Presentation du Dr Kei Otsuka, Professeur d'Economie du Développement à l’Institut National des Hautes Etudes Politiques (GRIPS) de Tokyo, Japon.

1. Development of Industrial
Clusters and Strategy for Its
Development
Kei Otsuka
Professor of Development Economics, Kobe University
Special Adviser to Director General of IFPRI
At Strategy and Development Roundtables, Dakar
May 9, 2018
1

2. Part I
Development of Industrial
Clusters
2

3. More than 20 case studies of development of
industrial clusters in Asia and Africa
3

4. Preamble
• We conducted a large number of case studies of the
development of industrial clusters in Japan (2), China (4),
Taiwan (2), Vietnam (2), Bangladesh (2), Pakistan (1),
Ghana (1), Ethiopia (3), Kenya (2), and Tanzania (2).
• We found more similarities than dissimilarities in the
pattern of cluster-based industrial development in
different industries in different countries.
4

5. I also pay attention to cluster’s historical
development and the role of Kaizen
5

6. Purpose
To identify key factors determining the success and failure of
development of industrial clusters based on the results of case studies in
Asia and sub-Saharan Africa, as well as historical experience in Japan
and Europe.
6

7. Industrial development ≈ Cluster-based
development in manufacturing sectors
• Industrial revolution in UK: Manchester (textile),
Birmingham (steel), and Glasgow (ship-building)
• Silicon Valley and many other clusters in the US
• Development in East Asia (Japan, Taiwan, and China).
• Successful development in South Asia (Bangladesh,
India, and Pakistan). 7
Most, if not all, successful industrialization is
cluster-based not only historically but also at
present in both developed and developing
countries.

8. 8
What is Industrial Cluster?
Definition
Geographical concentration of enterprises producing
similar and closely related products in a relatively
small area (e.g., assemblers and part-suppliers).
Type 1 - Cluster that characterized by the dominance of
SMEs.
Type 2
- Pyramidal type in which
there are one assembler,
many sub-contractors, and
so many sub-sub-
contractors (e.g., Toyota).
Pyramid Type
We focus on Type 1, because it is common in
developing countries.

9. 9
Why is industrial cluster so important?
- Clustering is the key to successful industrial
development because of agglomeration economies:
1. Information spillovers
(typically imitation)
2. Specialization and division of labor among
enterprises
(low transaction costs due to proximity of transacting partners)
3. Development of skilled labor markets
(engineers, designers, and others usually through poaching)
4. Availability of useful human resources for
innovations (engineers, designers, traders, etc.)
- If an enterprise is located outside the cluster, it will find it
difficult to learn from other enterprises, to sell/buy parts
and recruit workers with desired skills.

10. How is the cluster formed?
1. Pioneering entrepreneurs initiate new business by
imitation.
2. Pioneers earn sizable profit, even though they
produce low-quality products for domestic markets.
3. High profit attracts a swarm of followers, who are
often spin-offs.
4. In this way, the cluster is formed without exception
in which SMEs use the same materials and machines,
produce the same products, and sell them at the
same domestic markets.
10

11. Two Types of industrial clusters
1. Survival cluster: Sizable entrepreneurial profit 
Entry of imitators  Formation of industrial cluster
with a large number of small enterprises producing
low-quality products  Declining prices and profits
 Stagnation
(Many clusters in SSA are this type.)
2. Dynamic cluster: Declining prices and profits 
INNOVATIONS, leading to (1) quality
improvement, (2) increasing profit, (3) firm size
expansion, (4) exit of non-innovative enterprises,
and (5) initiation of export to developed countries
(Many clusters in Asia are this type.)
11

12. 12
Number of
firms
Time
Profitability
Profitability
Number
of firms
Profitability
Number of firms
Figure 1. An Illustration of Development
Patterns of Industrial Clusters

13. Key to success is multi-faceted
innovations
a) In order to restore the profitability, the quality of products
must be improved by employing engineers, designers, and
skilled workers and by using high-quality materials, parts,
and machineries.
b) Since consumers do not immediately perceive the quality
improvement, innovative enterprises must convey the
quality information by establishing brand names, opening
own retail stores, and so on.
c) Innovative enterprises should embark on exports, expand
production, absorb non-innovative enterprises, control
quality of products, and manage a large number of workers.
13

14. Determinants of multi-faceted innovations
• Education as only educated entrepreneurs can carry
out such complicated multi-faceted innovations
• Work experience at foreign companies
• Attitude to “learn from abroad”
“Learning from abroad” is the essence of East Asian
model of development: Recognition of this is weak in
India and sub-Saharan Africa (SSA) in general.
14

15. Lessons from the development of the garment industry
in Bangladesh
15
0
100
200
300
400
500
600
700
800
0
1000
2000
3000
4000
5000
6000
1983…
1984…
1985…
1986…
1987…
1988…
1989…
1990…
1991…
1992…
1993…
1994…
1995…
1996…
1997…
1998…
1999…
2000…
2001…
2002…
2003…
2004…
2005…
2006…
2007…
2008…
2009…
2010…
No.offirmsandworkers(1,000)
No. of firms
No. of workers (1,000)
Average no. of workers per firm
AverageNo.ofworkersperfirm

16. Lessons from Bangladesh (continued)
• Leaning from abroad: Daewoo company in Korea trained 130
young Bangladeshi who graduated from universities in Korea
for 8 months about production, management, and marketing in
1979.
• In a few years, 130 workers quitted and initiated new garment
companies and trading houses.
• Because of the intensive training, the quality of products was
high from the beginning and they were all exported.
• Recognizing the importance of training, managers and workers
in Bangladesh continued to learn from abroad.
• No Daewoo-type training afterwards: Training in Korea was a
mistake of Daewoo but extremely profitable investment for the
economy as a whole.
• Thus, there is need for the governments or international
organizations to offer intensive training programs.
Establishment of Ethiopian Kaizen Institute for training of
Ethiopian consultants is superb idea. 16

17. There are many potentially dynamic clusters
in SSA
• Garment cluster in Dar es Salaam
• Metalwork cluster in Nairobi
• Metalwork cluster in Kumasi
• Leather shoe cluster in Addis Ababa
• ……………..
To our knowledge, the development of such cluster has not been
supported by the industrial development policy in SSA.
17

18. 18
Figure 2 Results of RCT of Kaizen management training in Tanzania in terms
of the adoption of number of improved management practices and value added
Kaizen score continues to rise for a while, even
among enterprises which did not take any training,
and then declined as enterprises abandon less
useful management methods.
Enterprises which took both classroom training
and on-site training continue to improve their
performance.

19. Imitation, Innovation, and Industrial Policy
1. Imitation is desirable as new idea is widely used in the society.
2. But imitation reduces incentives to innovate because private benefit
of innovation is smaller than its social benefit.
3. The ease of imitation is a major shortcoming of industrial cluster.
4. That is why, innovation must be supported either by collective action
of producers, by the industrial policy, or by both. Producer
cooperatives or associations collectively internalize the benefits of
innovations in dynamic clusters.
19

20. Part II
Strategy for Development of
Industrial Clusters
20

21. “Training-Infrastructure-Finance (TIF)
Strategy for Industrial Development in
Sub-Saharan Africa”
https://www.jica.go.jp/jica-ri/ja/publication/other/20171122_01.html
21

22. I-1. Introduction
• There is neither universally accepted effective “strategy to develop
industries” nor efforts to develop such strategy by UN, World Bank,
and most development economists.
• We propose a strategy to support industrial development in SSA,
based on Japan’s experience in East Asia and numerous findings and
emerging insights generated by our own field studies in SSA.
22

23. 23
Figure 1. A Recommended
Logical Sequence of TIF Strategy for
Industrial Development Policies

24. I-2. Unique Features of TIF Strategy
1. Focus on the existing and spontaneously developed industrial clusters,
e.g., garment, shoe, weaving, metal processing, and food processing
clusters in Senegal etc., consisting of SMEs.
2. Invest in managerial human capital of entrepreneurs (owners and
managers of enterprises).
3. Invest in infrastructure, particularly the establishment of industrial parks.
4. Financially support competent entrepreneurs.
5. Attract FDI, as foreign enterprises bring about advanced technology and
management know-how.
6. Invest in training to enhance absorptive capacity of local entrepreneurs.
We propose sequential support from 2 to 6 to realize industrial development.
24

25. I-3. Justifications
• Private enterprises do not have strong incentives to provide training because
of the labor turnover. Also entrepreneurs are ignorant of the value of
training.
• The rate of return to investment in infrastructure, including industrial parks,
is bound to be low, if there are few promising entrepreneurs.  Training of
entrepreneurs should proceed infrastructure investment.
• Training is useful not only for improving the ability of entrepreneurs but
also for identifying promising and non-promising entrepreneurs. 
Targeted support for promising entrepreneurs by admitting them to
industrial parks and providing them financial support becomes feasible only
after the training.
• In order to attract and learn from FDI, further investments in human capital
of entrepreneurs and workers are required so as to enhance their absorptive
capacities.
25

26. II. What is Missing in SSA?
• Promising industries?
-- No! There are a large number of informal industrial clusters consisting of SMEs
in SSA. They have spontaneously formed and, hence, are market-led and
consistent with comparative advantage.
-- By and large, they are stagnant, importantly because entrepreneurs do not know
how to manage their enterprises. They just survive without much profit.
-- We propose to support the development of such industrial clusters, which, we
believe, have potential to grow and become formal sectors.
• Useful knowledge?
-- Yes! In particular, what is missing is the knowledge of the efficient management
of enterprises.
* Kaizen can kill two birds (management and workers) with one stone.
26

27. Basic Ideas for TIF Strategy
1. Focus on the development of unskilled labor-intensive industrial clusters, e.g.,
textile, garment, shoes, metal processing, and simple machineries. This is
consistent with miraculous development of East Asia, which follows the flying
geese pattern of development.
2. Emphasize the importance of learning from abroad or imitation by training,
exchange programs, and visiting foreign companies. This is exactly what Japan
has been doing since the Meiji era and what other high-performing Asian
countries have been doing in the postwar periods.
3. Develop potential of promising entrepreneurs, who are likely to be highly
educated. Managerial human capital is found to be the most critical missing
factor in developing countries according to the recent studies in development
economics (Bloom et al. 2013, 2016; Sonobe and Otsuka 2011, 2014).
27

28. III-1. Major Components of Strategic Support
1. We advocate a sequential support, i.e., TIF. A critical point is the T
increases the benefit of I and F.
2. Another important point is that T confers substantial benefits to
enterprises without improving I and F according to our own studies.
3. T makes it possible to offer targeted support of I and F for
competent entrepreneurs.
4. TIF sets the stage for FDI by setting up industrial parks and
developing supporting industries, e.g., machine repairs.
5. It is a mistake to assume that once FDI is made, domestic enterprises
automatically learn advanced technology and management methods.
Further T is needed to enhance absorptive capacity.
28

29. What is Kaizen?
• Kaizen is a human friendly management philosophy and know-how that brings
about continuous, participatory, incremental, and low-budget improvement of
quality, productivity, cost, delivery, safety, morale, and environment.
• Kaizen is a collection of ideas and insights that a large number of managers and
workers have created and refined through observations and experiments carried
out over several decades in Japan and other parts of the world.
• Kaizen is designed to keep workplace clean and tidy and to find out problems and
solutions by all. Some well-known examples are 5S (sort, set in order, shine,
standardize, and sustain), Mudadori (elimination of wasteful activities), and TQC
(total quality control).
See Applying Kaizen in Africa: A New Avenue for Industrial Development, edited by
K. Otsuka, K. Jin, and T. Sonobe, forthcoming. This is fortcoming free access book.
29

30. Before the training in Dar es Salaam
30

31. After training
31

32. Some Success Stories such as
Cherie Blair, the founder of CBFW and the wife of the
former UK prime minister Tony Blair, observed how Rose
Makoyola, a participant in the GRIPS/World Bank
training program, benefited from Kaizen (1 July 2013)
32

33. III-2. Investment in Industrial Parks
• The establishment of industrial parks is likely to fail to invite domestic
enterprises to the parks, unless there are many growing profitable
enterprises looking for larger spaces. That is why we need T in the first
place.
• In order to take advantage of agglomeration economies, we recommend to
construct industrial parks to construct new industrial cluster.
• The construction of industrial parks saves investment costs in infrastructure,
such as transportation, communication, electricity, water, and sewage
facilities and it offers favorable investment climate for FDI.
33

34. III-3. FDI and beyond
34
• TIF strategy is designed to invite FDI from the beginning.
• While local enterprises should learn from foreign firms, foreign firms
try not to disclose management know-how.
• Thus, attraction of FDI is not the end of industrial policy but the
beginning of a new phase of industrial development, in which
management ability plays a key role.
• That is why continuous training is necessary for the development of
local industries in developing countries.

35. Development of agro-industrial clusters:
Is there any difference?
• Insights from sake (Japanese wine) cluster, rice milling cluster in Bouake in
Cote d’Ivoire, sun-flower oil cluster in Tanzania, and potato processing
cluster in China.
• Basic stories are the same as other clusters. Thus, it is critically important to
provide training on management as well as technology.
• One possible difference is that agro-processing enterprise has strong
incentives to offer contract farming to improve the quality of inputs.
• Contract farming: Contractor provides high-quality inputs (e.g., improved
seeds and safe pesticide) on credit, production instruction, and undertakes
marketing, so as to solve the market failures.
• Training should be offered to agro-processors and farmers. Thus, there is a
possibility of collaboration among agro-processing enterprises, farmers, and
extension workers, as well as donors and international organization.
35

36. IV. Proposal for Industrial Development in SSA
• We would like to emphasize that there are many spontaneously developed
industrial clusters in SSA, which have potential to grow.
• Kaizen is shown to be an excellent first step for development of such
clustered industries.
• When enterprises which adopt Kaizen began growing, we should support
them by allocating space in industrial parks and providing cheap loans.
• TIF strategy is designed to take advantage of complementarities among T, I,
and F.
• TIF strategy is also designed to invite FDI.
• In order to learn from foreign enterprises, continuous efforts to learn
through contractual relationships and further training programs are
absolutely necessary.
• We are confident that the application of TIF strategy will lead to successful
cluster-based industrial development in SSA.
36