24. Technological Learning, Diffusion and Induced Innovation.

46

(3) Dynamism between Innovation, Diffusion, Learning and Spillover

Innovation

Diffusion

Embed to prod. facts. Disseminate into socio-economy

Labor Capital Economy Society Transfer of labor/capital

Reflect to prod. process Technol. spillover

Productivity improvement

Production increase/growth

Utilizat. of prod. growth

Further innovation investment

Effects of learning

Fig. 26. Innovation, Diffusion, Learning and Spillover Dynamism.

Assimila. of spillover tech.

Effects of learning 1962 Kaldor △(Y/L)/(Y/L) = F(△(K/L)/(K/L)) Technical Progress Function 1962 Arrow L =λ(K) = b K a Learning by doing

Embed to production factors 1964 Nelson (Capital) Vintage Model 1967 Denison (Labor) Quality of Labor

Technology spillover 1979 Griliches Technology Distance 1982 Scherer 1986 Jaffe Proximity

25

47

(4) Contribution of Learning to TFP Growth

Own efforts R&D Technology investment knowledge stock

Dependent Spillover Assimilation technology

Learning effectsEconomies of scaleLabor quality improvement Maturity of capital stockIndustrial structural changeImprovement of managementExternal economy, policy effects

Direct effects

Indirect effects

Technologicalimprovement

Productivity increase

Technological progress

TFP

25 Aug Accumulation of technology knowledge1 Sep Diffusion of technology

15 Sep Effects of learning6 Oct Technology spillover

47

3.0

1.3

-0.6-0.9

Gro

wth

rat

e of

GD

P(%

p.a

.)

1960-1975 1975-1985 1985-1990 1990-1995 1995-2001


Energy crises

High-tech. miracle

Lost decade

9.7

2.23.4

2.0 1.8

6.2

1.42.8

-0.3 0.2

(Rapid economic growth)

48

Fig. 26. Trend in Contribution of Learning to TFP and Consequent GDP Growth Rate in Japan (1960‐2001) － % p.a．

1960‐1973 1975‐1985 1985‐1990 1990‐1995 1995‐2001

GDP (TFP) 9.7 (6.2) 2.2 (1.4) 3.4 (2.8) 2.0 (-0.3) 1.8 (0.2)

Direct effect of R&D investment

1.0 0.2 0.5 0.2 0.3

Indirect effect of R&D investment

2.2 0.4 1.0 0.4 0.5

Learning and spillover effects

3.0 0.8 1.3 - 0.9 - 0.6

Japa

n’s T

FP c

ompo

sitio

n

Source: Watanabe (2005). Japan 0.6 1.1 U S 0.4 0.3

REG

0.8

9.7

2.23.4

2.0 1.8

3.8 3.4 3.22.4

3.9

6.2

1.5 1.4 1.02.8

0.9 -0.30.9 0.2

1.5


JP US

Con

trib

utio

n of

le

arni

ng to

TFP

Con

trib

utio

n of

in

nova

tion

Japan US comparison (growth rate-% p.a.)

GDP

TFP

1960-1975 1975-1985 1985-1990 1990-1995 1995-2001

(5) Japan’s Unique X Efficiency Depending on Learning

MPT

1. Japan accomplished conspicuous X-efficiency during the period of industrial society.

2. Which contributed to increase MPT.3. This can be attributed to Japan’s intensive

learning efforts.

Learning led to high MPT

4.2 Theory of Learning by Doing(1) Learning Incorporated Production Function

Arrow considered requirement of labor decreases as capital (K) increases with a following function:

Based on this analysis Arrow developed the following production function:

(2) Increasing Returns to Scale BehaviorAbove production function demonstrates increasing returns to scale.

e.g., in case n=1 (same as the case ),

While in general production function,(when △K/K > △L/L)

nbKKL )(

)(1111

1

11

bL

n

neaK

cKLaKV

)1( n

bL

eaK 1 )1( n

01

n

bc

(35)

(36)

(37)

1n

)1( /bLeaKV bLeKaV /lnlnln KKebLKKVV bL /)/(// / (when△L > 0 )

KKKKKKKKKKLLVV //)(/////

b: Coefficient > 0, n: Leaning coefficient > 0

a: Production capacity corresponding to K > 0

)()( )/()(,lnlnln,1 bLbL eLLbLKK

VVeKaVnofcaseIn

0)1(

111

nn

)1(

}{1

1

))(1())(1(

}){(1

,)(

)(

))(1(

))(1(

1

11111

1

11

11

111

bL

nbL

bL

n

nn

eaK

eaKV

ecK

LbLn

bLn

Lb

ncK

LbLK

n

n

nnbLn

n

KK

49

50

)1(ln

)1(011

)()(

)0()()(

0,0)()0()(

})({)(

)(,)()(

)()()()(

)()(

)()(

:)(,)(

:)(,)()(

)(),,(

0

0 0

0 0

0

0

11

0

1

100

0

0

nKbdKKb

nn

bccKKn

bdKbKKK

KaKadKKK

nbbKKaaK

LKKV

LKKLKK

KKVKKL

dKKK

dKKK

KdKKV

KdKKKLL

KLLKLFV

K

K

nnK

K

K

K

n

K

K

K

K

n

K

K

K

K

nnnn

cLcK

cLKLKLK

cLKK

cK n

1

111

11

11

}{})({})({,)(}))(({,)(1

1

)1(}1{1}{1})({}{ 11

1

11

11

11

11

111

ncK

LaKcK

LaKaKcK

LcKKaaKc

LcKaaKV nn

n

nnnn

nnn

)1(

,})({,)()(,)(ln

1

)()(

)()(

bL

bL

bL

bK

bLK

bLK

bLK

bK

eaKeaeaKeaeaKaeaKY

eLKLKeKe

bKb

12

3

4

5

6789

Number of labor

Production capacity

From 6

Substitute 8 for 7Provided that 10

11

12

13

14

From 13

Substitute 15 for 9

When n = 1, from 14,

Substitute 17 for 9

15

16

1718

Appendix: Mathematical Development of Arrow’s Learning Incorporated Production Function

0,)(),(

nbbKKL

KLFVn

)1(1

01

)1()1(1 11

1

neaK

nbcn

cKLaKV

bL

nn

KK

VV

Increasing returns to scale

Learning effects

KforLKsubstitutetheninKforcKSubstitute

functionreverse

n

)(13)/(

:

11

1

51

(1) Learning CoefficientIncorporation of the effects of learning in production function can be examined bylearning curve (or experience curve) that illustrates correlation betweencumulative production and prices.

Asuume that prices of product P and cumulative production Y* (=ΣY), learning curve can be depicted by the following equation: P = A・Y*-λ

where A: scale factor, λ: learning coefficient.

Given that P = P(Y*),

LR = - [P(2Y*) - P(Y*)]/P(Y*) Price declines by LR % for each doubling of cumulative production

PR = P(2Y*)/P(Y*) = 2 -λ = 1 - LR Speed of learning

P(2Y*)/P(Y*) = A2Y*-λ/AY*-λ= 2-λ= 1-LR

(3) Progress Ratio (PR)

(2) Learning Rate (LR)

4.3 Learning Curve

(38)

(39)

(40)

52

― 日本の産業の太陽電池開発における学習効果（1980-1995）

0 10 20 30 40 50 60 70 80

-20 80 180 280 380

1980

851990 95 98

79

99

累積太陽電池生産量（CMSCP）

太陽電池生産価格（PSC

）

0

1

2

3

4

5

-

2

0 2 4 6

198

0 8

5199

0 99

79 log (PSC) = 3.65 - 0.34 log (CMSCP) adj.R2

DW (-38.39) 0.987 1.18

累積太陽電池生産量（CMSCP）

太陽電池生産価格（PSC

）

Fig. 27. Learning Effects of PV Development in Japan’s Firms (1979-1999).

ln PSC = 3.65 – 0.34 ln CSCP 0.987 DW 1.18(-38.39)

(4) Measurement of Learning Effects

2.Radj

CumulativeCumulative solar cell production (CSCP)

(-38.39)

Cumulative solar cell production ( ln CSCP)

Pric

es o

f sol

ar c

ell (

PSC

)Pr

ices

of s

olar

cel

l (ln

PSC

)

1980

1990

2

53

表 1 日本の太陽電池生産代表企業の学習効果係数 (1976－1990)

Model: PSC = A*CMSCP (t 値) 2. R adj DW 産業全体 -0.35 (-22.80) 0.981 1.42 1976－1990

[ -0.37 (-73.88) 0.997 1.60 1976－1995 ]

PV 企業 A -0.12 (-28.31) 0.988 2.09 1976－1990

B -0.25 (-25.92) 0.988 2.63 1976－1990

C -0.29 (-7.47) 0.871 1.49 1976－1990

D -0.69 (-18.99) 0.978 2.53 1981－1990

E -0.41 (-9.34) 0.911 1.77 1976－1990

F -0.31 (-14.46) 0.962 1.52 1976－1990

PSC: 太陽電池生産価格（実質価格），CMSCP: 累積太陽電池生産量 A~F には次の太陽電池生産リーディング企業を含む．三洋電機，京セラ，シャープ，鐘淵化学，富士電機，日立

Table 7 Learning Coefficients in Japan’s Leading PV Producing Firms (1976-1990)

CSCPAPSC t-value

PV firms

Whole industry

PSC: Prices of solar cell (fixed prices), CSCP: Cumulative solar cell production.A-F: leading PV firms including Sanyo, Kyocella, Sharp, Kanegafuchi chemicals,

Fuji electrics and Hitachi.

54

PSC: Prices of solar cell

SCP: Production of solar cell

PVR: PV R&D by industry

Decline in solar cell production prices

Increase in solar cell production

Inducement of industry R&D

PV development initiated by MITI’s Sunshine Project

Increase in PV technology stock

*4 *5

*3

*6

*2

*1

SSPV: R&D budget in the Sunshine Project

TPV: Technology stock of PV

adj.R2 DW

*1 log(PVR) = 1.37 + 0.77 log(SSPV -1) 0.979 1.31 (29.35) *2 TPVt = PVRt-m + (1-ρ)TPVt-1 m = 2.8 年, ρ= 20.3% *3 log(SCP) = -8.26 + 2.19 log(TPV) + 5.94 log(Pey) 0.985 1.75 (35.48) (13.60) *4 log (PSC) = 11.97 - 0.97 log(TPV) - 1.04 log(Pey) 0.988 1.53 (-39.51) (-5.59)

Effects of technology stock and energy prices log (PSC) = 130.87 - 0.06 Year - 0.31 log(SCP) 0.995 1.38 (-9.32) (-18.87)

Effects of learning and economies of scale *5 log(SCP) = 16.97 - 2.02 log(PSC -1) + 2.98 log(Pey -1) 0.995 1.25 (-58.73) (13.61) *6 log(PVR) = 3.59 + 0.43 log(SCP) 0.973 1.03 (25.96)

Fig. 28. Virtuous Cycle of Japan’s PV Development (1976-1995).

Pey: Relative energy prices

(39)

4.4 Market Learning(1) Dynamic Learning Coefficient

Learning function according to diffusion of innovative good Y YAP (38’)

P: fixed prices of PCs; A: scale coefficient; Y : cumulative production; : learning coefficient.

These dynamic learning coefficients 1, 2, … n are learning results regarding as effects of “marketing learning” during a serial process of production, distribution and usage.

Learning coefficients in the market learning are a series of continual coefficients during production, distribution and usage phases when they are diffused, these coefficients are expressed as function of t as:

in

iin tattttt

0

321 ))(,),(),(),(()( (39)

bi is the ith coefficient of column t.

YtaAP in

ii lnlnln

0

is disturbance column which is independent from (t) .(t) can be computed by taking differentiation of equation (38’-2) by lnY.

(38’-2)

n

i

iita

YP

0lnln

55

56

(2) Dynamic Learning Coefficient of PCs: An Empirical AnalysisDynamic learning coefficient of China’s PC is measured by the following equation with the usage of fixed prices of PCs:

lnA a0 a1 a2 a3 a4 a5 d adj. R2 DW AIC

n = 3

11.79 (3.37)

0.37 (3.63)

-0.029 (-1.70)

3.2*10-3

(2.79) -8.7*10-5

(3.35) 0.30

(3.37)0.996 1.38 -51.79

n = 4

12.01 (122.99)

0.84 (5.35)

-0.17 (-3.98)

0.020 (4.07)

-9.3*10-4

(-3.82) 1.5*10-5

(3.47) 0.27

(3.93)0.997 2.03 -62.76

n = 5

11.89 (85.61)

-0.50 (1.48)

0.033 (0.26)

-3.3*10-3

(0.15) -9.5*10-4

(-0.56) 5.7*10-5

(0.88) 1.0*10-6

(1.12) 0.26

(3.85)0.997 1.29 -62.65

Table 8 Comparison of Dynamic Learning Coefficient of China’s PCs (1982-2002)

0.2

0.3

0.4

0.5

0.6

0.7

0.8

1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002

Dyn

amic

lear

ning

coe

ffic

ient

(λ)

Import parts dependence period

Market expansion period

Real Competitive Market Period

45342 10*5.110*3.9020.017.084.0)( ttttt

Fig. 29. Trend in Dynamic Learning Coefficient of China’s PCs (1982-2002).

n

iT

iiT dDPCtaAdDPCtAP

08181 lnlnln)(lnln

57

n

iT

iiT dDPCtaAdDPCtAP

08282 lnlnln)(lnln

a2 a3 a4 d adj. R2 DW AIC

n = 2 1.91（1.04）

0.08（0.20）

-0.018（-1.08）

2.2*10-4

（0.52）0.27

（1.04）0.923 0.56 -1.73

n = 3 7.66（5.90）

-1.40（-4.53）

0.11（5.13）

-8.5*10-3

(-6.34)2.1*10-4

(6.60)0.68

（4.74）0.979 1.38 -28.30

n = 4 5.63（2.35）

-0.84（-1.33）

0.039（0.52）

-6.8*10-5

(-0.01)-2.5*10-4

(-0.55)8.9*10-6

(1.01)0.47

（1.88）0.979 1.29 -27.81

0.6

0.8

1.0

1.2

1.4

1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002

3423 10*1.210*5.811.040.1)( tttt

Fig. 30. Trend in Dynamic Learning Coefficient of Japan’s PCs (1982-2002).

Table 9 Comparison of Dynamic Learning Coefficient of Japan’s PCs (1982-2002)

58

(i) Japan’s Dynamic Learning Coefficients of PCsThe value of Japan’s coefficient is higher than that of China’s, but it decreases continually.

0.2

0.3

0.4

0.5

0.6

0.7

0.8

1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002

Dyn

amic

lear

ning

coe

ffic

ient

(λ)

Import parts dependence period

Market expansion period

Real Competitive Market Period

45342 10*5.110*3.9020.017.084.0)( ttttt

Fig. 31. Trends in Dynamic Learning Coefficient of China’s PCs(1982-2002).

(ii) China’s Dynamic Learning Coefficients of PCsa. The value of coefficient was in the bottom in 1980s, and it increased in 1990s near to the value of

Japan’s;

b. It increases when the effect of technology spillover began to decrease at the beginning of 1990s;

c. That is, “Technology of China’s PCs was accelerated by marketing learning effects after full enjoyment of spillover technology during market expansion period .”

0.6

0.8

1.0

1.2

1.4

1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002

3423 10*1.210*5.811.040.1)( tttt

Fig. 32. Trends of Dynamic Learning Coefficient of Japan’s PCs (1982-2002).

59

Interpretation by the Empirical Analysis

(1) PCs prices decreased rapidly when entering into market expansion period;

(2) China’s government reduced tariff, abolished the system of import permission, theprotection ordinance of computer software was executed, and joint-production waspermitted ( it was restricted to foreign sale by themselves in domestic market)during 1991, 1992. Foreign producers reduced their prices to expand their salesshare in the market.

(3) China’s producers reduced prices to opposite foreign producers because of poorfunctionality. However, it is limited to only depend on the effect of technologyspillover, it should expand range of learning including setting up new factories,production and distribution.

(4) Lenovo’s PCs had No.1 share of China’s domestic market through the marketinglearning in 1996.

60

2.2

1980 1983 1986 1989 1992 1995 1998 2001 2004 2007 20100.8

1.0

1.2

1.4

1.6

1.8

2.01997 Sharp

1992 Canon 1999 Matsushita (Panasonic)

2000 Hitachi

(1) Learning Efforts of Leading High-Technology Firms1. While learning coefficient has declined in Japan’s leading firms due to puffing-up of the success in the1980s

(Not Invented Here: NIH syndrome), certain high technology firms have maintained intensive learning.2. Canon can be one of the typical example which learned system LSI, SCM and cell production system from

external market.

Inflection year

Learning coefficient(2004/1990)

Canon 1992 1.25

Sharp 1997 1.01

Panasonic 1999 0.97

Hitachi 2000 0.94

Fig. 33. Learning Coefficients in 4 Electrical Machinery Firms (1980-2003).

4.5 Hybrid Management of Technology: Learning vs NIHL

earn

ing

coef

ficie

nt

61

0.00

0.10

0.20

0.30

0.40

0.50

0.60

KEY

FAN

RO

M

MU

R

CA

N

TOK

KY

C

TDK

RIC

SHA

SEI

MTE

MA

I

FLT

TOS

HLT

NEC

SAY

SON

Fig. 34. Technoprenurial Positions of 19 Electrical Machinery Firms (2001-2004).

1 CAN Canon

2 SHA Sharp

3 RIC Ricoh

4 FAN Fanuc

5 SEI Seiko Epson

6 SAY Sanyo

7 SON Sony

8 MUR Murata

9 ROM Rohm

10 KYC Kyocera

11 KEY Keyence

12 HLT Hitachi

13 MAI Matsushita (Panasonic)

14 NEC NEC

15 TOS Toshiba

16 FLT Fujitsu

17 MTE Mitsubishi

18 TDK TDK

19 TOK Tokyo Electron

(2) Bi-polarization of Technopreneurial Trajectory

Firms technology progress W depends on the ratios of (i) R&D and operating income X and (ii) operating income and sales Y.

W = F (X, Y) Taylor expansion to the secondary term, lnW = a + b lnX + c lnY + d lnX ・ lnY （a, b, c, d: coefficients）

lnW, lnX, lnY → growth rate of ＴＦＰ，R/OI （R&D expenditure to OI），OI/S (operating income to sales)

SRd

SR

ROIc

ROIba

SOI

OIRd

SOIc

OIRba

TFPTFP

1

SOI

SOITS

Mar

gina

l Pro

duct

ivity

of T

ech.

Operating Income to Sales(OIS)

Group B Group A

y

yz

z

yy

z

160.0

0.047

FAN

CAN

SHA

MAIHLITOSMTEFLTNEC

xxx

x

xx

xxx

x

x

RIC

xSON

x

TS

( )

- 0.017

xTDK

x TOK

KEYx

xROMMURx

xKYC

yz

SEISAY

Group A

Group B

OIS (2004-2007) in 19 firms

1. Japan’s leading electrical machinery firms demonstrate bi-polarization in their technopreneurial trajectories.

2. Virtuous cycle in Group A between OIS and MPT, while vicious in GroupB..

62

Organizational Inertia by Firm Size

SRSDSRTDI /ln)(/lnln 21 (i) Differences of the endeavor to technological diversification challenge due to organizational inertia by firm size

(ii) D(S) can be depicted by a logistic growth function.

TDI : Technological diversification index，R/S : R&D intensity.

0.0

0.5

1.0

1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0

売上高 (1995年実質: 兆円)

組織の慣性

24.1807.5

1ln1

1ln

11

1

1

01

0

01

S

SSSS

SS

e

e

SD

1995-1998

S0: 2.7, S1: 5.1, ε: 0.9, η: 0.0005

24.1807.511

Se

SD

Sa

PanasonicHitachiNEC

Toshiba

Fujitsu

Mitsubishi

Sharp

Sony

Canon

Sanyo

Fig. 35. Organizational Inertia Corresponding to Firm Size in Japan’s Leading 10 Electrical Machinery Firms (1995-1998).

Sales S (1995 fixed prices: trillion Yen)

Organizational inertia

index D(S)

Organizational Inertia Index

63

(3) Hybrid Management -Fusing East and West (i) Japan is emerging from years of sluggish growth (Group A firms).

(ii) Its firms appear to have produced something.(iii) Management method that incorporates lessons from US firms (West)

while preserving the practices that once made Japanese firms famous (East).

Fig. 43. Scheme of Fusion.

Close Merge Fuse

64

4.6 Necessity for Effective Learning1. Success in Hybrid Management of Technology can be attributed to learning effects (West) based on

indigenous strength (East) thereby “fusing East and West” was realized.

2. Fundamental ability distinguishing between(1) To be learned and not to be learned, and(2) To be able to learn and not able to learn.

3. For that following abilities are required:(1) Curious to know new/exotic experience,(2) Willing to be actively involved in the experience,(3) Able to distinguish harm/good, and can/cannot,(4) Able to reflect on the experience,(5) Possessing analytical skills to conceptualize,(6) Possessing decision making and problem solving skills,(7) Envisioning fusion between East and West.

Absorptive capacityAssimilation capacity

Effective utilization of spillover technology

com8 techno-economic systems institutional innovation (3

Documents